J Pharmacol Sci 110, 133 – 149 (2009)2 Journal of Pharmacological Sciences ©2009 The Japanese Pharmacological Society Critical Review

Neurogenesis Mediated by γ-Aminobutyric Acid and Glutamate Signaling Noritaka Nakamichi1, Takeshi Takarada1, and Yukio Yoneda1,*

1Laboratory of Molecular Pharmacology, Division of Pharmaceutical Sciences, Kanazawa University Graduate School of Natural Science and Technology, Kanazawa, Ishikawa 920-1192, Japan

Received November 12, 2008

Abstract. In this review, we will summarize our ongoing studies on the functionality of both γ-aminobutyric acid (GABA) and glutamate receptors expressed by undifferentiated neural progenitor cells isolated from embryonic rodent brains. Cells were cultured with growth factors for the formation of round spheres by clustered cells under floating conditions, whereas a reverse transcription polymerase chain reaction analysis revealed expression of mRNA for particular subtypes of different ionotropic and metabotropic GABA and glutamate receptors in undifferen- tiated progenitors and neurospheres. Moreover, sustained exposure to either GABAergic or glutamatergic agonists not only modulated the size of neurospheres formed, but also affected spontaneous and induced differentiation of neural progenitor cells into particular progeny cell lineages such as neurons and astroglia. Both GABA and glutamate could play a pivotal role in the mechanisms underlying proliferation for self-replication along with the determination of subsequent differentiation fate toward particular progeny lineages through activation of their receptor subtypes functionally expressed by undifferentiated neural progenitor cells. Accordingly, neurogenesis seems to be also under control by GABAergic and glutamatergic signaling in developing brains as seen with neurotransmission in adult brains.

Keywords: neural progenitor, neurosphere, γ-aminobutyric acid (GABA), glutamate

Contents 3.3. A role of GABAAR in neurogenesis 3.4. Metabotropic GABABR signals 1. Introduction...... 134 3.5. A role of GABABR in neurogenesis 2. Isolation of neural progenitor cells ...... 134 4. Glutamatergic signaling...... 140 3. GABAergic signaling...... 135 4.1. Glutamatergic machineries 3.1. Expression of GABAergic machineries 4.2. Ionotropic NMDAR signals 3.2. Ionotropic GABAAR signals 4.3. A role of NMDAR in neurogenesis 4.4. Metabotropic GluR signals 4.5. A role of group III mGluR in neurogenesis *Corresponding author. [email protected] Published online in J-STAGE on May 30, 2009 (in advance) 4.6. Adult neurogenesis doi: 10.1254/jphs.08R03CR 5. Amino acid signaling to neurogenesis ...... 146

Invited article

Abbreviations used (in alphabetical order): AMPA, DL-α-amino-3-hydroxy-5-methylisoxasole-4-propionate; AMPAR, AMPA receptor; AP1, activator protein-1; ATRA, all-trans-retinoic acid; bFGF, basic fibroblast growth factor; bHLH, basic helix-loop-helix; BrdU, 5-bromo-2'- deoxyuridine; CNTF, ciliary neurotrophic factor; CPPG, RS-α-cyclopropyl-4-phosphonophenyl glycine; CRE, cAMP responsive element; CREB, CRE-binding protein; DCG-IV, (2S,2'R,3'R)-2-(2',3'-dicarboxycyclopropyl)glycine; DG, dentate gyrus; DHPG, 3,5-dihydroxyphenylglycine; EGF, epidermal growth factor; FBS, fetal bovine serum; GABA, γ-aminobutyric acid; GABAAR, GABAA receptor; GABABR, GABAB receptor; GABACR, GABAC receptor; GFAP, glial fibrillary acidic protein; iGluR, ionotropic glutamate receptor; KA, kainate; KAR, KA receptor; L-AP4, L-2-amino-4-phosphonobutyrate; LDH, lactate dehydrogenase; MAP2, microtubule-associated protein-2; mGluR, metabotropic glutamate receptor; MK-801, dizocilpine; MTT, 3-(4,5-dimethyl-2-thiazolyl)-2,5-diphenyl-2H-tetrazolium bromide; NCAM, neural cell adhesion molecule; NMDA, N-methyl-D-aspartate; NMDAR, NMDA receptor; PACAP, pituitary adenylyl cyclase activating peptide; PCNA, proliferating cell nuclear antigen; PI, propidium iodide; PKA, protein kinase A; PTSD, posttraumatic stress disorder; RAR, retinoic acid receptor; RT-PCR, reverse transcription polymerase chain reaction; RXR, retinoid X receptor; SVZ, subventricular zone; WIRS, water-immersion restraint stress.

133 134 N Nakamichi et al

1. Introduction our recent findings on the importance of signals mediated by different GABAergic and glutamatergic The prevailing view is that neural stem cells are receptor subtypes in the mechanisms related to self- primitive progenitor cells with abilities to proliferate for replication and multi-potentiality in undifferentiated self-renewal and to differentiate into different progeny neural progenitor cells isolated from fetal rodent brains. cell lineages including neurons, astroglia, and oligo- In 1991, Meier et al. published a pioneering review on dendroglia. These progenitor cells are enriched in the brain development by neurotransmitters (25). SVZ during development of the brain (1, 2), while in the adult brain, progenitor cells are also localized in the DG 2. Isolation of neural progenitor cells of the hippocampus as well as the SVZ (3 – 8). In contrast to other structures in the adult brain, the Mitotic cells can be clearly separated from postmitotic subgranular cells spontaneously undergo proliferation cells in embryonic brains at the late stage by using and migration to the granule cell layer of the DG (9). The density gradients generated with colloidal silica (26, 27). fact that these progenitors undergo cellular proliferation, Fetal postmitotic cells have a buoyant density lower than commitment, and differentiation into neurons and glia in 1.043 g/ml, whereas mitotic and progenitor cells have vitro (10) is suggestive of the proposal that cells are densities higher than 1.056 g/ml (28). In these previous derived from multi-potential neural stem cells (11). On studies, cells are derived from stem cells and contain a transplantation of neural stem cells into brains, cells mixture of lineages at various stages of differentiation. develop into mature neurons or glia with morphological The most immature cells have remarkably high densities and biochemical features similar to those of neighboring ranging from 1.065 to 1.075 g/ml, with ability to cells, which gives rise to the idea that cellular commit- differentiate to mature cells. The present preparation ment and/or differentiation is at least in part under procedures for neural progenitor cells are in principle control by microenvironments around the stem cells in based on the aforementioned differential densities of vivo (12). cells at different developmental stages. Emerging evidence that endogenous factors regulate In the upper cell layer prepared by Percoll density self-renewal capacity and multi-potentiality of neural gradient centrifugation, particular cells were immuno- progenitors expressed in the developing and matured reactive for either the neuronal marker MAP2 or the brains is available in the literature. These endogenous astroglial marker GFAP, but not for the neural pro- factors are believed to be derived from a variety of genitor marker nestin (29). In the lower cell layer, adjacent cells around the stem cell niche (13 – 16). several cells were immunoreactive for nestin, but Several extracellular molecules, such as growth factors expression of either MAP2 or GFAP was not detected. and neurotransmitters, have been implicated in the The lower cell layer was used as a source of progenitors extrinsic regulation of cell proliferation in the develop- expressing nestin and cultured in the presence of a ing telencephalon. For example, bFGF prolongs pro- growth factor for subsequent double immunocyto- liferation of progenitor cells with a concomitant increase chemical detection of MAP2 and GFAP. In the presence in the number of neurons in rat neocortex when either of EGF, round spheres were formed within 4 days with added to cultured cells in vitro (17) or microinjected into increasing cluster sizes proportional to culture durations embryonic rat brains in vivo (18). The neurotransmitters up to 10 days. In these neurospheres cultured for GABA and glutamate are shown to reduce the number of 10 days, cells were immunoreactive for nestin but not proliferating cells in dissociated or organotypic cultures for either MAP2 or GFAP. of rat neocortex (19). GABA partially blocks the bFGF- Floating neurospheres were then dispersed and the induced increase in cell proliferation (20), but promotes cells were seeded on wells previously coated with poly- cell proliferation in cultures of rat cerebellar progenitors L-lysine and subsequently cultured in the absence of a (21). Signals mediated by ionotropic GABAAR are shown growth factor for an additional several days toward to promote neuronal differentiation after depolarization spontaneous differentiation. Removal of EGF led to a in neural progenitors of adult mouse hippocampal slices complete abolition of immunoreactive nestin together (22). In our previous study, by contrast, sustained with a drastic increase in the number of cells immuno- exposure to a GABAAR agonist leads to increased reactive for either MAP2 or GFAP. Dispersed cells were proliferation along with facilitation of commitment also cultured for an additional 4 days in the presence of and/or differentiation toward an astroglial lineage in the either the neuronal inducer ATRA or the astroglial presence of CNTF in neural progenitor cells isolated inducer CNTF to promote the commitment toward a from the neocortex of embryonic rats (23) and mice (24). neuronal or astroglial lineage, respectively, after the In this review, therefore, we will mainly summarize removal of EGF. Sustained exposure to ATRA markedly Amino Acid in Neurogenesis 135

increased the number of cells immunoreactive for MAP2 GABABR1a, 1b, and GABABR2 subunits of GABABR compared to that found in the presence of CNTF. and for the ρ1, ρ2, and ρ3 subunits of GABACR, but Repetition of the dispersion and culture led to similar not for the α6 or γ1 subunit of GABAAR, on RT-PCR expression profiles of different immunoreactive marker analysis. In neocortical neurospheres grown for 10 days, proteins (23, 30). Accordingly, the lower cell layer mRNA expression was seen for the α2, α3, α4, α5, β1, indeed contains neural progenitor cells endowed to β2, β3, γ1, γ2, and γ3 subunits of GABAAR and proliferate for self-replication in a manner sensitive to a GABABR1a, 1b and GABABR2 subunits of GABABR growth factor and to differentiate into neuronal and and for the ρ1 and ρ2 subunits of GABACR, but not for astroglial lineages in response to the respective differen- the α1, α6, and δ subunits of GABAAR or the ρ3 subunit tiation inducers. of GABACR. Moreover, in the lower cell layer prepared from fetal mouse neocortex, mRNA expression was seen 3. GABAergic signaling for mGAT1, mGAT3, and mGAT4, but not mGAT2, isoforms of GABA transporters; vesicular GABA In signal transduction mechanisms mediated by the transporter required for the condensation of GABA; inhibitory neurotransmitter GABA, extracellular signals and both GAD65 and GAD67 isoforms of glutamate are transformed into intracellular signals through decarboxylase responsible for the synthesis of GABA ionotropic (GABAAR and GABACR) and metabotropic from glutamate. Similar expression profiles were found (GABABR) GABAR subtypes in the brain. The GABAAR with mRNA for these GABAergic signaling molecules, subtype is orchestrated by the heteromeric assembly of except for the mGAT2 isoform, in neurospheres cultured 5 different proteins, including α, β, and γ subunits, for 10 days. toward anion channels permeable to chloride ions, whereas the GABACR subtype is formed from the 3.2. Ionotropic GABAAR signals homomeric assembly of 5 ρ subunits for chloride 3.2.1. Proliferation mediated by GABAAR: In order channels. In contrast to other metabotropic receptors, to evaluate the role of GABAR expressed by undifferen- the metabotropic GABABR subtype is a heteromeric tiated progenitors, neurosphere formation was examined dimer from the assembly between GABABR1 and in cells cultured with the endogenous agonist GABA at GABABR2 subunits with the functional linkage to the different concentrations for a period of up to 10 days. inhibitory Gi/o protein to negatively regulate the activity The size of clustered neurospheres was drastically of adenylyl cyclase, which in turn leads to a decrease increased in proportion to the culture period from 4 to in intracellular concentrations of cAMP, opening of 10 days, while sustained exposure to GABA significantly potassium channels, and closing of calcium channels. increased the size of neurospheres in a concentration- GABAergic input is shown to partially block the bFGF- dependent manner at concentrations of 10 to 300 μM induced increase in cell proliferation (20), but promotes from 6 to 10 days. The increase by GABA was signifi- cell proliferation in cultures of rat cerebellar progenitors cantly prevented by either the GABAAR antagonist (21). Signals mediated by ionotropic GABAAR subtype bicuculline or the GABABR antagonist CGP54626. No promote neuronal differentiation after depolarization in significant difference was seen in the release of LDH neural progenitors of adult mouse hippocampal slices from neocortical neurospheres cultured for 10 days into (22). In our previous study, by contrast, sustained culture medium irrespective of sustained exposure to activation of the GABAAR subtype leads to increased any GABAR ligand. A negligibly small number of cells proliferation along with subsequent facilitation of were reactive with the membrane-impermeable dye PI commitment and/or differentiation toward an astroglial for DNA staining, but most cells were stained with the lineage in the presence of CNTF in neural progenitor membrane-permeable dye Hoeschst 33342. cells isolated from the fetal rat neocortex (23). More- Although neurospheres were proportionally increased over, we have attempted to evaluate the possible with days in culture, the GABAAR agonist muscimol modulation of the functionality by the metabotropic significantly increased the neurosphere size in a manner GABABR subtype expressed in neural progenitor cells sensitive to the GABAAR antagonist bicuculline. prepared from fetal mouse neocortex under the similar Sustained exposure to bicuculline alone significantly conditions (31). and concentration-dependently inhibited the cluster formation in the concentration range of 0.1 – 50 μM in 3.1. Expression of GABAergic machineries neurospheres, with concomitant significant inhibition In the neocortical lower cell layer before culture, of MTT reduction at 10 – 50 μM. The incorporation of mRNA expression was seen for the α1, α2, α3, α4, α5, the thymidine analog BrdU was found in neurospheres β1, β2, β3, γ2, γ3, and δ subunits of GABAAR, cultured in the presence of EGF for 6 days, and sustained 136 N Nakamichi et al exposure to muscimol increased the number of cells sustained exposure to muscimol induced a significant immunoreactive for BrdU in a manner sensitive to increase in the number of GFAP-positive cells with a bicuculline. Bicuculline alone decreased the number of concurrent significant decrease in the number of MAP2- cells immunoreactive for BrdU. Repetition of these positive cells upon differentiation induced by CNTF experiments for quantitative analysis confirmed the only. No significant changes were seen in the numbers significant increase in the number of BrdU-positive cells of cells immunoreactive for either MAP2 or GFAP in in neurospheres exposed to muscimol and the prevention cells with prior exposure to muscimol after spontaneous by bicuculline. Therefore, tonic activation of GABAAR and ATRA-induced differentiation. Prior exposure to would lead to a significant increase in proliferation muscimol decreased MAP2 expression and increased activity in neural progenitor cells before commitment GFAP expression in cells differentiated by CNTF, but without affecting cellular viability. bicuculline alone induced an increase in MAP2 expres- 3.2.2. Differentiation mediated by GABAAR: Re- sion with a decrease in GFAP expression. Quantitative moval of the growth factor EGF led to an appearance of calculation showed that muscimol significantly increased cells with marked expression of either MAP2 or GFAP GFAP expression with decreased MAP2 expression in 4 days later, with a concomitant complete abolition of the presence of CNTF in a manner sensitive to immunoreactive nestin. For the induction of cellular bicuculline. Bicuculline alone was effective in signifi- differentiation, dispersed cells were cultured for an cantly increasing MAP2 expression and decreasing additional 4 days in the presence of either ATRA or GFAP expression. CNTF to facilitate commitment toward a neuronal 3.2.3. Expression of CNTF receptor: Sustained ex- lineage or an astroglial lineage, respectively. Sustained posure to muscimol alone resulted in a significant and exposure to ATRA for 4 days increased the number predominant increase in mRNA expression of CNTF-α of cells immunoreactive for MAP2 compared to that receptors in undifferentiated neurospheres, without found in its absence. CNTF was effective in increasing significantly affecting that of other relevant molecules. the number of cells immunoreactive for GFAP and These included N-cadherin; β-1 integrin; NCAM; RAR- decreasing that for MAP2. In cells previously exposed α, -β, and -γ; and RXR-α. Simultaneous exposure to to muscimol for 10 days, a marked decrease was seen in bicuculline significantly prevented the increase in CNTF- MAP2 expression with a concomitant increase in GFAP α–receptor mRNA expression induced by muscimol, expression when cell differentiation was induced by but was ineffective in altering mRNA expression of CNTF. Previous exposure to muscimol did not markedly other molecules tested. affect the expression of either MAP2 or GFAP in differentiation induced by ATRA as well as spontaneous 3.3. A role of GABAAR in neurogenesis differentiation. As mentioned above, tonic activation of GABAAR Quantification of these data revealed that around led to a significant increase in the capabilities to form 50% of the cells were immunoreactive for GFAP and neurospheres and to express the astroglial marker GFAP around 30% for MAP2 after spontaneous differentiation. in response to CNTF through upregulation of CNTF- Although cells not immunoreactive for either MAP2 or receptor expression in cultured neural progenitor cells GFAP were almost 20% of the total cells in spontaneous (Table 1). Significant inhibition was also seen in the differentiation, ATRA increased the percentage of expression of the neuronal marker protein MAP2 in MAP2-positive cells with concomitant decreases in the presence of CNTF in cells previously exposed to those of GFAP-positive cells and cells not immuno- muscimol, but muscimol failed to significantly affect reactive for either marker protein. By contrast, CNTF the expression of both GFAP and MAP2 in the presence decreased the percentage of MAP2-positive cells along of ATRA. Our data clearly confirm the proposal that with an increase in that of GFAP-positive cells. Prior functional heteromeric GABAAR channels are constitu-

Table 1. GABAAR signaling to neurogenesis Differentiation Spontaneous Induced Proliferation ATRA CNTF NEUASTNEUASTNEUAST promoted no effect no effect no effect no effect suppressed promoted

Abbreviations: AST, astrocyte; NEU, neuron. Amino Acid in Neurogenesis 137 tively expressed to play a role in mechanisms related to differentiate into type-II astrocytes, which is specifi- to both self-replication and multi-potentiality in un- cally mediated via the signal transducer and activator differentiated neural progenitor cells before commit- of transcription 3 (33). ATRA is shown to regulate ment to any neural lineages. Activation of GABAAR differentiation and/or proliferation of the progeny cells would play a key role in the mechanisms associated with derived from neural stem cells in the CNS. Under proliferation, commitment, and differentiation of neural conditions that promote differentiation, cultures of progenitor cells at an early developmental stage in dissociated neurospheres show a dose-dependent increase embryonic neurogenesis in both mouse and rat brains. in the number of neurons in response to ATRA (34). Evidence is accumulating for the functional expres- These inducers could regulate both proliferation and sion of GABAAR in neural progenitor cells before differentiation of neural stem cells after commitment. commitment and differentiation. In the ventricular zone The presence of CNTF would be favorable for commit- of rat embryonic neocortex, GABA induces depolariza- ment to astroglial progenitors and that of ATRA for tion with decreased DNA synthesis through activation neuronal progenitors, respectively. Moreover, FBS was of GABAAR (19). Activation of GABAAR not only leads shown to increase the number of living cells as well as to the inhibition of proliferation of cortical progenitors the number of astrocytes during differentiation, but in the presence of bFGF with concomitant facilitation of bFGF facilitates proliferation with reduced differentia- differentiation to neurons (20), it also promotes neuronal tion to neurons in neural stem cells (35). These previous differentiation after depolarization in neural progenitors findings could account for marked expression of MAP2 of adult mouse hippocampal slices (22). Although these even in the presence of CNTF in cells previously previous findings disagree with the present significant exposed to muscimol. A large proportion of the neural increase in proliferative activity of neurospheres by progenitor cells would be already rendered to differ- tonic activation of GABAAR, GABA was also shown to entiate into neurons during culture under the conditions stimulate proliferation of cultured immature granule favorable for cell proliferation. Culture conditions cells of the rat cerebellum (21). Opposite effects of would be a determinant crucial for the fate of commit- GABA are also reported in the proliferation of neural ment to neuronal or astroglial progenitors in neural progenitors expressed in the ventricular zone and SVZ progenitor cells isolated from fetal mouse neocortex. of mouse neocortex (32). Exposure to GABA increases Neural stem cells commit to differentiation into specific neural progenitor proliferation in the ventricular zone, functional cells in subregions of the brain at an early but significantly decreases that in the SVZ, in a manner developmental stage in vitro (36 – 38). The fact that sensitive to a GABAAR antagonist in mouse organotypic muscimol did not affect spontaneous differentiation of slice cultures. The fact that extracellular environments neural progenitor cells toward the neuronal or astroglial are highly responsible for commitment of stem cells in lineage could be at least in part due to early commitment vivo (12) would account for the paradox. The abundance to specific cells during culture. The data cited above of cells already differentiated, such as neurons and are all suggestive of the proposal that GABAAR signals astroglia, is a determinant of proliferation and differenti- may facilitate the CNTF-induced commitment of neural ation activities of adjacent undifferentiated neural pro- progenitor cells into astroglia. genitors in vivo and in vitro. From this point of view, We have for the first time demonstrated that tonic the present technique has the advantage that extra- activation of GABAAR leads to the predominant up- cellular environments are artificially defined by the use regulation of CNTF-receptor expression during the of neural progenitor cells isolated from embryonic culture with EGF in undifferentiated neural progenitor brains under the in vitro culture conditions. In the cells of fetal mouse neocortex (24). Although differential absence of the influences by different endogenous expression profiles are shown with CNTF and CNTF factors in extracellular environments, activation of receptor-α in astrocytes and neurons of the fascia dentate GABAAR would lead to the stimulation of proliferation after the entorhinal cortex lesion (39), the exact underly- activity of undifferentiated neural progenitor cells toward ing mechanism is still unclear. An alternative possibility self-replication and subsequent astroglial commitment that activation of GABAAR would stimulate the expres- in the embryonic neocortex. sion of particular adhesion molecules responsible for Our studies confirm the promotion of subsequent commitment and/or differentiation toward an astroglial commitment and differentiation induced by CNTF toward lineage in response to CNTF in undifferentiated neural the astroglial lineage of undifferentiated neural pro- progenitor cells is not conceivable if you take into genitor cells isolated from embryonic rat (23) and mouse consideration the findings cited above. Neural pro- (24) neocortex through prior tonic activation of genitor cells are shown to express different adhesion GABAAR. CNTF is required for neural progenitor cells molecules, including NCAM, cadherin, and integrin 138 N Nakamichi et al

α5β1, α6aβ1, αvβ1, αvβ8, α5β1, and α6bβ1 (a low level) alone and the prevention by CGP54626. (40). The cell adhesion molecules NCAM and the Cells were cultured for 10 days in the presence of integrins play an important role in the mechanisms pertussis toxin, which is responsible for the uncoupling underlying cellular proliferation (40), migration (41), of GABABR to inhibitory Gi/o proteins toward the nega- and differentiation (42) in neural progenitor cells. Notch tive regulation of adenylyl cyclase. Sustained exposure is a signal of differentiation in cell adhesion (43), but to the toxin significantly decreased the neurosphere sustained activation of Notch inhibits both commitment size at a concentration range of 0.02 to 1.0 μg/ml. MTT and differentiation to either neurons or glia in neural reduction was slightly but significantly increased in progenitor cells (44). An AP1 site exists on the promoter neurospheres cultured in the presence of baclofen alone, regions of both NCAM (45) and integrin α6 (46), but decreased in those exposed to CGP54626 alone. No whereas the administration of an NMDAR antagonist significant difference was seen in the release of LDH up-regulates NCAM expression in granule cells of the from neurospheres into culture medium irrespective of adult rat DG (47). The interaction between CNTF and the sustained exposure to baclofen or CGP54626. A CNTF receptor-α leads to heightened expression of negligibly small number of cells were reactive with the connexin 43 mRNA through the Janus tyrosine kinase membrane-impermeable dye PI for DNA staining in /STAT pathway in cultured murine cortical astrocytes neurospheres exposed to a GABABR ligand, but most (48). The exact mechanism as well as the physiological cells were stained with the membrane-permeable dye significance of the upregulation of CNTF receptors, Hoechst 33342. Therefore, tonic activation of GABABR however, remains to be elucidated in future studies. It would lead to a significant increase in proliferation thus appears that functional heteromeric GABAAR activity in neural progenitor cells before commitment channels are constitutively expressed in undifferentiated and/or differentiation without affecting cellular viability. neural progenitor cells for the regulation of subsequent 3.4.2. Proliferation of GABABR1-null progenitors: commitment and differentiation to the astroglial lineage Neurospheres were prepared from whole brain of fetal in the developing brains. mice defective of the gene for the GABABR1 subunit essential for the orchestration of functional GABABR 3.4. Metabotropic GABABR signals (31). Genotyping by RT-PCR clearly revealed the 3.4.1. Proliferation mediated by GABABR: The complete absence of GABABR1-subunit mRNA from increase by GABA was significantly prevented by particular embryos delivered after the mating of mice simultaneous exposure to the GABABR antagonist with genetically heterozygous genes. Double immuno- CGP54626 with regards to the size of neurospheres cytochemical analysis using the antibody against and MTT reduction in neurospheres. Both proliferation GABABR1 subunit or GABABR2 subunit together with indices were significantly inhibited in neurospheres the anti-nestin antibody confirmed the complete absence exposed to CGP54626 alone, while GABAR ligands of GABABR1 subunit protein from these neurospheres failed to significantly alter the number of PI-positive prepared from GABABR1-null mice with expression of cells in the neurospheres. The GABABR-subtype agonist nestin protein. baclofen significantly increased MTT reduction and The size of clustered cells was invariably smaller neurosphere size in a manner sensitive to a GABABR- with neurospheres prepared from GABABR1-null mice subtype antagonist. The GABABR antagonist CGP54626 (gb1−/−) than those prepared from heterozygous significantly prevented the increase in neurosphere size (gb1+/−) and wild-type (gb1+/+) mice. In fact, the size by baclofen. In order to confirm the expression of the of clustered cells was significantly smaller in neuro- heteromeric functional GABABR subtype, neurospheres spheres prepared from GABABR1-null mice (gb1−/−) cultured for 10 days were subjected to double immuno- than in those from heterologous (gb1+/−) and wild-type staining for either the GABABR1 or GABABR2 subunit (gb1+/+) mice during culture for 4 to 10 days. The along with nestin. Both GABABR1 and GABABR2 GABABR antagonist CGP54626 significantly decreased subunits were clearly co-localized to cells expressing the size of neurospheres prepared from wild-type mice nestin (31). Baclofen markedly increased the number of without significantly affecting that in GABABR1-null cells immunoreactive for BrdU in a fashion sensitive mice. Accordingly, GABABR could be functionally to CGP54626 when exposed to neurospheres cultured expressed by undifferentiated neural progenitors before for 6 days. Moreover, CGP54626 alone profoundly cellular commitment and/or differentiation to stimulate decreased the number of cells immunoreactive for BrdU. proliferation activity toward self-replication in fetal Repetition of these experiments for quantitative analysis mouse brain. clearly confirmed a significant increase in the number of 3.4.3. Differentiation of GABABR1-null progenitors: BrdU-positive cells in neurospheres exposed to baclofen Neural progenitor cells were prepared from GABABR1- Amino Acid in Neurogenesis 139 null mice, followed by culture in the presence of EGF of both RAR-α and RAR-γ, with those of the other for 8 to 12 consecutive days and subsequent dispersion molecules being unchanged. after the removal of EGF to initiate spontaneous differentiation. Irrespective of the presence of the 3.5. A role of GABABR in neurogenesis GABABR1-subunit gene, cells cultured for an additional Tonic activation of GABABR led to a significant 4 days were highly immunoreactive to either the neuronal increase in the capability to form neurospheres in marker MAP2 or the astroglial marker GFAP. Both the cultured neural progenitor cells isolated from fetal size and shape of the GFAP-positive cells of GABABR1- mouse brain (Table 2). The data from GABABR1-null null mice were somewhat different from those of wild- mice give support to this proposal. We have already type mice under the culture conditions employed. On shown that sustained activation of GABAAR leads to spontaneous differentiation, almost 40% of the cells increased proliferation activity for self-renewal along were immunoreactive to either MAP2 or GFAP in with promoted differentiation into an astroglial lineage neurospheres prepared from wild-type animals irrespec- in response to CNTF (23, 24). Therefore, GABA could tive of the culture duration. In cells prepared from play a pivotal role in the positive regulation of cellular GABABR1-null mice, however, a significant decrease proliferation through activation of both GABAAR and was seen in the number of cells immunoreactive to GABABR expressed by undifferentiated neural pro- MAP2 with a significant increase in the number of cells genitors endowed with self-replication and multi- immunoreactive to GFAP on spontaneous differentiation. potentiality activities. Furthermore, our findings from The addition of ATRA markedly increased the experiments using GABABR1-null mice argue in favor number of cells expressing MAP2 with a concomitant of the idea that tonic activation of GABABR would lead decrease in the number of GFAP-expressing cells in to the promotion of both cellular proliferation and preparations from wild-type mice, while CNTF decreased subsequent differentiation toward a neuronal lineage MAP2 expression with a concurrent increase in GFAP irrespective of the presence of differentiation inducers expression. In cultures prepared from GABABR1-null in neural progenitors. At any rate, GABABR as well as mice, a significant decrease was seen in the number of GABAAR would play a key role in the mechanisms cells immunoreactive to MAP2 with a significant associated with both proliferation and differentiation of increase in the number of cells immunoreactive to GFAP neural progenitor cells at an early developmental stage on differentiation by ATRA. In cells after differentiation in embryonic neurogenesis in the brain. The possible dif- induced by CNTF, knockout of the GABABR1 gene ferences in animal species and/or ages would at least in failed to significantly affect the number of cells part account for the paradoxical data on the modulation immunoreactive to either MAP2 or GFAP. Accordingly, of neural progenitors by different signal mediators in GABABR could play a role in the mechanisms related the literature. Although the inhibition by a GABABR to subsequent differentiation into neurons and astroglia antagonist alone clearly gives support to the presence in mouse neural progenitor cells. of endogenous GABA released from neurospheres pre- We analyzed the expression profiles of mRNA for pared from embryonic brains, the possibility that adhesion molecules and differentiation inducer receptors GABABR displays low-level and ligand-independent in the neurospheres. There was marked expression of background activity in a manner sensitive to antagonists the mRNAs for RAR-α, RAR-β, RAR-γ, RXR-α, CNTF as seen in other ionotropic receptors is not excluded. receptor-α, N-cadherin, β-1 integrin, NCAM, and However, the effective inhibition by CGP54626 alone laminin-β1 in the undifferentiated neurospheres pre- would account for the weak promotion by baclofen of pared from wild-type mice. In neurospheres prepared proliferation in neural progenitor cells, in contrast to from GABABR1-null mice, however, a slight but the full efficiency of GABA that is an endogenous significant increase was seen in the mRNA expressions agonist for both GABAAR and GABABR subtypes.

Table 2. GABABR signaling to neurogenesis Differentiation Spontaneous Induced Proliferation ATRA CNTF NEUASTNEUASTNEUAST promoted promoted suppressed promoted suppressed promoted suppressed

Abbreviations: AST, astrocyte; NEU, neuron. 140 N Nakamichi et al

To our knowledge, this is the first direct demonstra- for subsequent activation of protein kinase C and release tion of a significant increase in mRNA expression of of Ca2+ from intracellular stores, respectively. Both both RAR-α and RAR-γ in neural progenitor cells pre- group II and III mGluR subtypes are coupled to the pared from fetal mice devoid of the GABABR1 subunit. inhibitory Gi/o protein to negatively regulate the activity In rat neural progenitors that have no GABABR2-subunit of adenylyl cyclase, which decreases intracellular con- mRNA expression, prior tonic activation of GABAAR centrations of cAMP (53 – 56). induces the promotion of subsequent differentiation induced by CNTF toward an astroglial lineage (23). In 4.1. Glutamatergic machineries neural progenitor cells of the embryonic rat spinal In the rat neocortical lower cell layer before culture, neural tube ventricular zone (49), activation of RAR-α mRNA expression was seen for NR1, NR2A-2C, and leads to the predominant formation of both astrocytes NR2D subunits of the NMDAR subtype; GluR1 – 4 and oligodendrocytes with the formation of moto- subunits of the AMPAR subtype; and GluR5, GluR6, neurons by activation of RAR-β, which is in good agree- GluR7, KA1, and KA2 subunits of the KAR subtype on ment with the present findings on expression profiles of RT-PCR analysis (57). In neocortical neurospheres RAR isoforms, MAP2, and GFAP in response to ATRA cultured for 12 days, mRNA expression was similarly in GABABR1-null mouse progenitor cells. The data seen for NR2A-2C subunits of NMDAR; GluR1 – 4 cited above are all suggestive of the proposal that signals subunits of AMPAR; and GluR5, GluR6, KA1, and KA2 mediated by GABABR may facilitate differentiation of subunits of KAR, but not for NR1 and NR2D subunits neural progenitor cells into neurons. It thus appears that of NMDAR and the GluR7 subunit of KAR. In contrast functional heteromeric GABABR could be constitutively to the temporal expression profile of mRNA for expressed by undifferentiated neural progenitor cells for group III mGluR isoforms, transient expression was the regulation of cellular proliferation and subsequent seen with the mRNA for the NR1 subunit absolutely differentiation in fetal mouse neocortex. Taken together, essential for the heteromeric assemblies toward the one fascinating but hitherto unidentified speculation is functional NMDAR channels in neurospheres cultured that self-replication and multi-potentiality would be for 2 to 6 days with subsequent disappearance during under control by delicate balancing between ionotropic culture with the growth factor bFGF under floating and metabotropic GABAergic signals through the respec- conditions. In dispersed cells cultured for an additional tive receptors expressed by neural progenitor cells 6 days after the removal of bFGF, however, marked before differentiation as seen in matured neurons. In expression was seen with NR1-subunit mRNA at a level undifferentiated progenitors, GABA could positively similar to that found in pre-neurospheres before culture. regulate the proliferation through ionotropic GABAAR By contrast, NR2A subunit mRNA was not expressed in and metabotropic GABABR subtypes, but differentially pre-neurospheres before culture with a gradual increase modulate subsequent differentiation to particular progeny proportional to the culture duration from 2 to 12 days. lineages. Constitutive expression was seen for NR2B-subunit mRNA in neurospheres cultured for up to 12 days with a 4. Glutamatergic signaling sudden threefold increase between 2 and 4 days. A gradual decrease was found in the expression of mRNA Glutamatergic signals are mediated by iGluR and for both NR2C and NR2D subunits during the culture mGluR glutamate receptors in the developing brain periods of up to 12 days. Marked mRNA expression was and matured brain. iGluRs are categorized into three seen for all NMDAR subunits examined in dispersed different cation channel subtypes according to the cells cultured after the removal of bFGF. nucleotide sequential homology and intracellular signal- Sustained exposure to either the NMDAR agonist ing systems. These include NMDAR, AMPAR, and NMDA or the group III mGluR agonist L-AP4 led to a KAR subtypes. By contrast, mGluRs are a family of significant decrease in MTT reduction, without signifi- type III G-protein–coupled receptors activated by cantly affecting LDH release during culture. However, glutamate and divided into three major subtypes other GluR agonists did not significantly affect either (group I: mGluR1 and 5 isoforms; group II: mGluR2 and MTT reduction or LDH release even when persistently 3 isoforms; group III: mGluR4, 6, 7, and 8 isoforms) exposed to neurospheres for 12 days. These included based on sequence homology, signal transduction the AMPAR agonist AMPA, the KAR agonist KA, the pathway, and pharmacology (50 – 52). The group I group I mGluR agonist DHPG, and the group II mGluR mGluR subtype is coupled to stimulatory Gq proteins to agonist DCG-IV. activate phospholipase C, which catalyzes the produc- tion of diacylglycerol and inositol (1,4,5)-triphosphate Amino Acid in Neurogenesis 141

4.2. Ionotropic NMDAR signals proportion to increasing culture periods from 2 to 4.2.1. Responses to NMDA: Brief exposure to 12 days in the absence of NMDAR ligands, while NMDA led to a marked increase in the fluorescence sustained exposure to NMDA significantly decreased intensity for Ca2+ imaging analysis in neurospheres the activity of BrdU incorporation in neurospheres. The cultured for 2 to 6 days in a manner sensitive to the aforementioned decreases by NMDA were not seen in NMDAR antagonist dizocilpine (MK-801), but not in neurospheres cultured in the presence of both NMDA those cultured for a period longer than 8 days. The and MK-801. However, no significant difference was calcium ionophore A23187 was invariably active in seen in the release of LDH from neocortical neuro- drastically increasing Ca2+ fluorescence in neurospheres spheres into the culture medium irrespective of the cultured for 2 to 12 days, but marked nestin expression sustained exposure to any NMDAR ligand. Therefore, was seen in neurospheres throughout the culture period. tonic activation of NMDAR could lead to a significant In cells cultured for 18 days with spontaneous differenti- decrease in proliferation activity in neural progenitor ation after the removal of bFGF, NMDA markedly cells before commitment and/or differentiation without increased the fluorescence in an MK-801–sensitive affecting cellular viability. manner. The addition of A23187 profoundly increased 4.2.3. Differentiation mediated by NMDAR: In the fluorescence in these differentiated cells, where cells previously exposed to NMDA for 2 to 12 days, a immunoreactive nestin was not expressed at all. significant increase was seen in MAP2 expression with In order to evaluate the transient expression of func- a concomitant significant decrease in GFAP expression tional NMDAR, neurospheres were similarly exposed irrespective of the presence of any differentiation to 50 mM potassium chloride for determination of inducers. Prior exposure to NMDA led to a significant intracellular free Ca2+ ions every 2 days. In contrast to increase in the number of cells with high fluorescence the response to NMDA, exposure to potassium chloride in response to the second brief exposure to NMDA in an at a high concentration invariably increased the fluores- MK-801–sensitive manner on spontaneous differentia- cence intensity in neurospheres irrespective of the tion compared to neurospheres not exposed to NMDA. culture duration up to 12 days. Quantification of these Therefore, activation of NMDAR would invariably lead data clearly confirmed the constant responsiveness to to suppression of the proliferative activity toward self- depolarization by potassium chloride with respect to the renewal and subsequent promotion of the neuronal increased intracellular free Ca2+ level in both neuro- differentiation with a concomitant diminution of the spheres cultured in the presence of bFGF and cells astroglial differentiation in neural progenitors isolated spontaneously differentiated after the removal of bFGF. from fetal rat neocortex. Accordingly, functional NMDAR seems to be tran- siently expressed during a culture period of 2 to 6 days, 4.3. A role of NMDAR in neurogenesis but not for 8 to 12 days, in neurospheres cultured in the As mentioned above, sustained exposure to NMDA presence of bFGF under floating conditions. markedly inhibited the formation of neurospheres 4.2.2. Proliferation mediated by NMDAR: Expo- and subsequent facilitation of differentiation to cells sure to NMDA alone markedly decreased the size of immunoreactive for different neuronal marker proteins neurospheres throughout the culture period up to 12 days in cultured neural progenitor cells (Table 3). The data in a manner sensitive to the prevention by MK-801. obtained in both immunocytochemistry and RT-PCR In proportion to increasing culture durations, MTT analyses gave rise to the idea that functional heteromeric reduction was significantly decreased in neurospheres NMDAR channels are expressed by neural progenitor cultured in the presence of NMDA alone as seen in the cells before commitment and/or differentiation to size of the clusters on phase contrast micrographs. The neurons. The fact that brief exposure to NMDA induces incorporation of BrdU was drastically increased in marked expression of c-Fos, Fos-B, Fra-2, and c-Jun

Table 3. NMDAR signaling to neurogenesis Differentiation Spontaneous Induced Proliferation ATRA CNTF NEUASTNEUASTNEUAST suppressed promoted suppressed promoted suppressed promoted suppressed

Abbreviations: AST, astrocyte; NEU, neuron. 142 N Nakamichi et al proteins in a manner sensitive to MK-801 (30) is also expression level up to 12 days. Expression levels were favorable for the high functionality of NMDAR subunits not significantly different with mRNA for all group III expressed by neural progenitor cells before differentia- mGluR subtype isoforms examined in cells further tion. cultured in the absence of bFGF for an additional 6 days Rapid responsiveness of AP1 member proteins (30, from those seen in neurospheres cultured for 12 days. 58) supports the speculation that brief exposure to The group III mGluR subtype agonist L-AP4 was NMDA could induce transient expression of the AP1 effective in markedly decreasing the size of neuro- complex via functional NMDAR channels assembled by spheres, while sustained exposure to either the group I heteromeric subunits expressed by neural progenitor mGluR subtype agonist DHPG or the group II mGluR cells, toward the modulation of de novo synthesis of subtype agonist DCG-IV did not markedly affect the inducible target proteins responsible for the regulation size of neurospheres. The formation of neurospheres was of cellular proliferation, commitment, differentiation, almost completely abolished after sustained exposure to and/or maturation. Activation of NMDAR would play a L-AP4, but not to either DHPG or DCG-IV at the same key role in the mechanisms associated with commitment concentration. At 10 days after plating, MTT reduction and differentiation of neural progenitor cells at an was significantly decreased in neurospheres cultured in early developmental stage in neurogenesis. The exact the presence of L-AP4, but not in those cultured with mechanism as well as physiological significance of early DHPG or DCG-IV, as assessed by measurement of the and transient expression of functional heteromeric total size (59). NMDAR channels in neural progenitor cells before L-PA4 at >1 μM was inhibitory when neurospheres differentiation, however, remains to be elucidated in were exposed to this compound for 10 consecutive days. future studies. The observations in the present study Sustained exposure to L-AP4 significantly decreased also directly demonstrate facilitation of commitment MTT reduction and the size of neurospheres in a and differentiation to neurons of neural progenitor cells concentration-dependent manner. No neurospheres were through prior activation of NMDAR. The data cited formed in the presence of L-AP4 at 50 μM, while no above are all suggestive of the proposal that NMDA significant difference was seen in the cumulative release signal input may suppress proliferation toward self- of LDH from cells cultured for 10 days into culture renewal along with the promotion of subsequent medium irrespective of the sustained exposure to L-AP4 commitment and differentiation to neurons in neural at concentrations effective in significantly inhibiting progenitor cells. Thus modulation of NMDAR-channel MTT reduction. A negligibly small number of cells functionality would be highly beneficial for the were reactive with the membrane-impermeable dye PI regeneration of central neurons without surgical implan- for DNA staining in neurospheres exposed to L-AP4 tations in patients with a variety of neurodegenerative for 10 days, but most cells were stained with the disorders in a particular situation. membrane-permeable dye Hoeschst33342. A significant decrease was seen in the size of neurospheres exposed to 4.4. Metabotropic GluR signals L-AP4 for a period of 4 to 10 days. Therefore, sustained 4.4.1. Proliferation mediated by group III mGluR: exposure to L-AP4 would lead to a significant decrease In the mouse neocortical lower cell layer before culture, in the proliferation activity toward self-replication in mRNA expression was seen for mGluR1 and mGluR5 neural progenitor cells before commitment and/or isoforms of the group I subtype, mGluR2 and mGluR3 differentiation without affecting cellular viability. isoforms of the group II subtype, and mGluR4 and Sustained exposure to L-AP4 led to a marked mGluR8 isoforms of the group III subtype, but not for decrease in the size of neurospheres, which was blocked mGluR6 and mGluR7 isoforms of the group III subtype, by the group III mGluR antagonist CPPG. Although on RT-PCR analysis. In neocortical neurospheres L-AP4 significantly inhibited MTT reduction in a CPPG- cultured for 10 days, however, mRNA expression was sensitive manner, CPPG alone significantly increased seen for mGluR5, mGluR2, mGluR3, and mGluR8 MTT reducing activity and size of the neurospheres. As isoforms, but not for mGluR1, mGluR4, mGluR6, and the group III mGluR subtype is responsible for the mGluR7 isoforms. The expression of mGluR4 mRNA negative regulation of adenylyl cyclase through inhibi- was almost constant in neurospheres cultured for up to tory Gi/o proteins, we attempted to determine whether 12 days, while a drastic increase was seen in mRNA the adenylyl cyclase activator forskolin affects MTT expression of both mGluR6 and mGluR7 isoforms in reduction and size of the neurospheres. Sustained proportion to increasing culture periods from 2 to exposure to forskolin significantly increased MTT 12 days. The expression of mGluR8 mRNA was doubled reduction and neurosphere size, which were both pre- in neurospheres cultured for 2 days with a constant vented by L-AP4. Accordingly, the inhibition by L-AP4 Amino Acid in Neurogenesis 143 is indeed mediated by the group III mGluR subtype group III subtype. In P19 cells cultured for 4 days negatively linked to adenylyl cyclase in undifferentiated with ATRA, however, mRNA was expressed for the neural progenitor cells. mGluR5, mGluR2, mGluR4, mGluR7, and mGluR8 4.4.2. cAMP/PKA pathway: In addition to forskolin, isoforms, but not for the mGluR1, mGluR3, or mGluR6 both dibutyryl cAMP and PACAP were effective in isoform. Therefore the mRNA expression pattern for significantly increasing MTT reduction and the size of the group III mGluR subtype in undifferentiated P19 the neurospheres, with adrenaline being ineffective. The cells was similar to that in the neural progenitor cells. PKA inhibitor H89 significantly prevented the increases Accordingly, P19 cells were exposed to the group III by forskolin in MTT reduction activity and the size in mGluR subtype agonist L-AP4 in the presence of ATRA neurospheres. H89 alone was also efficient in signifi- for 4 days. Sustained exposure to L-AP4 invariably led cantly inhibiting MTT reduction activity and neuro- to a decreased size of neurospheres in a CPPG-sensitive sphere size. L-AP4 markedly decreased the number of manner, while a marked increase was seen in the size cells immunoreactive to BrdU in a manner sensitive to of neurospheres exposed to forskolin. The increase by CPPG in the neurospheres. In contrast, forskolin signifi- forskolin was significantly decreased by L-AP4 and cantly increased the number of cells immunoreactive H89. H89 alone significantly decreased the size of for BrdU in the neurospheres, which was significantly neurospheres formed. These data clearly indicate the blocked by L-AP4. The increase by forskolin was high similarity between neural progenitors and P19 significantly inhibited by H89, moreover, while H89 cells in terms of the responsiveness to intracellular alone significantly decreased the number of cells cAMP signals. immunoreactive for BrdU. Thus, the cAMP/PKA Therefore, cells were transfected with a reporter pathway would be indeed involved in the mechanisms plasmid containing the full-length promoter region of underlying regulation of proliferation activity toward cyclinD1 (60) or four tandem copies of CRE along with self-replication in undifferentiated neural progenitor the TK-Renilla luciferase construct, with the internal cells. control vector phRL-SV40, by the calcium-phosphate 4.4.3. CyclinD1 gene expression: mRNA expression method. Approximately 30% of cells expressed green was altogether assessed with different adhesion mole- fluorescent protein in P19 cells transfected with the cules, in addition to cyclinD1, a key regulator of the EGFP-C2 plasmid under the transfection condition cell cycle. Marked expression was seen for cyclinD1 employed. The mouse embryonic carcinoma stem cell mRNA in neurospheres cultured for 10 days, while line P19 cells transfected with a reporter plasmid of the cyclinD1 mRNA expression was significantly inhibited full-length promoter region of cylinD1, or four tandems by L-AP4 but promoted by forskolin. The decrease by of CRE, were thus exposed to a test drug in the presence L-AP4 was significantly prevented by CPPG. Similarly, of ATRA. Exposure to L-AP4 significantly decreased there were marked expressions of the mRNAs for the cyclinD1 promoter activity in a CPPG-sensitive NCAM, N-cadherin, integrin-β1, and laminin-β1 in manner. Forskolin alone significantly increased the neurospheres cultured for 10 days. However, no signifi- promoter activity, while the increase by forskolin was cant alterations were found in mRNA expression of significantly decreased by L-AP4 and H89. Moreover, these adhesion molecules in neurospheres cultured for H89 alone was effective in significantly decreasing 10 days irrespective of the presence of the tested drugs. the promoter activity of cyclinD1. Similarly significant 4.4.4. CyclinD1 promoter activity: In silico analysis alterations were invariably seen with the promoter clearly showed the presence of particular nucleotide activity in P19 cells transfected with a reporter plasmid sequences recognized by the nuclear transcription factors containing the CRE tandem construct. Therefore, activa- AP1 and CREB upstream of the cyclinD1 promoter tion of the group III mGluR subtype would lead to region. Therefore, we first examined the similarity of suppression of the proliferation activity toward self- the murine embryonic carcinoma stem cell line P19 cells replication through downregulation of the cell cycle to murine neural progenitor cells (59). P19 cells cultured regulator cyclinD1 at the level of gene transcription in with ATRA within 4 days showed formation of neuro- neural progenitor cells. spheres immunoreactive for nestin, but neither MAP2 4.4.5. Differentiation mediated by group III mGluR: nor GFAP. RT-PCR analysis indicated that P19 cells Neural progenitor cells were cultured with EGF in either before culture with ATRA expressed mRNA for the the presence or absence of a test drug, followed by mGluR4 and mGluR8 isoforms of the group III subtype, further culture in the absence of those test drugs for an but not for mGluR1 and mGluR5 isoforms of the group I additional 4 days toward spontaneous differentiation subtype, mGluR2 and mGluR3 isoforms of the group II and subsequent double immunocytochemical detection subtype, or mGluR6 and mGluR7 isoforms of the of both MAP2 and GFAP. Prior exposure to L-AP4 144 N Nakamichi et al induced a marked increase in the number of cells culture with EGF, marked expression was similarly seen immunoreactive for GFAP with a concomitant decrease with mRNA for Mash1, Neuerogenin2, and Hes1, but in that for MAP2, whereas forskolin increased the not for Math3, NeuroD, or Hes5. In neural progenitors number of cells immunoreactive for MAP2 together cultured with EGF for 10 days, Hes5 mRNA was highly with a decrease in that for GFAP. Around 60% of expressed in addition to Mash1 and Hes1 mRNA. RT- cells were immunoreactive for GFAP and less than PCR analysis clearly revealed that neither L-AP4 nor 40% for MAP2 on spontaneous differentiation. Prior forskolin significantly affected mRNA expression of exposure to L-AP4 alone significantly increased the the activator-type Mash1 or the repressor-type Hes1 in number of cells immunoreactive for GFAP with a neurospheres cultured for 10 days. The present findings significant decrease in that for MAP2 in a CPPG- do not support the possible involvement of these sensitive fashion. In contrast, forskolin significantly regulatory bHLH genes in the mechanisms relevant to increased the number of cells immunoreactive for MAP2 the promotion of subsequent astroglial differentiation in along with a significant decrease in that for GFAP. neural progenitor cells exposed to the group III mGluR– As seen with spontaneous differentiation, prior subtype agonist L-AP4. sustained exposure to L-AP4 alone significantly increased the number of cells immunoreactive for GFAP 4.5. A role of group III mGluR in neurogenesis with a concomitant significant decrease in that for Tonic activation of the group III mGluR subtype MAP2 in a CPPG-sensitive manner irrespective of the leads to a significant decrease in the capability to form differentiation inducers added. Forskolin alone signifi- neurospheres in cultured neural progenitor cells cantly increased the number of cells immunoreactive (Table 4). The data cited above all support the idea for MAP2 along with a significant decrease in that for that activation of the group III mGluR subtype leads to GFAP in cells differentiated by either ATRA or CNTF. suppression of proliferation activity in neural pro- Accordingly, the group III mGluR subtype could be genitors through inhibition of cAMP/PKA signaling functionally expressed to play a pivotal role not only in processes as shown in the literature (53 – 56). As the proliferation toward self-replication through a mecha- group III mGluR subtype is negatively linked to adenylyl nism relevant to a cAMP/PKA signaling to cyclinD1 cyclase through the inhibitory Gi/o protein, however, gene expression, but also in subsequent differentiation the formation of cAMP should be under the basic and/or toward particular cell lineages, in undifferentiated tonic stimulation by some endogenous signals for a neural progenitor cells. group III mGluR agonist to elicit its pharmacological 4.4.6. Expression of regulatory genes: Next, we actions in undifferentiated neural progenitor cells. attempted to determine whether sustained exposure to From this point of view, it should be noted that PACAP L-AP4 modulates the expression profile of a variety significantly increases both MTT reduction and neuro- of bHLH genes responsible for the critical control of sphere size. This polypeptide was shown to regulate the cellular differentiation in neural stem cells. There is proliferation and differentiation in neural progenitor accumulating evidence that timing and regulation of cells during embryonic development (63). In our hands, cell differentiation are critically controlled by diverse mRNA expression was not detected for any isoforms of bHLH genes in neural stem cells. These include the adrenergic β-receptors known to be positively coupled activator type (Mash1, Math, NeuroD, and Neurogenin) to adenylyl cyclase in undifferentiated neural progenitor and the repressor type (Hes1 and Hes5) of bHLH genes cells. Accordingly, the group III mGluR subtype could (61, 62). In embryonic mouse whole brain, RT-PCR play a pivotal role in mechanisms underlying the analysis indicated marked mRNA expression was seen regulation of cellular proliferation toward self- for Mash1, Neurogenin2, Math3, NeuroD, and Hes1, replication, in association with particular endogenous but not for Hes5. In neural progenitor cells before signals positively coupled to cAMP formation, in

Table 4. Group III mGluR signaling to neurogenesis Differentiation Spontaneous Induced Proliferation ATRA CNTF NEUASTNEUASTNEUAST suppressed suppressed promoted suppressed promoted suppressed promoted

Abbreviations: AST, astrocyte; NEU, neuron. Amino Acid in Neurogenesis 145 undifferentiated neural progenitor cells. 4.6. Adult neurogenesis To our knowledge, this is the first direct demonstra- 4.6.1. Involvement of NMDAR in adult neuro- tion of a significant decrease in mRNA expression of genesis: Adult neuronal proliferation is susceptible to cyclinD1 after sustained activation of the group III the inhibition by excitatory inputs from the entorhinal mGluR subtype in neural progenitors. CyclinD1 is a cortex via the perforant path in a manner prevented by major cell-cycle regulator responsible for the promoted NMDAR antagonists (70). Systemic administration of transition to a proliferative stage, while overexpression NMDA not only increases DNA binding of the of cyclinD1 leads to shortening of the G1 phase and transcription factor AP1 (58, 71), but ,in fact, also subsequent rapid entry into the S phase (64). CyclinD1 decreases cellular proliferation activity determined by gene expression is mainly regulated at the transcriptional the incorporation of BrdU in a manner sensitive to the level with less involvement of the posttranscriptional NMDAR antagonist MK-801 (29) in the DG of the processes (65). In fact, CRE is shown to be required adult murine hippocampus. Systemic administration of for the basal and inducible expression of cyclinD1 gene NMDA markedly reduces expression of the neural in association with the cAMP/PKA signaling pathway progenitor marker protein nestin as well as PCNA in the in the pancreatic β-cell line INS-1 cells (66). Our DG, without significantly affecting that in the SVZ (29). findings from promoter analysis thus argue in favor of Moreover, sustained exposure to an NMDAR agonist the proposal that activation of the group III mGluR not only decreases the size of neurospheres formed by subtype suppresses self-replication through decreased clustered cells, but also facilitates neuronal commitment cyclinD1 gene expression due to reduction of the induced by ATRA, in cultured neural progenitor cells promoter activity in response to the attenuation of the isolated from the adult mouse hippocampus (30). In our cAMP/PKA signaling pathway after decreased intra- hands, astroglial differentiation predominates in neural cellular cAMP levels in undifferentiated neural pro- progenitor cells isolated from adult mouse hippocampus genitors. irrespective of the presence of differentiation inducers Nevertheless, the reason why group II and group III in vitro. mGluR subtypes differentially affect self-replication 4.6.2. Involvement of NMDAR-mediated neuro- activity in undifferentiated neural progenitor cells is not genesis in a psychiatric disease: PTSD is a long-lasting clear so far. The view that both mGluR subtypes are psychiatric disease after the traumatic experience of negatively coupled to adenylyl cyclase through the severe fatal stress with the consequence of hippocampal inhibitory Gi/o protein in almost all cells is prevailing, atrophy. Accumulating evidence that glutamatergic whereas several independent lines of evidence indicate signals are involved in the mechanisms relevant to the differential properties between group II and group III crisis of a variety of symptoms caused by stress is mGluR subtypes (67, 68). Similar differentiation available in the literature. WIRS leads to significant between group II and group III mGluR subtypes is also alterations of endogenous levels of both glutamate and seen with the inducible expression of the AP1 complex GABA in particular discrete brain structures in rats (72), in cultured rat neocortical neurons (69). The present for example, while stress is believed to promote the data are suggestive of the proposal that signals mediated excessive release of glutamate at glutamatergic synapses by group III mGluR subtype may promote subsequent in patients with depression and PTSD (73). A single differentiation of neural progenitor cells into an immobilization stress induces changes in expression astroglial lineage. The group III mGluR subtype is not levels of mRNA for particular subunits of ionotropic required for the expression of particular adhesion glutamate–receptor subtypes, such as NMDAR and molecules responsible for cellular proliferation, commit- AMPAR, in selected regions of the rat hippocampus ment, and/or differentiation toward a particular lineage and hypothalamus (74). Restraint and tailshock stress in undifferentiated neural progenitor cells. Functional impairs long-term potentiation but enhances long-term group III mGluR subtype could be constitutively ex- depression through activation of NMDAR in the hippo- pressed in undifferentiated neural progenitor cells before campal CA1 pyramidal subfield (75). Moreover, an commitment for the regulation of cellular proliferation antagonist for NMDAR, but not for AMPAR, prevents toward self-replication and subsequent differentiation the dendritic atrophy induced by daily restraint stress in into particular lineages in fetal mouse brains. Sustained hippocampal CA3 pyramidal neurons, which is a highly activation of the group III mGluR subtype would result replicable consequence of chronic stress (76). in the suppression of proliferation activity toward self- In order to test the possible involvement of NMDAR renewal together with the facilitation of subsequent responsible for long-term memory as well as prolifera- differentiation into astroglia. tion of neural progenitor cells, mice were intra- peritoneally administered the NMDAR antagonist MK- 146 N Nakamichi et al

801, followed by measurements of different behavioral together, NMDAR would mediate the prolonged and responses induced by WIRS 14 days after exposure (77). repeated suppression of progenitor cell proliferation in In a forced swimming test, MK-801 significantly the DG toward long-lasting and bidirectional abnormal improved the increased immobility in mice exposed to behaviors in patients with PTSD in a particular situation. WIRS when injected once 30 min before exposure, without altering the immobility time in mice not exposed 5. Amino acid signaling to neurogenesis to WIRS. In addition, no significant increase was seen in the spontaneous locomotor activity in mice exposed Our studies on neurogenesis have begun with the to WIRS with prior administration of MK-801 when findings that systemic administration of NMDA pre- determined 14 days after exposure. Moreover, mice dominantly increases AP1 DNA binding in a manner were injected with MK-801, followed by exposure to sensitive to MK-801 in the DG of the adult murine WIRS for 3 h, and subsequent flashback stress by hippocampus (58, 71). On the basis of these previous forced swimming 9 days after the initial conditioning in vivo studies, we have analyzed the mechanisms by WIRS. Prior single administration of MK-801 was underlying amino acid signaling to neurogenesis in the also effective in significantly preventing the decrease in developing brain in vitro. It is likely that self-renewal the number of BrdU-positive clustered cells in mice activity would be under control by the delicate balancing exposed to WIRS when determined 5 days after the between GABAergic and glutamatergic signals through flashback stress by forced swimming. Similarly, mice the respective receptors expressed by neural progenitor were administered with MK-801, followed by WIRS for cells before differentiation as seen in matured neurons. 3 h, and subsequent measurement of BrdU incorporation In undifferentiated progenitors expressed in the develop- 5 days later. The administration of MK-801 alone did ing and adult brains, GABA could positively regulate not significantly affect the number of BrdU-positive proliferation for self-replication through activation of clustered cells in the DG 5 days after administration, either the GABAAR or GABABR subtype, but glutamate while prior administration of MK-801 significantly would negatively modulate proliferation through activa- reversed the significant decrease in the number of BrdU- tion of either NMDAR or group III mGluR subtype positive clustered cells to the control level in mice (Fig. 1). Activation of the GABAAR subtype could exposed to WIRS. Thus, NMDAR could play a role in lead to the increased self-replication activity in neural the mechanisms relevant to the input and/or consolida- progenitor cells along with the promotion of subsequent tion of fatal traumatic stress signals in the DG. differentiation into astroglia (23, 24), while activation of To our knowledge, this is the first direct demonstra- the GABABR subtype would result in the similarly tion of the prevention by prior administration of MK- efficient promotion of self-renewal activity together 801 of the suppressed incorporation of BrdU in the DG with the facilitation of subsequent differentiation into along with the improvement of behavioral abnormalities neurons (31). On the other hand, activation of the in mice exposed to WIRS. In our previous studies, an NMDAR subtype could lead to the suppressed self- intraperitoneal injection of NMDA led to marked renewal capacity along with the promotion of sub- decreases in the incorporation of BrdU and the expres- sequent differentiation into neurons (57), while activa- sion of both nestin and PCNA (24), in addition to the tion of the group III mGluR subtype would result in the predominant induction of AP1 (23), in the DG of adult similarly efficient suppression of self-renewal activity mice. These alterations are all antagonized by the prior together with the facilitation of subsequent differentia- administration of MK-801 (29, 58). Thus, one possible tion into astroglia (59). Therefore, modulation of the but hitherto unidentified speculation is that WIRS would functionality of particular GABA and/or glutamate facilitate the release of glutamate to activate NMDAR receptor subtypes would be of a great benefit for the toward subsequent suppression of the proliferation of regeneration and supplementation of neuronal and/or neural progenitor cells in the DG. Stress is shown to astroglial lineages without surgical implantations of promote glutamate release in patients with depression neural progenitor cells in patients suffering from a and PTSD (73), indeed, while systemic administration of variety of brain diseases relevant to neuronal and/or NMDA induces the suppression of BrdU incorporation, astroglial dysfunctions. Alternatively, prior modulation nestin expression, and PCNA expression in the DG of the activity of GABAergic and glutamatergic recep- 2 days later (29). Similarly, flashback by forced swim- tors could be a novel strategy for subsequent regulation ming significantly decreases the number of BrdU- of the differentiation toward a particular progeny cell positive clustered cells in the DG (77). The memory lineage of implanted neural progenitor cells. acquired in contextual fear-conditioning is shown to be associated with the hippocampal functions (78). Taken Amino Acid in Neurogenesis 147

Fig. 1. Hypothetical GABAergic and glutamatergic signaling to neurogenesis. Although GABAergic signal inputs would similarly lead to the stimulation of proliferation for self-replication through activation of either the ionotropic GABAAR or the metabotropic GABABR subtype expressed by undifferentiated neural progenitor cells, GABAAR signals could promote subsequent differentiation to an astroglial lineage and GABABR signals could induce facilitation of subsequent differentiation to neurons, respectively. Glutamatergic signal inputs could invariably result in the suppression of proliferation for self-renewal of undifferentiated neural progenitor cells, by contrast, while prior activation of the ionotropic NMDAR subtype would lead to the promotion of subsequent differentiation into neurons. Activation of the metabotropic group III mGluR subtype would promote the subsequent differentiation to an astroglial lineage. Accordingly, neurogenesis may be under control by the delicate balancing between GABAergic and glutamatergic signal inputs in neural progenitor cells as seen in matured neurons. In the absence of GABAergic and glutamatergic signals, neural progenitor cells are undoubtedly endowed to proliferate for self-renewal and differentiate into progeny lineages.

Acknowledgments 1441. 7 Gage FH, Kempermann G, Palmer TD, Peterson DA, Ray J. This work was supported in part by Grants-in-Aid for Scientific Multipotent progenitor cells in the adult dentate gyrus. J Research to N.N. (20790061), T.T. (20790250), and Y.Y. (18053009) Neurobiol. 1998;36:249–266. from the Ministry of Education, Culture, Sports, Science, and Tech- 8 Garcia-Verdugo JM, Doetsch F, Wichterle H, Lim DA, Alvarez- nology, Japan. Buylla A. Architecture and cell types of the adult subventricular zone: in search of the stem cells. J Neurobiol. 1998;36:234–248. References 9 Cameron HA, Woolley CS, McEwen B, Gould E. Differentia- tion of newly born neurons and glia in the dentate gyrus of the 1 Doetsch F, Caille I, Lim DA, Garcia-Verdugo JM, Alvarez- adult rat. Neuroscience. 1993;56:337–344. Buylla A. Subventricular zone astrocytes are neural stem cell in 10 Gage FH, Coates PW, Palmer TD, Kuhn HG, Fisher LJ, Suhnen the adult mammalian brain. Cell. 1999;97:703–716. JO, et al. Survival and differentiation of adult neuronal 2 Johansson CB, Momma S, Clarke DL, Risling M, Lendahl U, progenitor cells transplanted to the adult brain. Proc Natl Acad Frisen J. Identification of a neural stem cell in the adult Sci U S A. 1995;92:11879–11883. mammalian central nervous system. Cell. 1999;96:25–34. 11 Temple S, Alvarez-Buylla A. Stem cells in the adult mammalian 3 Altman J, Das G. Postnatal neurogenesis in the guinea-pig. central nervous system. Curr Opin Neurobiol. 1999;9:135–141. Nature. 1967;214:1098–1101. 12 Suhonen JO, Peterson DA, Ray J, Gage FH. Differentiation of 4 Kaplan MS, Hinds JW. Neurogenesis in adult rat: electron adult hippocampus-derived progenitors into olfactory neurons microscopic analysis of light radioautographs. Science. 1977; in vivo. Nature. 1996;383:624–627. 197:1092–1094. 13 Lim DA, Tramontin AD, Trevejo JM, Herrera DG, Garcia- 5 Kaplan MS, Bell DH. Neuronal proliferation in the 9-month-old Verdugo JM, Alvarez-Buylla A. Noggin antagonizes BMP rodent: radioautographic study of granule cells in the hippo- signaling to create a niche for adult neurogenesis. Neuron. campus. Exp Brain Res. 1983;52:1–5. 2000;28:713–726. 6 Kaplan MS, Bell DH. Mitotic neuroblasts in the 9-day-old and 14 Shen Q, Goderie SK, Jin L, Karanth N, Sun Y, Abramova N, 11-month-old rodent hippocampus. J Neurosci. 1984;4:1429– et al. Endothelial cells stimulate self-renewal and expand 148 N Nakamichi et al

neurogenesis of neural stem cells. Science. 2004;304:1338– 31 Fukui M, Nakamichi N, Yoneyama M, Ozawa S, Fujimori S, 1340. Takahata Y, et al. Modulation of cellular proliferation and 15 Li L, Xie T. Stem cell niche: structure and function. Ann Rev differentiation through GABAB receptors expressed by undiffer- Cell Dev Biol. 2005;21:605–631. entiated neural progenitor cells isolated from fetal mouse brain. 16 Nicoleau C, Benzakour O, Agasse F, Thiriet N, Petit J, Prestoz J Cell Physiol. 2008;216:507–519. L, et al. Endogenous hepatocyte growth factor is a niche signal 32 Haydar TF, Wang F, Schwartz ML, Rakic P. Differential for subventricular zone neural stem cell amplification and self- modulation of proliferation in the neocortical ventricular and renewal. Stem Cells. 2009;27:408–419. subventricular zones. J Neurosci. 2000;20:5764–5774. 17 Cavanagh JF, Mione MC, Pappas IS, Parnavelas JG. Basic 33 Aberg MA, Ryttsen F, Hellgren G, Lindell K, Rosengren LE, fibroblast growth factor prolongs the proliferation of rat cortical MacLennan AJ, et al. Selective introduction of antisense progenitor cells in vitro without altering their cell cycle oligonucleotides into single adult CNS progenitor cells using parameters. Cereb Cortex. 1997;7:293–302. electroporation demonstrates the requirement of STAT3 activa- 18 Vaccarino FM, Schwartz MI, Raballo R, Nilsen J, Rhee J, Zhou tion for CNTF-induced gliogenesis. Mol Cell Neurosci. 2001; M, et al. Changes in cerebral cortex size are governed by 17:426–443. fibroblast growth factor during embryogenesis. Nature Neurosci. 34 Wohl CA, Weiss S. Retinoic acid enhances neuronal prolifera- 1999;3:246–253. tion and astroglial differentiation in culture of CNS stem cell- 19 LoTurco JJ, Owens DF, Heath MJS, Davis MBE, Kriegstein AR. derived precursors. J Neurobiol. 1998;37:281–290. GABA and glutamate depolarize cortical progenitor cells and 35 Qian X, Davis AA, Goderie SK, Temple S. FGF2 concentration inhibit DNA synthesis. Neuron. 1995;15:1287–1298. regulates the generation of neurons and glia from multipotent 20 Antonopoulos J, Pappas IS, Parnavelas JG. Activation of the cortical stem cells. Neuron. 1998;18:81–93. GABAA receptor inhibits the proliferative effects of bFGF in 36 Retaux S, Rogard M, Bach I, Failli V, Besson MJ. Lhx9: a novel cortical progenitor cells. Eur J Neurosci. 1997;9:291–298. LIM-homeodomain gene expressed in the developing forebrain. 21 Fiszman ML, Borodinsky LN, Neale JH. GABA induces pro- J Neurosci. 1999;19:783–793. liferation of immature cerebellar granule cells grown in vitro. 37 Xiao Q, Xu HY, Wang SR, Lazar G. Developmental changes of Brain Res Dev Brain Res. 1999;115:1–8. NADPH-diaphorase positive structures in the isthmic nuclei of 22 Tozuka Y, Fukuda S, Namba T, Seki T, Hisatsune T. GABAergic the chick. Anat Embryol (Berl). 2000;201:509–519. excitation promotes neuronal differentiation in adult hippo- 38 Bachler M, Neubuser A. Expression of members of the Fgf campal progenitor cells. Neuron. 2005;47:803–815. family and their receptors during midkacial development. Mech 23 Yoneyama M, Fukui M, Nakamichi N, Kitayama T, Taniura H, Dev. 2001;100:313–316. Yoneda Y. Activation of GABAA receptors facilitates astroglial 39 Lee MY, Deller T, Kirsch M, Frotscher M, Hofmann HD. differentiation induced by ciliary neurotrophic factor in neural Differential regulation of ciliary neurotrophic factor (CNTF) progenitors isolated from fetal rat brain. J Neurochem. and CNTF receptor alpha expression in astrocytes and neurons 2007;100:1667–1679. of the fascia dentate after entorhinal cortex lesion. J Neurosci. 24 Fukui M, Nakamichi N, Yoneyama M, Yoshida K, Ozawa S, 1997;17:1137–1146. Kitayama T, et al. Upregulation of ciliary neurotrophic factor 40 Jacques TS, Relvas JB, Nishimura S, Pytela R, Edwards GM, receptor expression by GABAA receptors in undifferentiated Streuli CH, et al. Neural precursor cell chain migration and neural progenitors of fetal mouse brain. J Neurosci Res. division are regulated through different β1 integrins. Develop- 2008;86:2615–2623. ment. 1998;125:3167–3177. 25 Meier E, Hertz L, Schousboe A. Neurotransmitters as develop- 41 Prestoz L, Relvas JB, Hopkins K, Patel S, Sowinski P, Price J, mental signals. Neurochem Int. 1991;18:1–15. et al. Association between integrin-dependent migration 26 Pertoft H, Laurent TC. Isopycnic separation of cells by centrifu- capacity of neural stem cells in vitro and anatomical repair gation in percoll gradients. Prog Clin Biol Res. 1982;102 following transplantation. Mol Cell Neurosci. 2001;18:473–484. (pt. A):95–104. 42 Amoureux MC, Cunningham BA, Edelman GM, Crossin KL. 27 Maric O, Maric I, Ma W, Lahojuji F, Somogyi R, Wen X, et al. N-CAM binding inhibits the proliferation of hippocampal Anatomical gradients in proliferation and differentiation of progenitor cells and promotes their differentiation to a neuronal embryonic rat CNS accessed by buoyant density fractionation: phenotype. J Neurosci. 2000;20:3631–3640. α3, β3 and γ2 GABAA receptor subunit co-expression by post- 43 Coffman CR, Skoglund P, Harris WA, Kintner CR. Expression mitotic neocortical neurons correlates directory with cell of an extracellular deletion of Notch diverts cell fate in Xenopus buoyancy. Eur J Neurosci. 1997;9:507–522. embryos. Cell. 1993;73:659–671. 28 Palmer TD, Markakis EA, Willhoite AR, Safar F, Gage FH. 44 Yamamoto S, Nagao M, Sugimori M, Kosako H, Nakatomi H, Fibroblast growth factor-2 activates a latent neurogenic program Yamamoto N, et al. Transcription factor expression and Notch- in neural stem cells from diverse regions of the adult CNS. J dependent regulation of neural progenitors in the adult rat spinal Neurosci. 1999;19:8487–8497. cord. J Neurosci. 2001;21:9814–9823. 29 Kitayama T, Yoneyama M, Yoneda Y. Possible regulation by N- 45 Feng ZF, Li L, Ng PY, Porter AG. Neuronal differentiation and methyl-D-aspartate receptors of proliferative progenitor cells protection from nitric oxide-induced apoptosis require c-Jun- expressed in adult mouse hippocampal dentate gyrus. J Neuro- dependent expression of NCAM 140. Mol Cell Biol. 2002;22: chem. 2003;84:767–780. 5357–5366. 30 Kitayama T, Yoneyama M, Tamaki K, Yoneda Y. Regulation of 46 Nishida K, Kitazawa R, Mizuno K, Maeda S, Kitazawa S. neuronal differentiation by N-methyl-D-aspartate receptors Identification of regulatory elements of human alpha 6 integrin expressed in neural progenitor cells isolated from adult mouse subunit gene. Biochem Biophys Res Commun. 1997;241:258– hippocampus. J Neurosci Res. 2004;76:599–612. 263. Amino Acid in Neurogenesis 149

47 Nacher J, Rosell DR, Alonso-Llosa G, McEwen BS. NMDA 517–530. receptor antagonist treatment induces a long-lasting increase in 62 Ross SE, Greenberg ME, Stiles CD. Basic helix–loop–helix the number of proliferating cells, PSA-NCAM-immunoreactive factors in cortical development. Neuron. 2003;39:13–25. granule neurons and radial glia in the adult rat dentate gyrus. Eur 63 Dicicco-Bloom E, Lu N, Pintar JE, Zhang J. The PACAP J Neurosci. 2001;13:512–520. ligand/receptor system regulates cerebral cortical neurogenesis. 48 Ozog MA, Bernier SM, Bates DC, Chatterjee B, Lo CW, Naus Ann N Y Acad Sci. 1998;865:274–289. CC. The complex of ciliary neurotrophic-ciliary neurotrophic 64 Sherr CJ. G1 phase progression: cycling on cue. Cell. 1994; receptor alpha up-regulates connexin 43 and intercellular 79:551–555. coupling in astrocytes via the Janus tyrosine kinase/signal 65 Yan Y-X, Nakagawa H, Lee M-H, Rustgi AK. Transforming transducer and activator of transcription pathway. Mol Biol Cell. growth factor-α enhances cyclin D1 transcription through the 2004;15:4761–4774. binding of early response protein to a cis-regulatory element in 49 Goncalves MB, Boyle J, Webber DJ, Hall S, Minger SL, Corco- the cyclin D1 promoter. J Biol Chem. 1997;272:33181–33190. ran JP. Timing of the retinoid-signalling pathway determines the 66 Kim M-J, Kang J-H, Park YG, Ryu GR, Ko SH, Jeong I-K, et al. expression of neuronal markers in neural progenitor cells. Dev Extendin-4 induction of cyclin D1 expression in INS-1 β-cells: Biol. 2005;278:60–70. involvement of cAMP-responsive element. J Endocrinol. 2006; 50 Conn PJ, Pin JP. Pharmacology and functions of metabotropic 188:623–633. glutamate receptors. Ann Rev Pharmacol Toxicol. 1997;37:205– 67 Neugebauer V, Chen PS, Willis WD. Groups II and III 237. metabotropic glutamate receptors differentially modulate brief 51 Schoepp DD, Jane DE, Monn JA. Pharmacological agents acting and prolonged nociception in primate STT cells. J Neurophysiol. at subtypes of metabotropic glutamate receptors. Neuro- 2000;84:2998–3009. pharmacol. 1999;38:1431–1476. 68 Capogna M. Distinct properties of presynaptic group II and 52 Zhai Y, George CA, Zhai J, Nisenbaum ES, Johnson MP, III metabotropic glutamate receptor-mediated inhibition of Nisenbaum LK. Group II metabotropic glutamate receptor perforant pathway-CA1 EPSCs. Eur J Neurosci. 2004;19:2847– modulation of DOI-induced c-fos mRNA and excitatory 2858. responses in the cerebral cortex. Neuropsychopharmacol. 69 Sugiyama C, Nakamichi N, Ogura M, Honda E, Maeda S, 2003;28:45–52. Taniura H, et al. Activator protein-1 responsive to the group II 53 Cartmell J, Schoepp DD. Regulation of neurotransmitter release metabotropic glutamate receptor subtype in association with by metabotropic glutamate receptors. J Neurochem. 2000;75: intracellular calcium in cultured rat cortical neurons. Neurochem 889–907. Int. 2007;51:467–475. 54 Schoepp DD. Unveiling the functions of presynaptic metabo- 70 Cameron HA, McEwen B, Gould E. Regulation of adult tropic glutamate receptors in the central nervous system. J neurogenesis by excitatory input and NMDA receptor activation Pharmacol Exp Ther. 2001;299:12–20. in the dentate gyrus. J Neurosci. 1995;15:4687–4692. 55 Moldrich RX, Chapman AG, De Sarro G, Meldrum BS. 71 Yoneda Y, Ogita K. Rapid and selective enhancement of DNA Glutamate metabotropic receptors as targets for drug therapy in binding activity of the transcription factor AP1 by systemic epilepsy. Eur J Pharmacol. 2003;476:3–16. administration of N-methyl-D-aspartate in murine hippocampus. 56 Kenny PJ, Markou A. The ups and downs of addiction: role of Neurochem Int. 1994;25:263–271. metabotropic glutamate receptors. Trends Pharmacol Sci. 2004; 72 Yoneda Y, Kanmori K, Ida S, Kuriyama K. Stress-induced 25:265–272. alterations in metabolism of gamma-aminobutyric acid in rat 57 Yoneyama M, Nakamichi N, Fukui M, Kitayama T, Georgiev brain. J Neurochem. 1983;40:350–356. DD, Makanga JO, et al. Promotion of neural differentiation 73 Sapolsky RM. The possibility of neurotoxicity in the hippo- through activation of NMDA receptors transiently expressed by campus in major depression: a primer on neuron death. Biol undifferentiated neural progenitor cells in fetal rat neocortex. Psychiatry. 2000;48:755–765. J Neurosci Res. 2008;86:2392–2402. 74 Bartanusz V, Aubry JM, Pagliusi S, Jezova D, Baffi J, Kiss JZ. 58 Yoneda Y, Ogita K, Azuma Y, Kuramoto N, Manabe T, Stress-induced changes in messenger RNA levels of N-methyl- Kitayama T. Predominant expression of nuclear activator D-aspartate and AMPA receptor subunits in selected regions of protein-1 complex with DNA binding activity following the rat hippocampus and hypothalamus. Neuroscience. 1995;66: systemic administration of N-methyl-D-aspartate in dentate 247–252. granule cells of murine hippocampus. Neuroscience. 1999;93: 75 Kim JJ, Foy MR, Thompson RF. Behavioral stress modifies 19–31. hippocampal plasticity through N-methyl-D-aspartate receptor 59 Nakamichi N, Yoshida K, Ishioka Y, Makanga JO, Fukui M, activation. Proc Natl Acad Sci U S A. 1996;93:4750–4753. Yoneyama M, et al. Group III metabotropic glutamate receptor 76 Magarinos AM, McEwen BS. Stress-induced atrophy of apical activation suppresses self-replication of undifferentiated dendrites of hippocampal CA3c neurons: involvement of gluco- neocortical progenitor cells. J Neurochem. 2008;105:1996– corticoid secretion and excitatory amino acid receptors. 2012. Neuroscience. 1995;69:89–98. 60 D’Amico M, Wu K, Fu M, Rao M, Albanese C, Russell RG, 77 Tamaki K, Yamada K, Nakamichi N, Taniura H, Yoneda Y. et al. The inhibitor of cyclin-dependent kinase 4a/alternative Transient suppression of progenitor cell proliferation through reading frame (INK4a/ARF) locus encoded proteins p16INK4a NMDA receptors in hippocampal dentate gyrus of mice with and p19ARF repress cyclin D1 transcription through distinct traumatic stress experience. J Neurochem. 2008;105:1642–1655. cis elements. Cancer Res. 2004;64:4122–4130. 78 Kim JJ, Fanselow MS. Modality-specific retrograde amnesia of 61 Bertrand N, Castro DS, Guillemot F. Proneural genes and the fear. Science. 1992;256:675–677. specification of neural cell types. Nat Rev Neurosci. 2002;3: