Stem cells from the adult human brain develop into functional neurons in culture

Available online at www.sciencedirect.com R

Experimental Cell Research 289 (2003) 378 –383

www.elsevier.com/locate/yexcr

Stem cells from the adult human brain develop into functional neurons in culture Ulf Westerlund,a,* Morten C. Moe,b Mercy Varghese,b Jon Berg-Johnsen,b Marcus Ohlsson,a Iver A. Langmoen,a and Mikael Svenssona a

b

Department of Clinical Neuroscience, Section of Neurosurgery, Karolinska Institutet, Stockholm, Sweden Institute for Surgical Research and Department of Neurosurgery, Rikshospitalet, University of Oslo, Oslo, Norway Received 8 April 2003, revised version received 7 May 2003

Abstract Recent research communications indicate that the adult human brain contains undifferentiated, multipotent precursors or neural stem cells. It is not known, however, whether these cells can develop into fully functional neurons. We cultured cells from the adult human ventricular wall as neurospheres and passed them at the individual cell level to secondary neurospheres. Following dissociation and plating, the cells developed the antigen profile of the three main cell types in the brain (GFAP, astrocytes; O2, oligodendrocytes; and ␤-III-tubulin/ NeuN, neurons). More importantly, the cells developed the electrophysiological profiles of neurons and glia. Over a period of 3 weeks, neuron-like cells went through the same phases as neurons do during development in vivo, including up-regulation of inward Na⫹ -currents, drop in input resistance, shortening of the action potential, and hyperpolarization of the cell membrane. The cells developed overshooting action potentials with a mature configuration. Recordings in voltage-clamp mode displayed both the fast inactivating TTX-sensitive sodium current (INa) underlying the rising phase of the action potential and the two potassium currents terminating the action potential in mature neurons (IA and IK, sensitive to 4-AP and TEA, respectively). We have thus demonstrated that the human ventricular wall contains multipotent cells that can differentiate into functionally mature neurons. © 2003 Elsevier Inc. All rights reserved. Keywords: Neuronal stem cells; Adult human stem cells; Sodium channels; Potassium channels; Delayed rectifier current; Tetrodotoxin (TTX); 4-Aminopyridine (4-AP); Tetraethylammonium (TEA)

Introduction Brain tissue consists of neurons, which are responsible for carrying out its complex functional tasks, and supportive cells (astro- and oligodendrocytes). The hallmark of distinction between these two classes of cells is the functional properties at the cellular level. Although recent communications suggest the existence of stem cells in the adult human brain [1–5], this hypothesis is almost exclusively based on immunocytochemical data. Specifically, it has not been shown that these cells develop into fully functional neurons.

* Corresponding author. Department of Neurosurgery, Karolinska Hospital, 171 76 Stockholm, Sweden. Fax: ⫹46-8-5177-1778. E-mail address: [email protected] (U. Westerlund). 0014-4827/03/$ – see front matter © 2003 Elsevier Inc. All rights reserved. doi:10.1016/S0014-4827(03)00291-X

Neuronal signaling depends upon rapid changes in the electrical potential across the neuronal cell membrane. These changes are made possible by ionic channels which are able to conduct ions across the membrane at a rate of up to 108 ions/s [6]. The channels have three following important objectives: (1) to pass ions across the membrane; (2) to select specific ions; and (3) to open and close in response to specific stimuli. Although the membrane of glia cells classically was considered to have only passive K⫹-channels, it has been established that they also contain other ionic channels, including voltage-gated Na⫹ -channels [7–9]. Objective differentiation between neurons and glia can therefore not be done without a closer study of functional properties. Mature neurons characteristically fire single and multiple action potentials that are overshooting (the potential goes beyond 0 mV) and are short-lasting (⬍2 ms). The upward

U. Westerlund et al. / Experimental Cell Research 289 (2003) 378 –383

phase of the action potential is caused by a rapidly inactivating Na⫹ -current that is blocked by tetrodotoxin (TTX) [10], whereas the downward phase depends on K⫹ -channels, especially the fast inactivating IA (sensitive to 4-aminopyridine (4-AP)), and the slowly inactivating IK (sensitive to tetraethylammonium (TEA)) [11,12]. The question we asked in the present study was whether self-renewing cells from the adult human lateral ventricle wall are capable of generating not only cells with the immunological properties of neurons and glia, but also with the distinct electrophysiological properties of these cell types.

Materials and methods Biopsies were taken from four temporal lobes resected from adult patients with epilepsy. In all cases, samples were restricted to the wall of the lateral ventricle. The tissue was taken after the temporal lobe had been removed from the patient. Leibowitz-15 (L15) medium (stored at ⫹4°C) was used for transporting the biopsies from the operating theater. All patients underwent MR scans to exclude tumors and were screened for the presence of infectious disease. The research protocol was approved by the ethical committee. The harvested tissue was mechanically dissociated with a scalpel and placed in a dissociation medium, consisting of 200 ␮l DNAse (4000 U/ml, Sigma) and 500 ␮l Trypsin (1.33 mg/ml, Sigma) for 4 min in a ⫹37°C water bath, triturated with a small tip, and further incubated for 4 min. Following dissociation, the tissue was passed through a 70-␮m strainer (Falcon), centrifuged at 1100 rpm for 4 min, and resuspended as single cells in a neurosphere medium, consisting of DMEM/F12 (96%, Gibco Life Tech), Hepes buffer (1 M, 0.9%, Gibco Life Tech), B27 supplement (2%, Gibco Life Tech), EGF (0.1%, Labora), bFGF (0.1%, R&D Systems), and Penicillin-Streptokinase (100 U/ml). Cells were cultured in either a 10-cm petri dish (Falcon), a 96/ 24-well plate (Falcon), or a four-well chamber slide (Nunc) in a ⫹37°C incubator, with 6% CO2 and 20% O2. EGF, bFGF, and B27 supplement was administered to the cell cultures twice every week, and additional DMEM/F12 was administered once every week. Large neurospheres typically formed by 3–7 weeks and were dissociated with trypsine (1.33 mg/ml) for 4 ⫹ 4 min. After rinsing twice in L15 medium, single cells were resuspended in 50:50 fresh and old neurosphere medium. For adherent cultivation we used EDTA (0.02%) in PBS for 10 s, and then Trypsin-EDTA (Gibco) for 20 s to 2 min until cells were floating freely. Thereafter 5% fetal calf serum (FCS) was added before the cells were rinsed twice with L15 and resuspended in 50:50 fresh and old neurosphere medium. Differentiation of the cells was induced by adding 5% FCS and removing mitogens. Micropipettes for isolating single cells were pulled from borosilicate glass capillaries (Harvard App., GC150F-15)

379

by a pipette puller (Heka, PIP5). The micropipette was connected to a 10-mL syringe via a three-way switch for aspiration, and the contraption was maneuvered with a joystick via an Eppendorf Micro manipulator (InjectMan). Immunocytochemistry The slides were air dried for 15 min and then fixed in 4% paraformaldehyde for 10 min. After rinsing in phosphate buffer, they were incubated with 1% bovine serum albumin, 0.3% Triton X-100 (Sigma, USA), and 0.1% NaN3 in phosphate-buffered saline for 45 min in a moist chamber. The slides were then incubated with the primary antibodies O4 (Chemicon), GFAP (Dako), NeuN (Chemicon), ␤-III-tubulin (Sigma), and Nestin (gift from Prof. U. Lendahl). As secondary antibodies, we used the fluorescent markers Cy3 (1:1000, Jackson, USA), FITC (1:150, Jackson, USA), or an avidin-biotin complex (ABC-elite, 1:50, Vector, USA)/ diaminobenzidine (DAB; 50 mg/100 ml, Sigma, USA)). The slides were incubated for 1 h with the fluorescent secondary antibody (-ies), rinsed in PBS, and air dried on a ⫹50°C heater. This was followed by rinsing in xylene and mounting coverglass with DPX (DPX mounting medium, BDH Laboratory Supplies, UK). Slides that were not investigated with a fluorescent marker were incubated with biotinylated secondary antibodies for 1 h at room temperature. The ABC kit was applied for 1 h followed by incubation with DAB in Tris–HCI buffer (0.1 M, pH 7.4) containing 1% nickel sulphate and 1% cobalt chloride for 15 min and for an additional 5 min after adding 0.02% hydrogen peroxide. The slides were then rinsed in Tris–HCl buffer and gradually dehydrated with ethanol and Xylene, followed by mounting with Xylene/DPX. Neurophysiology Stem cells grown in culture dishes were placed in a recording chamber on the stage of an inverted microscope (Nikon, Tokyo, Japan). The cells were perfused with DMEM/F12 (Invitrogen Corp., UK) at 28 –32°C and bubbled with 95% air and 5% CO2. Current- and voltage-clamp recordings were made using the whole-cell patch-clamp technique [13]. Electrodes were pulled from thick-walled borosilicate glass to resistances of 4 – 6 M⍀. The pipette solution contained (in mM) 125 K-gluconate, 10 Hepes, 10 EGTA, 5 KCl, 2 Mg-ATP, 0.2 CaCl2 (pH ⫽ 7.3). An Axoclamp 2A amplifier (Axon Instruments, USA) and WinWCP software (University of Strathclyde, Scotland) were used to control pipette potentials and to inject current. Data were sampled at 10 kHz and filtered at 3 kHz through an A/D interface. Intracellular potentials were not corrected for liquid junction potentials. The measured time constant (␶in) and the input resistance (Rin) were estimated in current-clamp by the responses to square hyperpolarizing current pulses. Rin was derived from the linear portion of the current-voltage curve, and ␶in was

380

U. Westerlund et al. / Experimental Cell Research 289 (2003) 378 –383

Fig. 1. (A) Neurosphere developed by multiple divisions of a single cell picked from a preceding neurosphere. (B) Enriched culture of free floating cells following dissociation of neurosphere. (C) Cells have adhered and some have started to differentiate after adding FCS. (D) Cells were cultured for up to 11 months. Scale bar: 60 ␮m (A, D), 15 ␮m (B, C).

calculated by minimizing the squared deviation between the function and the data between 5 and 25 ms after the onset of the pulse. The voltage-clamp protocol for testing active membrane properties consisted of a 100-ms hyperpolarizing pulse from a holding potential of ⫺70 to ⫺90 mV to remove inactivation, followed by 200-ms depolarizing steps with 10-mV increments at 0.5 Hz, taking the membrane from ⫺90 to ⫹60 mV. In current-clamp, current pulses (400 ms, 0.5 Hz) were given through the patch pipette to examine whether the cells were capable of producing action potentials. External test solutions included 1 ␮M TTX to block voltage-dependent sodium channels, 1 mM 4-AP, and 5 mM TEA to block potassium currents. The results are presented as mean ⫾ SEM. Differences in passive membrane properties were tested with independentsample t-tests and considered significant when P ⬍ 0.05. Results Following dissociation the cells were kept in a serumfree environment, where single cells subsequently divided to form neurospheres, clearly visible within 2–3 weeks (Fig. 1A). The neurospheres were cultured until they reached a critical size (i.e., just before the center became necrotic). This typically took 3–7 weeks, the spheres then consisting of up to a few hundred cells. To test the cells under strict clonal conditions single cells were picked up by a micropipette, isolated, and cultivated to form further neurospheres. Single cells kept in a 96-well plate in conditioned, serumfree media gave rise to new neurospheres. To test whether cells can be propagated past one cycle, single cells were isolated from subsequent monoclonal neurospheres. Again, isolated cells gave rise to neurospheres, showing that at least two passages are possible.

Trypsination of neurospheres gave rise to free floating cells (Fig. 1B). Cells with apparent morphological changes appeared already 1 day (D1) after induced differentiation (Fig. 1C). Cells were kept in culture for up to 11 months (Fig. 1D). At different time points after induction, cells of single clones were analyzed with immunocytochemical markers. Cells at early stages of differentiation expressed nestin (Fig. 2A, red), which normally is expressed transiently in neuronal and glial precursors [14]. A few days after induction, cells positive for more mature markers appeared, and individual cells stained with either the astrocyte marker GFAP [15] (Figs. 2B and C, red), the neuron marker ␤-III-tubulin [16] (Fig. 2B, green), or the oligodendrocyte marker O4 [17] (Fig. 2C, green). Concurrently, the number of nestin-positive cells declined. Quantification of these three cellular phenotypes on D7 after a second passage (n ⫽ 719) showed 63% GFAP-, 14% O4-, and 23% ␤-III-tubulinpositive cells. ␤-III-Tubulin-positive cells (Figs. 2B, green, and 3A, red) were initially negative for the neuron-specific nuclear marker NeuN [18], but from D7 NeuN-staining started to appear (Fig. 3B) and successively increased. The cells were investigated by whole-cell patch-clamp recordings at different developmental stages (Fig. 3E). Polarized cells with an early neuron-like appearance (D7–11, n ⫽ 15) had a resting membrane potential of ⫺49 ⫾ 2 mV, an input resistance of 653 ⫾ 31 M⍀, and a membrane charging time constant of 30 ⫾ 3 ms. Compared to this, neuron-like cells that had matured somewhat further (D16 – 25, n ⫽ 15) had a more negative membrane potential (⫺58 ⫾ 3 mV; P ⬍ 0.05) and a lower input resistance (396 ⫾ 21; P ⬍ 0.005). They also tended to have a somewhat shorter time constant (24 ⫾ 2 ms; not significant). This is in keeping with observations made in the developing brain [19,20]. Cells with a glial-like development (D16 –25, n ⫽ 10) differed from their neuronal counterparts by a more negative resting membrane potential (⫺68 ⫾ 3mV, P ⬍ 0.05), a lower input resistance (84 ⫾ 12 M⍀; P ⬍ 0.005), and a much faster time constant (8 ⫾ 2 ms; P ⬍ 0.005). Depolarizing humps could be evoked in some cells with a neuron-like appearance from D9 (current-clamp; Fig. 3F, left). A few days later depolarizing pulses elicited a single immature action potential (Fig. 3F, center), and from D16 more mature-looking overshooting action potentials were observed (Fig. 3F, right). There was, however, considerable variation in the rate of development. In general, development was faster in more populated areas of the cell cultures. Ten of 15 (67%) neuron-like cells recorded from during D16 –D25 had the proper balance of channels to support generation of repetitive action potentials (Fig. 3G, left). The specific sodium channel blocker TTX [10] abolished the action potentials (Fig. 3G, right). In voltage-clamp, depolarizing pulses evoked an initial brief inward current (Fig. 3G, left inset). The current was activated at a membrane potential of about ⫺60 mV and was completely blocked by TTX (Fig. 3G, right inset). It

U. Westerlund et al. / Experimental Cell Research 289 (2003) 378 –383

381

Fig. 2. (A) Early (D1) after induced differentiation cells were positive for nestin (red), indicating immaturity. (B–C) Cells staining with GFAP (B and C; red), ␤-III-tubulin (B; green), and O4 (C; green) were observed within the first week of differentiation. Fig. 3. (A) ␤-III-Tubulin-positive cell at D7. (B) Phase contrast microscopy of a neuron-like cell at D28. The region with soma and proximal neurites is shown at larger magnification in C–D. (C) Stained for the nuclear marker NeuN (red). (D) Merge of B ⫹ C. Scale bar ⫽ 10 ␮m (A), 30 ␮m (B–D). (E) Image of a D26 cell with neuron-like morphology during whole-cell patch-clamp recording. (F) Responses to sub- and suprathreshold current pulses recorded in current-clamp at D9 (90 and 100 pA), D12 (140 and 190 pA), and D16 (70 and 80 pA). Note that the resting membrane potential becomes more negative with increasing maturity. (G) Responses to polarizing current pulses in a D20 neuron-like cell. Action potentials (left) were blocked by 1 ␮M TTX (right). The hyperpolarizing pulses were used to calculate passive membrane properties (current range ⫺33 to ⫹100 pA, left, and ⫺30 to ⫹180 pA, right). Voltage-clamp revealed an initial transient inward current (left box) that was blocked by 1 ␮M TTX (right box). (H) Voltage-clamp further revealed slower outward currents, with an initial transient component blocked by 1 mM 4-AP, and a delayed slowly inactivating component blocked by 5 mM TEA. Voltage protocol: 200-ms depolarizing steps with 10 mV increments from ⫺70 mV at 0.5 Hz. Scale bars: 5 mV and 10 ms (F), 10 mV and 30 ms (G), 400 pA and 500 ␮s (G, boxes), and 500 pA and 25 ms (H).

thus resembles the classic sodium current, INa [10]. Thirteen of 15 (87%) D16 –D25 neuron-like cells tested expressed this current. The fast inward current was followed by an

outward current that was further investigated after TTX administration. The outward current was activated between ⫺50 and ⫺40 mV and steadily increased within the range

382

U. Westerlund et al. / Experimental Cell Research 289 (2003) 378 –383

tested (Fig. 3H). It had a transient component that peaked within 20 ms and disappeared following bath application of 4-AP (Fig. 3H, middle). This current thus resembles the A-type potassium current, IA [12]. The slowly inactivating component that remained after 4-AP application was almost completely blocked by TEA (Fig. 3H, bottom), thus resembling the delayed rectifier current, IK(DR) [12]. Fourteen of 15 (93%) D16 –D25 neuron-like cells tested expressed both IK-like and IA-like currents.

Discussion Single stem cells from the adult human brain multiplied in vitro and gave rise to cells with the characteristic immunological features of neurons, astrocytes, and oligodendrocytes. Glia- and neuron-like cells had the characteristic electrophysiological features of these cell types. Specifically, we showed that neuron-like cells developed the characteristic properties of mature neurons; overshooting, shortlasting action potentials with a mature configuration. During the third week, neuron-like cells developed the ability to fire repetitive action potentials. As in mature neurons, the action potentials were blocked by TTX. The upward phase of the action potential was caused by a rapidly inactivating Na⫹ current that was blocked by TTX, consistent with INa in mature neurons [10]. We also found the two K⫹-currents that—in addition to inactivation of the sodium current— have a main role in terminating the action potential in mature neurons, one fast inactivating and sensitive to 4-AP (compatible with IA), the other slowly inactivating and sensitive to TEA (compatible with IK) [11,12]. Even though the membranes of glia cells not only contain passive K⫹ -channels, but also other ionic channels, including voltage-gated Na⫹ -channels [7–9], they do not develop the typical functional properties of neurons that we have witnessed here. The notion that the cells develop into neurons is further confirmed by our observation of that they went through the same developmental phases as neurons do in vivo. Neurons that mature under normal conditions in vivo undergo characteristic changes consisting of an upregulation of inward Na⫹-current and a drop in input resistance; at the same time as the action potential becomes shorter in duration, the membrane becomes more hyperpolarized, and the neurons become capable of firing repetitively [20,21]. In the present article we have shown that neurons derived from human stem cells undergo the same transformations during maturation. During the first days of differentiation the cells stained with nestin, a marker of immature cells [14]. By the end of the first week cells stained with specific markers for astrocytes, oligodendrocytes, and neurons. The immunocytochemical data presented here thus corroborate earlier studies which have indicated that the human brain harbors cells that may multiply and give rise to offspring with certain neuronal properties [1–5]. We did, however, rarely observe active

membrane properties at this point. Neuronal functioning developed more slowly, and mature physiological activity (overshooting, repetitive action potentials of short width) in neuron-like cells first appeared during the third week. When working with immunocytochemical markers, it is thus important to realize that one is not necessarily observing functionally active neurons. It should be kept in mind that morphological and immunocytochemical appearances are not always predictive of functional differentiation; many cells do not proceed along their developmental pathway, and the ability to reach a functionally mature state differs widely [22–25] (see also discussion in [26]). It is therefore important to investigate the electrophysiological properties of the cells. The potential use of human embryonic stem cells (ESC) in the treatment of certain neurological disorders has been investigated in experimental models, and to some extent in clinical protocols [27,28]. The utilization of ESCs has, however, not been without problems; ethical objections have been raised; the survival of transplanted neurons is poor; large amounts of human fetal mesencephalic tissue are needed for therapeutic effects; and the availability of such tissue is inadequate. Generation of neural tissue in vitro could potentially overcome some of these obstacles, and the discovery of endogenous stem cells in the adult human brain [1–5] and the developing techniques for isolation, propagation, expansion, and differentiation of stem cells [29 –32] may open novel therapeutic strategies for neurodegenerative diseases, either by transplanting stem cells that have been propagated and differentiated in vitro, or by stimulating neurogenesis in the brain. In summary the present article demonstrates that stem cells harvested from the adult human brain differentiate into fully functional neurons in vitro.

Acknowledgments We thank Carolyn Horrocks, Katarina Jansson, and Anders Ha¨ gerstrand at Neuronova AB for various and necessary help with the cell culture. This work was supported by the Swedish Research Council and the Norwegian Foundation for Health and Rehabilitation.

References [1] Y. Arsenijevic, J.G. Villemure, J.F. Brunet, J.J. Bloch, N. Deglon, C. Kostic, A. Zurn, P. Aebischer, Isolation of multipotent neural precursors residing in the cortex of the adult human brain, Exp. Neurol. 170 (2001) 48 – 62. [2] C.B. Johansson, M. Svensson, L. Wallstedt, A.M. Janson, J. Frisen, Neural stem cells in the adult human brain, Exp. Cell Res. 253 (1999) 733–736. [3] B. Kirschenbaum, M. Nedergaard, A. Preuss, K. Barami, R.A. Fraser, S.A. Goldman, In vitro neuronal production and differentiation by precursor cells derived from the adult human forebrain, Cereb. Cortex 4 (1994) 576 –589.

U. Westerlund et al. / Experimental Cell Research 289 (2003) 378 –383 [4] V.G. Kukekov, E.D. Laywell, O. Suslov, K. Davies, B. Scheffler, L.B. Thomas, T.F. O’Brien, M. Kusakabe, D.A. Steindler, Multipotent stem/progenitor cells with similar properties arise from two neurogenic regions of adult human brain, Exp. Neurol. 156 (1999) 333–344. [5] N.S. Roy, S. Wang, L. Jiang, J. Kang, A. Benraiss, C. HarrisonRestelli, R.A. Fraser, W.T. Couldwell, A. Kawaguchi, H. Okano, M. Nedergaard, S.A. Goldman, In vitro neurogenesis by progenitor cells isolated from the adult human hippocampus, Nat. Med. 6 (2000) 271–277. [6] S.A. Siegelbaum, J. Koester, Ion channels, in: E.R. Kandel, J.H. Schwartz, T.M. Jessell, (Eds), Principles of Neural Science Elsevier, New York, 1991. [7] S. Bevan, S.Y. Chiu, P.T. Gray, J.M. Ritchie, The presence of voltage-gated sodium, potassium and chloride channels in rat cultured astrocytes, Proc. R. Soc. Lond. B Biol. Sci. 225 (1985) 299 –313. [8] H. Sontheimer, J.A. Black, B.R. Ransom, S.G. Waxman, Ion channels in spinal cord astrocytes in vitro. I. Transient expression of high levels of Na⫹ and K⫹ channels, J. Neurophysiol. 68 (1992) 985– 1000. [9] A. Bordey, H. Sontheimer, Ion channel expression by astrocytes in situ: comparison of different CNS regions, Glia 30 (2000) 27–38. [10] R. Henderson, J.M. Ritchie, G.R. Strichartz, Evidence that tetrodotoxin and saxitoxin act at a metal cation binding site in the sodium channels of nerve membrane, Proc. Natl. Acad. Sci. USA 71 (1974) 3936 –3940. [11] J. Mitterdorfer, B.P. Bean, Potassium currents during the action potential of hippocampal CA3 neurons, J. Neurosci. 22 (2002) 10106 –10115. [12] J.F. Storm, Potassium currents in hippocampal pyramidal cells, Prog Brain Res 83 (1990) 161–187. [13] O.P. Hamill, A. Marty, E. Neher, B. Sakmann, F.J. Sigworth, Improved patch-clamp techniques for high-resolution current recording from cells and cell-free membrane patches, Pflugers Arch. 391 (1981) 85–100. [14] U. Lendahl, L.B. Zimmerman, R.D. McKay, CNS stem cells express a new class of intermediate filament protein, Cell 60 (1990) 585–595. [15] P.C. Barber, R.M. Lindsay, Schwann cells of the olfactory nerves contain glial fibrillary acidic protein and resemble astrocytes, Neuroscience 7 (1982) 3077–3090. [16] E.E. Geisert Jr., A. Frankfurter, The neuronal response to injury as visualized by immunostaining of class III beta-tubulin in the rat, Neurosci. Lett. 102 (1989) 137–141. [17] S.C. Barnett, A.M. Hutchins, M. Noble, Purification of olfactory nerve ensheathing cells from the olfactory bulb, Dev. Biol. 155 (1993) 337–350.

383

[18] R.J. Mullen, C.R. Buck, A.M. Smith, NeuN, a neuronal specific nuclear protein in vertebrates, Development 116 (1992) 201–211. [19] D.A. McCormick, D.A. Prince, Post-natal development of electrophysiological properties of rat cerebral cortical pyramidal neurones, J. Physiol. 393 (1987) 743–762. [20] F.M. Zhou, J.J. Hablitz, Postnatal development of membrane properties of layer I neurons in rat neocortex, J. Neurosci. 16 (1996) 1131–1139. [21] H.L. Picken Bahrey, W.J. Moody, Early development of voltagegated ion currents and firing properties in neurons of the mouse cerebral cortex, J. Neurophysiol. 89 (2002) 1761–1773. [22] E.F. Ryder, E.Y. Snyder, C.L. Cepko, Establishment and characterization of multipotent neural cell lines using retrovirus vector-mediated oncogene transfer, J. Neurobiol 21 (1990) 356 –375. [23] N.J. Sucher, N. Brose, D.L. Deitcher, M. Awobuluyi, G.P. Gasic, H. Bading, C.L. Cepko, M.E. Greenberg, R. Jahn, S.F. Heinemann, et al., Expression of endogenous NMDAR1 transcripts without receptor protein suggests post-transcriptional control in PC12 cells, J. Biol. Chem 268 (1993) 22299 –22304. [24] T.G. Hales, R.F. Tyndale, Few cell lines with GABAA mRNAs have functional receptors, J. Neurosci 14 (1994) 5429 –5436. [25] G.L. Ye, X. Song Liu, J.F. Pasternak, B.L. Trommer, Maturation of glutamatergic neurotransmission in dentate gyrus granule cells, Brain Res. Dev. Brain Res. 124 (2000) 33– 42. [26] H.J. Song, C.F. Stevens, F.H. Gage, Neural stem cells from adult hippocampus develop essential properties of functional CNS neurons, Nat Neurosci 5 (2002) 438 – 445. [27] A. Bjorklund, O. Lindvall, Cell replacement therapies for central nervous system disorders, Nat. Neurosci. 3 (2000) 537–544. [28] I.A. Langmoen, M. Ohlsson, U. Westerlund, M. Svensson, A new tool in restorative neurosurgery: Creating niches for neuronal stem cells, Neurosurgery 52 (2003) 1150 –1153. [29] V. Ourednik, J. Ourednik, K.I. Park, E.Y. Snyder, Neural stem cells— a versatile tool for cell replacement and gene therapy in the central nervous system, Clin. Genet. 56 (1999) 267–278. [30] F.H. Gage, Mammalian neural stem cells, Science 287 (2000) 1433– 1438. [31] C.Y. Liu, M.L. Apuzzo, D.A. Tirrell, Engineering of the extracellular matrix for neural stem cell programming and neurorestoration, Neurosurgery 52 (2003) 1154 –1165. [32] A. Villa, E.Y. Snyder, A. Vescovi, A. Martinez-Serrano, Establishment and properties of a growth factor-dependent, perpetual neural stem cell line from the human CNS, Exp. Neurol. 161 (2000) 67– 84.