Synaptic integration of newly generated neurons in rat dissociated hippocampal cultures

Molecular and Cellular Neuroscience 47 (2011) 203–214

Contents lists available at ScienceDirect

Molecular and Cellular Neuroscience j o u r n a l h o m e p a g e : w w w. e l s e v i e r. c o m / l o c a t e / y m c n e

Synaptic integration of newly generated neurons in rat dissociated hippocampal cultures Juliette E. Cheyne a, 1, Louise Grant a, Charlotte Butler-Munro a, Janie W. Foote a, Bronwen Connor b, Johanna M. Montgomery a,⁎ a

Department of Physiology, Centre for Brain Research, Faculty of Medical and Health Sciences, University of Auckland, Private Bag 92019, Auckland, New Zealand Department of Pharmacology and Clinical Pharmacology, Centre for Brain Research, Faculty of Medical and Health Sciences, University of Auckland, Private Bag 92019, Auckland, New Zealand

b

a r t i c l e

i n f o

Article history: Received 14 December 2010 Revised 20 April 2011 Accepted 26 April 2011 Available online 4 May 2011 Keywords: Neurogenesis Hippocampal culture Synapse 5′-Bromo-2-deoxyuridine Retrovirus Glutamate receptors GABAA receptors

a b s t r a c t In the dentate gyrus of the hippocampus new neurons are born from precursor cells throughout development and into adulthood. These newborn neurons hold significant potential for self-repair of brain damage caused by neurodegenerative disease. However, the mechanism by which newborn neurons integrate into the brain is not understood due to a lack of knowledge of the molecular and functional characteristics of the synapses formed by newborn neurons. Here we report that dissociated hippocampal cultures continue to produce new granule cells in vitro that fire action potentials and become synaptically integrated into the existing network of mature hippocampal neurons. Quantification of the expression of synaptic proteins at newborn and mature granule cell synapses revealed synapse development onto newborn neurons occurs sequentially with initial synaptic contacts evident from 6 days after cell birth. These data also showed that the dendrites of newborn neurons have a high density of Piccolo and Bassoon puncta on them and therefore have a high potential to be integrated into the neuronal network through new synaptic connections. Electrophysiological recordings from newborn neurons reveal these synapses are functional within 10 days of cell birth. GABAergic input synapses were found to mature faster in newborn neurons than glutamatergic synapses where sequential recruitment of postsynaptic glutamate receptors occurred. Group I metabotropic glutamate receptors (mGluR1/5) were present at higher levels compared with ionotropic glutamate receptors (NMDA and AMPA receptors), suggesting that metabotropic and ionotropic receptors play differential roles at glutamatergic synapses in the integration and the maturation of newborn neurons. These data show that dissociated hippocampal cultures can provide a useful model system in which to study the integration of newborn neurons into existing neuronal circuits to increase our understanding of how the function of newborn neuron synapses could contribute to restoring damaged neuronal networks. © 2011 Elsevier Inc. All rights reserved.

Introduction In the dentate gyrus of the hippocampus neurogenesis continues throughout development and into adulthood. A portion of the newborn granule cells in the adult brain functionally incorporate into the hippocampal circuitry (van Praag et al., 2002) where they

Abbreviations: BrdU, 5′-Bromo-2-deoxyuridine; DAI, days after infection; DIV, days in vitro; LTP, long term potentiation; mEPSCs, miniature excitatory postsynaptic currents; MAP2, microtubule associated protein 2; PTVs, Piccolo Transport Vesicles. ⁎ Corresponding author at: Department of Physiology, The University of Auckland, Private Bag 92019, Auckland, New Zealand. Fax: + 64 9 3737499. E-mail addresses: [email protected] (J.E. Cheyne), [email protected] (L. Grant), [email protected] (C. Butler-Munro), [email protected] (B. Connor), [email protected] (J.M. Montgomery). 1 Present address: Netherlands Institute for Neuroscience, Meibergdreef 47, 1105 BA Amsterdam, The Netherlands. 1044-7431/$ – see front matter © 2011 Elsevier Inc. All rights reserved. doi:10.1016/j.mcn.2011.04.006

show enhanced plasticity relative to their mature neighbors (Schmidt-Hieber et al., 2004) during a critical period between 1 and 1.5 months after their birth (Ge et al., 2007b). Newborn granule cells can be activated during memory formation (Ramirez-Amaya et al., 2006), long term potentiation (LTP; Bruel-Jungerman et al., 2006) and are preferentially incorporated into spatial memory networks (Kee et al., 2007). However, most newborn granule cells do not incorporate into neuronal circuits and fail to survive (Dayer et al., 2003). Newborn neurons in the developing and adult hippocampus receive GABAergic inputs before glutamatergic inputs (Esposito et al., 2005; Overstreet-Wadiche et al., 2005). In vivo retroviral labeling of newly generated neurons in the adult hippocampus has allowed anatomical analysis of both the input (Toni et al., 2007) and output synapses (Faulkner et al., 2008) formed by these new neurons, revealing that contacts are preferentially formed with multi-synapse boutons and spines respectively. Input and output synapses both take

204

J.E. Cheyne et al. / Molecular and Cellular Neuroscience 47 (2011) 203–214

up to 2 months to reach anatomical maturity. Use of optogenetic retroviral vectors has shown that mature newly generated neurons in the adult brain form functional glutamatergic output synapses onto hilar interneurons, mossy cells and CA3 pyramidal cells (Toni et al., 2008). Electrophysiological recordings have shown that immature granule cells express functional GABA and glutamate receptors even when they are synaptically silent (Ambrogini et al., 2006). However there is limited anatomical data on the presence of synaptic proteins in newborn neurons. In precursor cells and newly generated neurons in hippocampal slices of adult rats (Nacher et al., 2007), the NR1 and NR2B subunits of the NMDA receptor were found to be present in most young granule cells and GFAP positive precursor cells (type-1) but absent from transiently amplifying precursors (type-2–3). However, due to newborn neurons being tightly packed into slices next to mature neurons the authors were not able to quantify differences in the expression of the NMDA receptor subunits between newborn neurons and mature neurons (Nacher et al., 2007). The continued occurrence of neurogenesis in vitro in hippocampal slice cultures (Kamada et al., 2004; Raineteau et al., 2004) offers advantages for studying the regulation of neurogenesis by various drugs or growth factors. Using this approach it has been shown that serum (Raineteau et al., 2004) and pilocarpine-induced seizure activity (Poulsen et al., 2005) inhibit neurogenesis whereas growth factors (Raineteau et al., 2004; Chechneva et al., 2005; Poulsen et al., 2005) and glutamate receptor antagonists (Poulsen et al., 2005) increase the generation of neurons by slice cultures. The dendrites and axons of newborn neurons in slice cultures have also been examined in detail along with mEPSCs and mIPSCs, showing that new neurons can integrate and mature normally in vitro (Raineteau et al., 2006). Neuronal precursors derived from embryonic stem cells can also be plated onto cultured hippocampal slices, where they functionally incorporate to provide an in vitro model of cell transplantation (Benninger et al., 2003). Here we describe how newborn neurons continue to be generated in vitro in dissociated hippocampal cultures. These neurons are easily amenable to anatomical and functional characterisation and provide a new model system for studying the synaptic integration of newly generated neurons. Our data show that the molecular makeup of the synapses of new neurons differs from those formed by mature granule cells and we predict that these molecular differences may provide clues to understanding how newborn neurons integrate into neuronal networks and how this integration of new neurons may be enhanced. Results Neurogenesis continues in dissociated hippocampal cultures Neurogenesis has been observed to continue in in vitro preparations including the organotypic hippocampal slice (Kamada et al., 2004; Raineteau et al., 2004). We aimed to determine whether neurogenesis also continues in the dissociated hippocampal cell culture preparation. This in vitro system is widely utilised to examine the molecular mechanisms of synapse formation and function as the neuronal and glial cell monolayer provides easy accessibility to exogenously express synaptic proteins and to perform detailed imaging and physiology on them. We observed that dissociated hippocampal cultures continue to produce new cells that could be phenotypically identified as neurons. The addition of BrdU (0.2 μM) or retroviral vectors (pQCMV-GFP or pQCAG-GFP, see methods; Gordon et al., 2007) at 0, 7, or 14 DIV revealed a population of cells at all 3 time points that were positive for either BrdU expression in the nucleus or expressed GFP respectively (Fig. 1A–E), indicating that these cells originated from cell division. Immunostaining for the neuronal marker protein microtubule associated protein 2 (MAP2) (De Camilli et al., 1984; Caceres et al., 1986) revealed that a subpopulation of these newly generated cells were neurons (Fig. 1A–E). Between 0–3

DIV, 15.9 ± 1.4% of total neurons (n = 9652) were newborn neurons. This increased to 24.2 ± 9.3% and then 34.3 ± 7.7% of total neurons being newborn neurons generated between 7–10 DIV (n = 10894) and 14–17 DIV (n = 15736) respectively. Therefore, neurogenesis is continuing beyond the day of plating of dissociated hippocampal cultures and in fact is increasing with time in vitro. To follow the differentiation and development of newly generated neurons born specifically at 7 DIV in dissociated hippocampal cultures and to determine whether these newborn neurons continue to undergo further neurogenesis, we labelled newborn neurons by retroviral transduction at 7 DIV and immunostained for the neuronal marker MAP2. GFP positive and MAP2 positive cells were then manually counted at 6, 8, 10, 12, 14 and 16 days after infection (DAI; Fig. 1F). The percentage of new neurons out of total retroviral-labelled cells increased with time in culture, from 26.60% of GFP expressing cells (n = 327) being new neurons (n = 82) at 6 DAI to 50.94% of GFP expressing cells (n = 159) being new neurons at 16 DAI (n = 81). These data suggest that labelled progenitor cells give rise to new cells that differentiate into neurons in vitro. As expected, the MAP2 negative dividing cells were positive for the expression of glial fibrillary acidic protein (not shown), a marker of progenitor cells and astrocytes (Rickmann et al., 1987; Seri et al., 2001; Garcia et al., 2004; Namba et al., 2005). Newborn neurons are immature or integrated granule cells We classified the newly generated neurons (labelled by retroviral infection at 7 DIV) into 2 groups based on their morphology: immature and integrated. Immature newborn neurons were small cells with short dendrites and limited dendritic branching (e.g. Figs. 1B–D and 2C-D). These neurons were frequently found in small clusters that likely originated from the transduction of one precursor cell. Immunostaining for presynaptic synapsin-I revealed that these neurons were devoid of synapses onto them (Fig. 1D). Integrated newborn neurons were large with long dendrites that extensively branched and had many synapsin positive puncta (Fig. 1E). The percentage of immature new neurons decreased with time in culture from 6 DAI to 16 DAI (Fig. 1F). This decrease was not due to an increase in immature neuron cell death over this time period as the identification of dead GFP-positive immature newborn neurons by propidium iodide labelling revealed cell death decreased from 17.14 ± 1.40 at 6 DAI to 7.26 ± 2.68 at 16 DAI. In contrast, the percentage of integrated new neurons continued to increase from 40.0% at 6 DAI (n = 20) up to 79.75% at 16 DAI (n = 63; Fig. 1F), consistent with this population increasing due to the maturation of immature newborn neurons into integrated newborn neurons. Therefore while immature and mature newborn neurons exist in dissociated hippocampal cultures, only the neurons with the more mature morphology have integrated into the neuronal network. To determine whether newborn neurons in dissociated hippocampal cultures labelled with retrovirus are granule cells, hippocampal cultures were immunostained for the mature granule cell marker calbindin D28K (Fig. 2A,B; Rami et al., 1987; Baimbridge, 1992). All immature newborn neurons lacked calbindin (n = 123), however integrated newborn neurons either expressed none (n = 1), low (n = 6) or high levels of calbindin (n = 68). Overall, 98.67% of integrated newborn neurons (n = 75) were calbindin D28K positive granule cells (n = 74; Fig. 2A). Those newborn neurons that lacked calbindin D28K are likely interneurons, as have previously been reported to be generated in the postnatal and adult hippocampus (Liu et al., 2003; Raineteau et al., 2004). We next compared and quantified the morphology of GFP-positive, MAP2 expressing immature and integrated newborn neurons with large mature granule cells (identified by positive calbindin D28K and MAP2 expression). At 10 DAI new neurons were found to have significantly smaller cell bodies, less dendritic branching and lower MAP2 expression

J.E. Cheyne et al. / Molecular and Cellular Neuroscience 47 (2011) 203–214

205

Fig. 1. Production of newborn neurons in postnatal dissociated rat hippocampal cultures. A–C: Immunostaining for BrdU incorporation (red) revealed a population of newly generated cells that expressed the neuronal microtubule associated protein MAP2 (green). A: Low power image of mature non-dividing neurons (BrdU negative, MAP2 positive) and neurons that have originated from cell division in vitro (BrdU positive and MAP2 positive). B: High power image of boxed region in part A. C: Immature BrdU and MAP2 positive newborn neurons were frequently localised in clusters in dissociated hippocampal cultures. D, E: Newborn neurons positive for GFP (green) and MAP2 (blue) expression, indicating successful transduction with retrovirus at 7 DIV (see methods). D: Immature newborn neuron devoid of synapsin positive puncta (red), compared with neighboring GFP negative, MAP2 positive mature neuron with numerous synapsin puncta (red). E: Integrated GFP positive, MAP2 positive newborn neuron with numerous synapsin puncta (red). Note GFPnegative mature neuron alongside. F: Quantification of newly generated neurons in hippocampal dissociated cultures at 6, 8, 10, 12, 14 and 16 days after infection with retrovirus (DAI) at 7 DIV. Scale bars for parts A–B: 50 μm. Scale bar for parts C–E: 10 μm.

than mature granule cells at 17 DIV (Figs. 2C–F and 3A). Specifically, immature (n = 14) and integrated (n = 10) new neurons had significantly smaller cell body diameter (Immature p b 0.001; Integrated p b 0.05), significantly less dendritic branching (Immature p b 0.001; Integrated p b 0.001) and a significantly lower mean intensity of MAP2 (Immature p b 0.001; Integrated p b 0.001) than mature granule cells (17 DIV; n = 15). In addition, comparing the immature newborn neurons to the integrated newborn neurons revealed that there were significant differences in cell body diameter, dendritic branching and MAP2 expression between these two classes of newborn neurons (p b 0.001

in all cases). Treatment with fluorescent phalloidin and immunostaining for α-actinin (Fig. 2C–F) revealed that newborn neurons also had fewer dendritic spines compared to mature neurons, as assessed by α-actinin expression (red immunostaining Fig. 2E–F). At 10 DAI, newborn neurons have a lower density (in puncta per μm: Newborn: 0.96 ± 0.12, n = 6; Mature: 1.45 ± 0.21, n = 6; p b 0.05), lower intensity (in arbitrary units: Newborn: 33.85 ± 0.50, n = 6; Mature: 36.31 ± 0.61, n = 6; p b 0.01) and smaller area (in μm2: Newborn: 0.056 ± 0.0046, n = 6; Mature: 0.096 ± 0.0085, n = 6; p b 0.005) of α-actinin puncta than mature neurons.

206

J.E. Cheyne et al. / Molecular and Cellular Neuroscience 47 (2011) 203–214

Fig. 2. Newborn neurons in dissociated hippocampal cultures (labelled with retrovirus at 7 DIV) are granule cells which mature in vitro. A: Identification of newborn granule neurons as GFP- (green) and MAP2-positive (blue) neurons that express the granule cell marker calbindin D28K (red). Note the GFP-negative, calbindin-positive mature granule cell above. B: In contrast, a mature pyramidal neuron is positive for MAP2 but lacks GFP and calbindin expression. C,D: GFP-positive retroviral transduced (green) immature newborn neurons (infected with retrovirus at 7 DIV) with low MAP2 expression (blue), fewer dendritic branches, smaller cell bodies and dendritic growth cones (red, phalloidin) but no dendritic spines. E: Example of a GFP-positive integrated newborn neuron (infected with retrovirus at 7 DIV) with a more mature morphology as evidenced by increased MAP2 expression (blue), dendritic branching and the presence of dendritic spines labelled with α-actinin expression (red). F: Mature GFP negative neuron with extensive dendritic branching, a large cell body, high MAP2 expression (blue) and a high density of dendritic spines labelled with α-actinin expression (red). Scale bar for parts A–F: 10 μm.

Synaptic integration of newborn neurons To examine synapses formed onto the integrated newborn neurons and to determine if this differed from synapses onto mature granule cells, neurons were infected with retrovirus at 7 DIV and synapsin-I expression was analyzed at 6 (Newborn n = 14; Mature n = 15), 8 (Newborn n = 12; Mature n = 12), 10 (Newborn n = 12; Mature n = 10), 12 (Newborn n = 4; Mature n = 3), 14 (Newborn n = 5; Mature n = 5) and 16 DAI (Newborn n = 4; Mature n = 7) (Fig. 3B). Only integrated newborn neurons were examined as immature newborn neurons have no synapsin puncta on their dendrites. As early as 6 DAI, the density of synapsin puncta on the dendrites of newborn neurons was not significantly different from the density of synapsin puncta on the dendrites of 13 DIV mature neurons

(Fig. 3B), indicating that integrated newborn neurons have a similar density of synapses on them as mature neurons. Specifically, at 6 DAI there were no differences in the density, intensity or the area of synapsin-I puncta in newborn and mature neurons (Newborn n = 14; Mature n = 15; p N 0.05 in all cases). Over the following 10 days the density of synapsin puncta continued to increase in both newborn and mature neurons. Intriguingly, from 12 DAI the density of synapsin puncta increased significantly more in the newborn neurons, such that by 16 DAI newborn neurons have a higher density of synapsin-I puncta onto them compared to 23 DIV mature neurons (p b 0.05) with no effect on the intensity of synapsin-I puncta or the area of synapsin-I puncta (p N 0.05 in both cases). The absence of synapsin puncta on immature newborn neurons and the increase in density of synapsin with increasing DAI for

J.E. Cheyne et al. / Molecular and Cellular Neuroscience 47 (2011) 203–214

207

Fig. 3. Morphological and functional development of newborn neurons. A: Quantification of cell body diameter, dendritic branching and level of MAP2 expression in immature and integrated newborn neurons at 10 DAI (labelled at 7 DIV) compared to mature granule neurons within the same cultures. Data are presented as a % of mature granule cell levels. B: Quantification of synapsin positive puncta forming onto GFP-positive and MAP2-positive newborn neurons at 6, 8, 10, 12, 14 and 16 DAI and compared to mature neurons from within the same cultures. In parts A and B, data are presented as mean ± SEM; * p b 0.05, ** p b 0.01, *** p b 0.005; Scale bars: 10 μm. C: Example action potentials in a GFP-positive newborn granule cell in response to a 1 second current injection. D: Example action potentials in a mature granule cell at 13 DIV in response to a 1 second current injection. E: Membrane properties of newborn and mature neurons in dissociated hippocampal cultures as assessed by whole cell patch clamping. A comparison of average resting membrane potential, capacitance and input resistance in GFP positive immature newborn neurons, GFP-positive integrated neurons at 6 DAI, and mature granule cells that were identified by their action potential properties (see methods). Data are presented as mean ± SEM; ** p b 0.01, *** p b 0.005. F: Example mEPSC currents from integrated newborn neurons at 10 and 17 DAI.

integrated newborn neurons suggests the presence of synapses onto newborn neurons is a major factor contributing to their survival and integration. We therefore examined the effect of increasing synaptic strength via the chemical induction of long-term potentiation (cLTP, see methods) on neurogenesis in our dissociated hippocampal cultures. We found that the induction of cLTP significantly increased both the number of immature newborn neurons by an average of 27.84 ± 3.96% and the number of integrated newborn neurons by 48.89 ± 8.79%, suggesting that stimulation of NMDARs can enhance the survival and/or the integration of newborn neurons. Electrical and synaptic properties of newborn granule cells To determine whether retroviral labelled granule cells have excitable membrane properties whole cell patch clamping was performed on new neurons transduced with either pQCMV-GFP or pQCAG-GFP retroviral vectors (Fig. 3C–F). Immature newborn neurons (e.g. Fig. 1B–D) had similar properties at all time points examined. These immature neurons did not fire action potentials upon injection of current (n = 9). In contrast, at 6 DAI integrated newborn neurons fired single action potentials (Fig. 3C; n = 9) in response to current injection. Control mature granule cells within the same culture preparation were observed to fire multiple mature action potentials (Fig. 3D; 13 DIV; n = 6). However no newborn

neurons at 6 DAI fired action potentials spontaneously in contrast to the majority of control mature granule cells at 13 DIV which fired action potentials spontaneously (n = 4 out of 5). We also examined the resting membrane potential, membrane capacitance and input resistance of newborn neurons and mature granule cells (Fig. 3E). The resting membrane potential of newborn neurons became more hyperpolarized, their membrane capacitance increased and their input resistance decreased (Fig. 3E), reflecting neuronal maturation. However at 6 DAI, resting membrane potential remained significantly different from control mature neurons (Fig. 3E; p b 0.01) indicating that these newborn neurons were not yet mature. To assess whether there are functional synapses forming onto newborn neurons, miniature EPSCs were measured in newborn neurons at 10/11 DAI and then again at 17/18 DAI (Fig. 3F). Functional synapses were detectable on all newborn neurons at 10/11 DAI, with the average mEPSC frequency 0.67 ± 0.33 Hz and the average mEPSC amplitude measuring 26.37 ± 4.32 pA (n = 7). Newborn neurons at 10/11 DAI also continued to show depolarized resting membrane potentials (−38.9 ± 4.96 mV; n = 9). After a further 7/8 DIV, newborn neurons at 17/18 DAI had a significantly more hyperpolarized average resting membrane potential (−53.78 ± 8.96 mV; n = 9; p b 0.001). Moreover, average mEPSC frequency had significantly increased in newborn neurons to 2.04 ± 0.77 Hz (p b 0.05), but average mEPSC amplitude was not changed (29.29 ± 3.24 pA; n = 6, p N 0.05). This is

208

J.E. Cheyne et al. / Molecular and Cellular Neuroscience 47 (2011) 203–214

supportive of an increase in the number of functional synapses occurring with maturation of newborn neurons, corresponding to the increase in synapsin-I density over this same period of development. Presynaptic protein expression at newborn neuron synapses The observed increases in synapsin-I expression and mEPSC frequency suggest that synapse formation is occurring frequently onto newborn neurons resulting in a higher density of synapses. We aimed to examine what pre- and postsynaptic proteins are present at synapses onto newborn neurons. We performed retroviral infection of hippocampal cultures at 7 DIV and performed this analysis at 10 DAI when both immature and integrated newborn neurons are abundant in the dissociated culture system to enable a direct comparison between both these 2 stages of differentiation and also with mature granule cells. We first examined the expression of Piccolo and Bassoon, proteins that mark both presynaptic terminals and also Piccolo Transport Vesicles (PTVs), which are known to carry presynaptic proteins critical for synapse formation (Zhai et al., 2000). At 10 DAI all immature neurons were found to lack detectable levels of Bassoon and Piccolo (not shown). In contrast, all integrated neurons had Bassoon and Piccolo puncta surrounding their dendrites (Fig. 4A). Overall, we found that at 10 DAI integrated newborn neurons have a higher density of puncta containing the presynaptic structural proteins Bassoon (n = 9; p b 0.01) and Piccolo (n = 11; p b 0.05) on their dendrites compared to mature granule cells (Bassoon n = 7; Piccolo n = 11) or mature pyramidal cells (Bassoon n = 6; Piccolo n = 11), but these puncta of Piccolo and Bassoon were not significantly different in either intensity or area (Fig. 4A–D; p N 0.05 in all cases). Synaptic inputs onto newborn neurons are GABAergic and glutamatergic We next examined whether the synapses onto newborn neurons labelled with retrovirus at 7 DIV are GABAergic or glutamatergic and what postsynaptic receptors types are expressed at these synapses. We first examined the expression of the α1 and β2/3 subunits of the GABAAR at 10 DAI in newborn neurons hippocampal cultures. All newborn neurons whether immature (n = 68) or integrated (n = 26) were found to completely lack the α1 subunit of the GABAAR, consistent with appearance of this subunit at 21 DIV in cultured hippocampal neurons (not shown; Swanwick et al., 2006). All immature newborn neurons also lacked the β2/3 subunit of the GABAAR (n = 27). In contrast, all integrated neurons expressed the β2/3 subunit of the GABAAR (n = 5, Fig. 5A). Expression of the β2/3 subunit of the GABAAR in integrated newborn neurons at 10 DAI (n = 5) was no different to that in mature granule cells (n = 7) or pyramidal cells (n = 4). Specifically, we found no difference in the density, intensity or area (p N 0.05 in all cases) of β2/3 subunit of the GABAAR puncta between newborn and mature neurons (Fig. 5A–D). This suggests that newborn neurons express similar levels of β2/3 subunit-containing GABAARs as mature granule cells. We also examined the expression of GAD-65, which localises at GABAergic synapses (Kaufman et al., 1991; Dupuy and Houser, 1996; Swanwick et al., 2006). Immature newborn neurons lacked GAD-65 puncta (not shown), whereas integrated newborn neurons exhibited significant GAD-65 labelling, reflecting GABAergic synapses onto them (Fig. 5A). When GAD-65 expression was compared between 10 DAI integrated newborn neurons (n = 7) and 17 DIV mature granule cells (n = 4) and pyramidal neurons (n = 6) no differences were observed in the density, intensity or area of GAD-65 puncta (p N 0.05 in all cases). This shows that within 10 days after cell birth newborn neurons already have the same density of GABAergic synapses with a similar level of GAD-65 expression as mature neurons. To examine whether excitatory glutamatergic synapses also form onto newborn neurons in dissociated hippocampal neurons, immu-

Fig. 4. Expression of the presynaptic proteins Piccolo and Bassoon onto the dendrites of 10 DAI newborn neurons (labelled with retrovirus at 7 DIV), mature granule cells and mature pyramidal neurons in dissociated hippocampal cultures. A: Example immunostaining for Bassoon (top) and Piccolo (below) in GFP positive newborn neurons (left) and mature GFP-negative neurons (right). B,C and D: Quantification of Piccolo and Bassoon puncta intensity, density and area respectively, in GFP positive newborn neurons and GFPnegative calbindin positive mature granule cells and mature pyramidal cells. Data are normalised to mature granule cells and presented as mean ± SEM; * p b 0.05, ** p b 0.01; Scale bar: 5 μm.

J.E. Cheyne et al. / Molecular and Cellular Neuroscience 47 (2011) 203–214

209

nostaining was performed for the obligatory NR1 subunit of the NMDA receptor (Monyer et al., 1992), the GluR1 subunit of the AMPA receptor or for Group I metabotropic glutamate receptors (mGluR5/1) at 10 DAI (Fig. 6A). All immature neurons were all found to lack detectable levels of any of the GluR subtypes we analysed (not shown). In contrast, all integrated neurons expressed all the GluR subtypes that we analysed, but the level of expression varied depending on the receptor subtype. At 10 DAI integrated newborn neurons (n = 6) have a lower density of NR1 puncta than mature granule cells (n = 35) or pyramidal cells (n = 40; Fig. 6A–D, p b 0.05 in both cases). No significant differences in the intensity (p N 0.05) or area of NR1 puncta were observed between newborn and mature granule cells or pyramidal neurons (p N 0.05). This reveals that newborn neurons have a lower density of NMDARs but suggests that the NR1 puncta in newborn neurons may contain the same amount of NMDARs as synapses of mature neurons. We also observed a lower density of GluR1 puncta in 10 DAI integrated newborn neurons (n = 6) compared with mature granule cells (n = 13) or pyramidal neurons (n = 14; Fig. 6C; p b 0.05). Moreover the intensity of GluR1 puncta in newborn neurons was also lower compared to mature granule cells or pyramidal neurons (Fig. 6B, p b 0.01). No significant difference in the area of GluR1 puncta was observed (Fig. 6D, p N 0.05). This suggests that newborn neuron synapses contain fewer AMPARs than synapses of mature granule cells. Interestingly, the expression of mGluRs was not significantly different in integrated newborn neurons (n = 7) compared to mature granule cells (n = 4) or pyramidal neurons (n = 4) at 10 DAI (Fig. 6A–D). Specifically, we found no difference in the density, intensity or area of mGluR5/1 puncta in integrated newborn neurons versus mature neurons (p N 0.05 in all cases). Therefore, in newborn neurons the expression of mGluR5/1 reaches mature levels within 10 days after their birth. Discussion

Fig. 5. Development of GABAergic synapses on the dendrites of 10 DAI newborn neurons (labelled with retrovirus at 7 DIV), mature granule cells and pyramidal neurons in dissociated hippocampal cultures. A: Example immunostaining for GAD (top) and beta 2/3 subunit of the GABAA receptor (bottom) in GFP-positive newborn neurons (left) and GFP-negative mature neurons (right). B,C,D: Quantification of GAD and beta2/3 puncta intensity, density and area respectively in GFP-positive newborn neurons and GFPnegative calbindin positive mature granule cells and mature pyramidal cells. Data are normalised to mature granule cells and presented as mean ± SEM; Scale bar: 5 μm.

Here we characterise a new model system for studying the incorporation of newborn neurons into existing neuronal networks. We have found that neurogenesis of granule cells continues in dissociated hippocampal cultures and by pairing retroviral labelling of newborn neurons with immunocytochemistry and electrophysiology we have shown that the integration of newborn neurons and the development of functional synaptic connections onto newborn neurons can proceed rapidly. Similar to what occurs in the brain (Ambrogini et al., 2004; Esposito et al., 2005; Gozlan and Ben-Ari, 2003; Liu et al., 1996), these new cells were identified as granule cells that received a high density of GABAergic synapses early in their development. Newborn neurons in dissociated culture also fired action potentials and exhibited membrane properties similar to young granule cells in the developing hippocampus (Liu et al., 2000). We propose that this in vitro model system of neurogenesis will be useful for answering questions that are difficult to experimentally perform in vivo. The ease of performing immunostaining, physiology, imaging, localisation and quantification of proteins in newborn neurons is a significant advantage offered by the dissociated culture system. Another major advantage of dissociated hippocampal cultures as a model system to study neurogenesis is that within the cultures newly generated neurons become synaptically integrated into existing neuronal networks. This differs from neurosphere-derived cultures in which all neurons are triggered to differentiate simultaneously and thus form a developing network where all neurons are essentially the same age. In addition, neurons do not need to be directed to differentiate as this occurs naturally in the culture environment. The high connectivity between neurons in dissociated culture enables study of the structure and function of both the input and output synapses of newborn neurons and how plasticity at these synapses could drive the functional integration of newborn neurons.

210

J.E. Cheyne et al. / Molecular and Cellular Neuroscience 47 (2011) 203–214

Newborn neurons were classified as either immature neurons with small cell bodies, short dendrites with limited dendritic branching that were devoid of synaptic contacts, or integrated neurons with branching dendrites with many synapsin puncta. The percentage of integrated neurons increased with time with some new neurons

integrating rapidly within 6 DAI. However, not all new neurons become synaptically integrated and a proportion of immature neurons fail to integrate and subsequently undergo apoptosis as occurs in vivo (Biebl et al., 2000; Sun et al., 2004). At present it is unknown whether immature new neurons in hippocampal cultures have a defined time window after birth in which they need to integrate in order to survive. Our data suggest that the differences in functional synapses, morphology, ability to fire action potentials, presynaptic protein expression on input synapses and the presence versus absence of glutamate receptor expression in integrated versus immature newborn neurons respectively reveals that the formation of synaptic contacts onto newborn neurons is critical for their morphological maturation, survival and integration. What specific factor(s) drive the development to the integrated phenotype remains to be determined but our data suggest that stimulation of NMDARs via cLTP can drive this process. Moreover, the rapid increase in the formation of functional synapses and synapsin-positive presynaptic puncta onto newborn neurons between 6 and 12 DAI when new neurons appear to be preferentially recruited into neuronal networks suggests that synapse formation from mature neurons onto newborn neurons plays a major role in driving newborn neuron integration. Further evidence of significant synapse formation occurring onto integrating newborn neurons was revealed by the higher density of puncta containing the presynaptic structural proteins Piccolo and Bassoon surrounding the dendrites of integrating newborn neurons. As the density of Piccolo and Bassoon puncta was also significantly higher than the density of synapsin puncta, these data suggest that not all of the Bassoon and Piccolo puncta are localised at new synaptic contacts onto newborn neurons. Indeed our mEPSC data at this time point (10/11 DAI) reveal that mEPSC frequency was significantly lower compared with 17/18 DAI neurons, therefore not all Bassoon and Piccolo puncta represent functional synaptic contacts at 10 DAI. It has previously been shown that many presynaptic active zone proteins traffic to developing synapses in large dense-core vesicles known as Piccolo Transport Vesicles (PTVs) (Shapira et al., 2003; Ziv and Garner, 2004), however synapsin-I is not contained in these vesicles (Ahmari et al., 2000). It is likely that a proportion of the Bassoon and Piccolo puncta we are detecting are therefore not located at synapses but are PTVs trafficking along axons on route to form new synapses onto newborn neurons. In contrast to integrated newborn neurons, no Piccolo or Bassoon expression was detectable onto immature newborn neurons. Therefore immature newborn neurons likely represent an early stage of neuron differentiation that is refractory to synaptogenesis and the trafficking of PTVs to develop presynaptic terminals onto newly generated neurons. It will be of significant interest to determine what signals a change in maturation to then allow newborn neurons to accept presynaptic input. The formation of the postsynaptic apparatus is also important for newborn neurons to receive information from surrounding neurons and consequently alter their excitability and output. We observed that integrated newborn neurons express α-actinin, a major component of postsynaptic dendritic spines (Wyszynski et al., 1998; Nakagawa et al., 2004), suggesting integrated neurons contain dendritic spines, but at a lower density than mature neurons. Previous studies have shown that new neurons in the adult mouse hippocampus lack dendritic spines at 14 DAI (Esposito et al., 2005) but begin to form spines within 15–16 days after their generation (Zhao et al., 2006). Spine growth Fig. 6. Development of glutamatergic synapses on the dendrites of 10 DAI newborn neurons (labelled with retrovirus at 7 DIV), mature granule cell and pyramidal cells in dissociated hippocampal cultures. A: Example immunostaining for NR1 subunit of the NMDA receptor (top), GluR1 subunit of the AMPA receptor (middle) and mGluR1/5 (bottom) in GFP-positive newborn neurons (left) and GFP-negative mature neurons (right). B,C,D: Quantification of NR1, GluR1 and mGluR1/5 puncta intensity, density and area respectively in GFP-positive newborn neurons and GFP-negative mature granule cells and pyramidal neurons. Data are normalised to mature granule cells and presented as mean ± SEM. * p b 0.05; ** p b 0.01; Scale bar: 5 μm.

J.E. Cheyne et al. / Molecular and Cellular Neuroscience 47 (2011) 203–214

peaks in the first 3–4 weeks and spine motility of new granule cells is greatest 1 month (Zhao et al., 2006). The density of spines increases with development of new neurons in the mouse brain but newly generated neurons in the adult mouse hippocampus still have a significantly lower density of dendritic spines than mature neurons even at 60 DAI (Toni et al., 2007). We found newly generated neurons in dissociated cultures to have a higher spine density compared with new neurons in the adult brain which may be the result of differences in age, species and conditions. However this result suggests that maturation of new neurons occurs faster in developing systems in vitro than in the adult brain in line with previous research that has shown that neuronal development is faster in the developing brain than in the adult brain (Overstreet-Wadiche et al., 2006; Zhao et al., 2006). Our analysis of postsynaptic glutamate receptor expression suggests that glutamatergic synapses form onto dendritic spines of newborn neurons but that these glutamatergic inputs develop slower than GABAergic inputs. This result is in agreement with previous literature that has shown that new granule cells both in the adult and in the developing brain receive GABAergic inputs before glutamatergic inputs (Ambrogini et al., 2004; Esposito et al., 2005; Gozlan and Ben-Ari, 2003; Liu et al., 1996). Due to the high intracellular Cl− content of young neurons these GABAergic inputs are depolarizing and are thought to be important for the survival and integration of these young cells (Ben-Ari, 2002; Ge et al., 2006, 2007a). Postsynaptic NMDAR expression has previously been shown to play a primary role in the survival of newborn neurons (Tashiro et al., 2006). We observed that postsynaptic NR1 was not detectable in immature newborn neurons. Integrated newborn neurons however do express significant levels of NR1 which were at a lower density than those on mature neurons but intensity measurements suggest that these NR1 puncta contain similar amounts of NMDARs as those of mature neurons. While NMDAR expression is not required for synapses to form, these receptors are critical for synapse maturation and the ability to undergo synaptic plasticity and in this way NMDARs play a critical role in the survival of newly generated neurons (Tashiro et al., 2006). The expression of NMDARs at synapses of newly generated neurons and the observed increase in integrated newborn neurons in response to NMDAR stimulation via cLTP suggests that Ca2+ influx through NMDARs enables the induction of LTP and increases neuronal integration. AMPARs may then be recruited and synaptic circuits that include newly generated neurons strengthened. In contrast, immature neurons that lack NMDARs will not be able to undergo NMDAR-dependent plasticity, and this may decrease their odds of integration. The fact that we observed a significant effect of cLTP on the number of immature newborn neurons is therefore unlikely due to a direct effect on these neurons, and likely reflects that the induction of cLTP promotes a more permissive environment for continued production and differentiation of immature newborn neurons as well as for the maturation of integrated newborn neurons. In the adult brain NR1 and NR2B are present in newly generated neurons, precursor cells (type 1) but absent from transiently amplifying progenitors (type 2–3) (Nacher et al., 2007). There is also evidence that suggests that neural progenitor cells express NMDAR subunits but lack functional NMDAR channels (Muth-Kohne et al., 2010). In support of this in differentiated mouse neuroectodermal progenitor cultures functional NMDARs do not appear until 6 days after induction of differentiation however individual subunits expression begins earlier (Jelitai et al., 2002). Therefore it appears that during development of new neurons expression of functional NMDARs may only occur after differentiation is complete and neurite outgrowth has begun (Varju et al., 2001), suggesting again that differentiation is not complete in our immature newborn neuron population. In contrast to NR1, the density and the intensity of expression of the AMPAR subunit GluR1 was lower in integrated newborn neurons, suggesting that a proportion of their synapses lack AMPARs and are silent at resting membrane potentials. This is consistent with the lower frequency of mEPSCs at this same developmental time point.

211

Therefore these neurons have significant potential to further strengthen their connections with the existing neuronal networks. Again, the lack of detectable GluR1 on immature neurons, and the frequent observation of these neurons in small isolated clusters will result in a lack of stimulation of these cells which may also decrease their survival and/or integration. Interestingly, group I mGluR expression was at mature levels by 10 DAI in integrated newborn neurons suggesting that these receptors are important in the early integration of newborn neurons. mGluR5 but not mGluR1 has previously been found in newborn cells in zones of active neurogenesis in the embryonic and postnatal brain (Di Giorgi Gerevini et al., 2004). In addition, several previous reports suggest a role for mGluR5 in the proliferation and self renewal of progenitor cells (Cappuccio et al., 2005; Di Giorgi-Gerevini et al., 2005). Activation of mGluR5 in cultured ESCs regulates proliferation through the expression of key transcription factors that are important for self renewal, Oct-4 and Nanog (Cappuccio et al., 2005). Furthermore, activation of mGluR5 increases proliferation of cultured mouse NPCs whereas antagonism reduces proliferation and survival (Di Giorgi-Gerevini et al., 2005). In addition, mGluR5 KO mice and mice treated with the mGluR5 antagonist MPEP show a reduced number of dividing progenitors in both the SVZ and the dentate gyrus (Di Giorgi-Gerevini et al., 2005). Taken together these results suggest that mGluR5 expression is very important for the proliferation of progenitor cells. However other published results show that mRNA levels of mGluR1 and mGluR5 are late to rise during the neuronal differentiation of the mouse embryocarcinoma cell line P19 and primary embryonic neurons (Heck et al., 1997). These results suggest that group I mGluRs are not involved in differentiation of neurons but are involved in early development processes such as neurite outgrowth (Heck et al., 1997). Our results support a role for group I mGluRs in neuronal outgrowth as they are highly expressed in young integrating neurons at a time where extensive neurite outgrowth is occurring. In addition, group I mGluR-mediated regulation of NMDARs and synaptic plasticity may be important early in the integration of new neurons. High levels of group I mGluRs in young neurons could result in robust regulation of NMDAR currents (O'Connor et al., 1994) and modulate LTP (Bortolotto et al., 1994), which may be important for successful neuronal integration. Conclusions In conclusion, here we have characterised a new model system for studying the integration of newborn neurons into existing circuits. Importantly, our analysis of the properties of these neurons has shown that new neurons develop similarly in hippocampal cultures as they do in vivo. The high density of presynaptic protein puncta on their dendrites shows there is high potential for mature neurons to form new synapses onto them to preferentially recruit them into neuronal networks. We have observed that integrated newly generated neurons, but not immature newborn neurons, express many of the synaptic proteins which are found at mature synapses and that these newborn neurons undergo synaptic transmission and plasticity and thereby contribute to the neuronal network. The existence of immature and integrated newborn neurons in this in vitro preparation may therefore provide an interesting model system to drive expression of synaptic proteins in immature neurons to determine the underpinnings of newborn neuron integration and how this could contribute to brain function and repair. Experimental methods Retroviral packaging Retroviral expression plasmids pQCMV-GFP (Gordon et al., 2007) and pQCAG-GFP (Zhao et al., 2006; Tashiro et al., 2006) were packaged into vesicular stomatitis virus (VSV-G) retroviral vector

212

J.E. Cheyne et al. / Molecular and Cellular Neuroscience 47 (2011) 203–214

particles using GP2-293 cells (BD Biosciences) as described previously (Gordon et al., 2007). Briefly, 2.0 × 10 6 GP2-293 cells were cotransfected with pQCMV-GFP or pQCAG-GFP and VSV-G plasmids by calcium phosphate precipitation (Clontech) for 10 h. The media containing retrovirus was collected 48 h after transfection and concentrated by centrifugation overnight at 7500 ×g or for 2 h at 65,000 ×g. The concentrated viral vector pellet was resuspended to 0.5% of the original volume in 40 mg/ml lactose (Sigma) in phosphate buffered saline (PBS) and stored at −80 °C until use. Retroviral titre was determined by transducing human embryonic kidney cells (HEK293) with 10-fold serial dilutions of concentrated virus. Retroviral titres between 10 8 and 10 10 colony forming units/mL led to successful labelling of newly generated cells in hippocampal cultures. As previously reported for pQCMV-GFP the GFP expression in infected cells is downregulated with neuronal differentiation and/or maturation (Gordon et al., 2007). To enhance GFP fluorescence in newborn cells cultures were also immunostained for GFP expression. We found that with pQCAG-GFP the expression of GFP in newly generated neurons remained relatively stable in agreement with previous studies (Zhao et al., 2006). Hippocampal cell culture Dissociated hippocampal cultures were prepared from male and female P0 Wistar rats using procedures approved by the University of Auckland Animal Ethics committee. Hippocampal cultures were prepared and maintained as previously described (Cheyne and Montgomery, 2008). Briefly, hippocampal neurons were dissociated from P0 rat hippocampi with papain and plated onto poly-D-lysine coated coverslips. At 0, 7 or 14 days in vitro (DIV) newly generated cells were labelled with either 0.2 μM 5′-bromo-2-deoxyuridine (BrdU, Applichem) for 3 days or retrovirus encoding GFP for 5 h. Polybrene (1 mg/mL; Sigma) was used to improve retroviral transduction. Neurons were fixed with 3.7% formaldehyde at 6, 7, 8, 10, 12 or 14 days after labelling. Chemical LTP was induced by addition of 200 μM glycine, 20 μM bicuculline (Sigma 14343) and 1 μM strychnine to the culture medium for 5 min at 7 days after beginning BrdU (BrdU from 7–10 DIV) at 14 DIV. Neurons were then fixed at 18 DIV. To assess newborn neuron viability, propidium iodide (5 mM) was added to the culture media for 1 h at 37 °C. Imaging was then performed immediately in neural basal medium without phenol red. Dead newborn neurons were identified as propidium iodide and GFPpositive neurons. Whole cell patch clamping Neurons were bathed in carbogen (95% O2, 5% CO2) bubbled ACSF (in mM: 120 NaCl, 3 KCl, 2 CaCl2, 1.25 NaH2PO4, 2 MgSO4, 20 D-(+)glucose, 26 NaHCO3). The internal solution consisted of (in mM): 120 K gluconate, 40 HEPES, 5 MgCl2, 2 NaATP, 0.3 NaGTP). Recordings were performed at room temperature (21 °C). Hippocampal cultures were visualised using an Olympus microscope (BX51WI) using differential interference microscopy (DIC) and retroviral labelled neurons were visualised via excitation at 530–550 nm. Electrode resistance was between 5–10 MΩ. Patch clamp recordings were obtained using a MultiClamp 700B Commander (Molecular Devices). Data acquisition and analysis was performed using AxoGraph X (AxoGraph Scientific) and pClamp 9 (Molecular Devices) software. Events were sampled at 10 kHz and low pass filtered at 2 kHz. Series resistance (Rs) was monitored throughout all experiments and results were not included if significant variation (N20%) occurred during the experiment. Action potentials were induced in neurons by a 1000 ms current injection of 50–400 pA. Action potential firing patterns and after-hyperpolarisation size of individual neurons were used to classify cells types (Toni et al., 2008). Neurons that fired repeatedly but showed little after-hyperpolarisation were classified as granule

cells. Miniature excitatory postsynaptic currents (mEPSCs) were detected in voltage clamp (−65 mV) by continuous data acquisition in ACSF containing 1 μM tetrodotoxin and 100 μM picrotoxin as previously described (Waites et al., 2009). mEPSC events were detected and analysed in MiniAnalysis (Synaptosoft version 6.0.3). As mEPSC amplitude distribution did not follow a normal distribution the nonparametric Kolmogorov–Smirnov test was used to determine the probability of a significant difference between mEPSC amplitudes in mature and newborn neurons. Immunocytochemistry Neurons were permeabilized with 0.25% Triton and non-specific binding prevented by a 30 min incubation in either blocking solution (2% BSA, 2% glycine, 0.2% gelatine, 50 mM NH4Cl) or 3% goat or horse serum (Vector). Primary antibody binding was performed overnight at 4 °C. Secondary antibody was performed for 1 h at room temperature. Coverslips were washed and mounted onto slides (Vectashield) for imaging. To immunostain for BrdU, cells were fixed, permeabilized and washed as described above then incubated in 2 M HCl for 1 h at 37 °C to denature the DNA. Cells were washed for 10 min in 0.1 M Na + tetraborate (pH 8.5, Sigma) then washed twice with 1× PBS. Staining then proceeded as described above. The following primary antibodies were used: mouse anti-BrdU (1:2500, Chemicon MAB3222), mouse anti-MAP2 (1:500, Sigma Aldrich M4403), rabbit anti-MAP2 (1:500, Chemicon AB5622), chicken anti-MAP2 (1:2000, Abcam ab5392), rabbit anti-GFAP (1:500, DAKO Z0334), chicken anti-GFP (1:1000, Abcam AB13970), mouse anti-synapsin-I (1:500, BD Pharmingen 611392), rabbit anti-synapsin-I (1:250, Sigma S193), mouse anticalbindin D28K (1:5000, Swant 300), goat anti-calbindin D28K (1:250, Santa Cruz sc-7691), Alexa Fluor 568 phalloidin (1:40, Molecular Probes A12380), mouse anti-α actinin (1:2000, Sigma A7811), mouse anti-Bassoon (1:400, gift of CC Garner), rabbit anti-Piccolo (1:400, gift of CC Garner), rabbit anti-GAD-65/67 (1:100, Sigma G5163); for GABARs, mouse anti-GABAA β2/3 (1:500, Chemicon AB5952) and rabbit anti-GABAA α1 (1:1000, gift from the University of Zurich), mouse anti-NR1 subunit of NMDAR (1:750, Chemicon MAB363, methanol), rabbit anti-NR1 subunit of NMDAR (1:1000, Chemicon AB9864), rabbit anti-GluR1 subunit of AMPAR (1:25, Calbiochem PC246 or 1:1000, Chemicon AB1504) and rabbit anti-mGluR1/5 (1:500, Sapphire Bioscience NB300-126). The different cell types in the dissociated hippocampal cultures were defined by immunocytochemical labelling with MAP2 as a neuronal marker in all cases. Specifically, newborn neurons were defined as calbindin D28K and GFP/BrdU-positive neurons, mature granule cells were identified as large calbindin-positive neurons and pyramidal neurons were defined as calbindin D28K negative. Image acquisition and analysis Z-stack (0.5 μm) images of dendrites were obtained on a Zeiss Axioscope with a CCD digital camera or an Olympus FV1000 using a 40×, 60× or 63× oil objective as previously described (Cheyne and Montgomery, 2008). Exposure times were kept constant for newly generated and mature neurons to ensure that intensity measures were directly comparable. Image analysis was performed using Image J. Z-stacks were converted to 8-bit, merged into maximum projections and background subtracted to remove the diffuse protein expression within dendrites. Images in which staining was dim or the background was high were excluded from the analysis. Dendritic branching was analysed in 40 x images of the whole cell by manual counting of the number of branch points. MAP2 intensity was measured in the entire dendritic tree and averaged for each cell. For analysis of the punctate expression of synapsin, Bassoon, Piccolo, GluR1, NR1, mGluR1/5, β2/3 and GAD dendritic regions were isolated such that that no branching or overlapping regions were included. Images were manually

J.E. Cheyne et al. / Molecular and Cellular Neuroscience 47 (2011) 203–214

thresholded to select only puncta that were greater than twofold above image background. Puncta were analyzed using the Analyze Particles function in ImageJ so that the average puncta intensity, area and number of puncta could be determined. The number of puncta was then divided by the length of dendrite analyzed to get the density of puncta per μm of dendrite. All data is presented with 2 decimal places plus or minus (±) the standard error of the mean (SEM) where n = number of fields or neurons analysed. Statistical analysis was performed using Student's t-tests (one-tailed distribution, unequal variance, * = p b 0.05, ** = p b 0.01 and *** = p b 0.005). Acknowledgments We acknowledge the assistance of the Connor lab for retroviral production and the Montgomery lab for helpful discussion. We are grateful to Janusz Lipski for helpful guidance and advice. This work was supported by the Health Research Council and The Neurological Foundation of New Zealand. References Ahmari, S.E., Buchanan, J., Smith, S.J., 2000. Assembly of presynaptic active zones from cytoplasmic transport packets. Nat. Neurosci. 3, 445–451. Ambrogini, P., Lattanzi, D., Ciuffoli, S., Agostini, D., Bertini, L., Stocchi, V., Santi, S., Cuppini, R., 2004. Morpho-functional characterization of neuronal cells at different stages of maturation in granule cell layer of adult rat dentate gyrus. Brain Res. 1017, 21–31. Ambrogini, P., Minelli, A., Lattanzi, D., Ciuffoli, S., Fanelli, M., Cuppini, R., 2006. Synapticallysilent immature neurons show GABA and glutamate receptor-mediated currents in adult rat dentate gyrus. Arch. Ital. Biol. 144, 115–126. Baimbridge, K.G., 1992. Calcium-binding proteins in the dentate gyrus. Epilepsy Res. 7, 211–220. Ben-Ari, Y., 2002. Excitatory actions of GABA during development: the nature of the nurture. Nat. Rev. Neurosci. 3, 728–739. Benninger, F., Beck, H., Wernig, M., Tucker, K.L., Brustle, O., Scheffler, B., 2003. Functional integration of embryonic stem cell-derived neurons in hippocampal slice cultures. J. Neurosci. 23, 7075–7083. Biebl, M., Cooper, C.M., Winkler, J., Kuhn, H.G., 2000. Analysis of neurogenesis and programmed cell death reveals a self-renewing capacity in the adult rat brain. Neurosci. Lett. 291, 17–20. Bortolotto, Z.A., Bashir, Z.I., Davies, C.H., Collingridge, G.L., 1994. A molecular switch activated by metabotropic glutamate receptors regulates induction of long-term potentiation. Nature 368, 740–743. Bruel-Jungerman, E., Davis, S., Rampon, C., Laroche, S., 2006. Long-term potentiation enhances neurogenesis in the adult dentate gyrus. J. Neurosci. 26, 5888–5893. Caceres, A., Banker, G.A., Binder, L., 1986. Immunocytochemical localization of tubulin and microtubule-associated protein 2 during the development of hippocampal neurons in culture. J. Neurosci. 6, 714–722. Cappuccio, I., Spinsanti, P., Porcellini, A., Desiderati, F., De Vita, T., Storto, M., Capobianco, L., Battaglia, G., Nicoletti, F., Melchiorri, D., 2005. Endogenous activation of mGlu5 metabotropic glutamate receptors supports self-renewal of cultured mouse embryonic stem cells. Neuropharmacology 49 (Suppl. 1), 196–205. Chechneva, O., Dinkel, K., Schrader, D., Reymann, K.G., 2005. Identification and characterization of two neurogenic zones in interface organotypic hippocampal slice cultures. Neuroscience 136, 343–355. Cheyne, J.E., Montgomery, J.M., 2008. Plasticity-dependent changes in metabotropic glutamate receptor expression at excitatory hippocampal synapses. Mol. Cell. Neurosci. 37, 432–439. Dayer, A.G., Ford, A.A., Cleaver, K.M., Yassaee, M., Cameron, H.A., 2003. Short-term and long-term survival of new neurons in the rat dentate gyrus. J. Comp. Neurol. 460, 563–572. De Camilli, P., Miller, P.E., Navone, F., Theurkauf, W.E., Vallee, R.B., 1984. Distribution of microtubule-associated protein 2 in the nervous system of the rat studied by immunofluorescence. Neuroscience 11, 817–846. Di Giorgi Gerevini, V.D., Caruso, A., Cappuccio, I., Ricci Vitiani, L., Romeo, S., Della Rocca, C., Gradini, R., Melchiorri, D., Nicoletti, F., 2004. The mGlu5 metabotropic glutamate receptor is expressed in zones of active neurogenesis of the embryonic and postnatal brain. Brain Res. Dev. Brain Res. 150, 17–22. Di Giorgi-Gerevini, V., Melchiorri, D., Battaglia, G., Ricci-Vitiani, L., Ciceroni, C., Busceti, C.L., Biagioni, F., Iacovelli, L., Canudas, A.M., Parati, E., De Maria, R., Nicoletti, F., 2005. Endogenous activation of metabotropic glutamate receptors supports the proliferation and survival of neural progenitor cells. Cell Death Differ. 12, 1124–1133. Dupuy, S.T., Houser, C.R., 1996. Prominent expression of two forms of glutamate decarboxylase in the embryonic and early postnatal rat hippocampal formation. J. Neurosci. 16, 6919–6932. Esposito, M.S., Piatti, V.C., Laplagne, D.A., Morgenstern, N.A., Ferrari, C.C., Pitossi, F.J., Schinder, A.F., 2005. Neuronal differentiation in the adult hippocampus recapitulates embryonic development. J. Neurosci. 25, 10074–10086.

213

Faulkner, R.L., Jang, M.H., Liu, X.B., Duan, X., Sailor, K.A., Kim, J.Y., Ge, S., Jones, E.G., Ming, G.L., Song, H., Cheng, H.J., 2008. Development of hippocampal mossy fiber synaptic outputs by new neurons in the adult brain. PNAS 105, 14157–14162. Garcia, A.D., Doan, N.B., Imura, T., Bush, T.G., Sofroniew, M.V., 2004. GFAP-expressing progenitors are the principal source of constitutive neurogenesis in adult mouse forebrain. Nat. Neurosci. 7, 1233–1241. Ge, S., Goh, E.L., Sailor, K.A., Kitabatake, Y., Ming, G.L., Song, H., 2006. GABA regulates synaptic integration of newly generated neurons in the adult brain. Nature 439, 589–593. Ge, S., Pradhan, D.A., Ming, G.L., Song, H., 2007a. GABA sets the tempo for activitydependent adult neurogenesis. Trends Neurosci. 30, 1–8. Ge, S., Yang, C.H., Hsu, K.S., Ming, G.L., Song, H., 2007b. A critical period for enhanced synaptic plasticity in newly generated neurons of the adult brain. Neuron 54, 559–566. Gordon, R.J., Tattersfield, A.S., Vazey, E.M., Kells, A.P., McGregor, A.L., Hughes, S.M., Connor, B., 2007. Temporal profile of subventricular zone progenitor cell migration following quinolinic acid-induced striatal cell loss. Neuroscience 146, 1704–1718. Gozlan, H., Ben-Ari, Y., 2003. Interneurons are the source and the targets of the first synapses formed in the rat developing hippocampal circuit. Cereb. Cortex 13, 684–692. Heck, S., Enz, R., Richter-Landsberg, C., Blohm, D.H., 1997. Expression of eight metabotropic glutamate receptor subtypes during neuronal differentiation of P19 embryocarcinoma cells: a study by RT-PCR and in situ hybridization. Brain Res. Dev. Brain Res. 101, 85–91. Jelitai, M., Schlett, K., Varju, P., Eisel, U., Madarasz, E., 2002. Regulated appearance of NMDA receptor subunits and channel functions during in vitro neuronal differentiation. J. Neurobiol. 51, 54–65. Kamada, M., Li, R.Y., Hashimoto, M., Kakuda, M., Okada, H., Koyanagi, Y., Ishizuka, T., Yawo, H., 2004. Intrinsic and spontaneous neurogenesis in the postnatal slice culture of rat hippocampus. Eur. J. Neurosci. 20, 2499–2508. Kaufman, D.L., Houser, C.R., Tobin, A.J., 1991. Two forms of the gamma-aminobutyric acid synthetic enzyme glutamate decarboxylase have distinct intraneuronal distributions and cofactor interactions. J. Neurochem. 56, 720–723. Kee, N., Teixeira, C.M., Wang, A.H., Frankland, P.W., 2007. Preferential incorporation of adult-generated granule cells into spatial memory networks in the dentate gyrus. Nat. Neurosci. 10, 355–362. Liu, Y.B., Lio, P.A., Pasternak, J.F., Trommer, B.L., 1996. Developmental changes in membrane properties and postsynaptic currents of granule cells in rat dentate gyrus. J. Neurophysiol. 76, 1074–1088. Liu, X., Tilwalli, S., Ye, G., Lio, P.A., Pasternak, J.F., Trommer, B.L., 2000. Morphologic and electrophysiologic maturation in developing dentate gyrus granule cells. Brain Res. 856, 202–212. Liu, S., Wang, J., Zhu, D., Fu, Y., Lukowiak, K., Lu, Y.M., 2003. Generation of functional inhibitory neurons in the adult rat hippocampus. J. Neurosci. 23, 732–736. Monyer, H., Sprengel, R., Schoepfer, R., Herb, A., Higuchi, M., Lomeli, H., Burnashev, N., Sakmann, B., Seeburg, P.H., 1992. Heteromeric NMDA receptors: molecular and functional distinction of subtypes. Science 256, 1217–1221. Muth-Kohne, E., Terhag, J., Pahl, S., Werner, M., Joshi, I., Hollmann, M., 2010. Functional excitatory GABA(A) receptors precede ionotropic glutamate receptors in radial glia-like neural stem cells. Mol. Cell. Neurosci. 43, 209–221. Nacher, J., Varea, E., Miguel Blasco-Ibanez, J., Gomez-Climent, M.A., Castillo-Gomez, E., Crespo, C., Martinez-Guijarro, F.J., McEwen, B.S., 2007. N-methyl-D-aspartate receptor expression during adult neurogenesis in the rat dentate gyrus. Neuroscience 144, 855–864. Nakagawa, T., Engler, J.A., Sheng, M., 2004. The dynamic turnover and functional roles of alpha-actinin in dendritic spines. Neuropharmacology 47, 734–745. Namba, T., Mochizuki, H., Onodera, M., Mizuno, Y., Namiki, H., Seki, T., 2005. The fate of neural progenitor cells expressing astrocytic and radial glial markers in the postnatal rat dentate gyrus. Eur. J. Neurosci. 22, 1928–1941. O'Connor, J.J., Rowan, M.J., Anwyl, R., 1994. Long-lasting enhancement of NMDA receptor-mediated synaptic transmission by metabotropic glutamate receptor activation. Nature 367, 557–559. Overstreet-Wadiche, L.S., Bromberg, D.A., Bensen, A.L., Westbrook, G.L., 2005. GABAergic signaling to newborn neurons in dentate gyrus. J. Neurophysiol. 94, 4528–4532. Overstreet-Wadiche, L.S., Bensen, A.L., Westbrook, G.L., 2006. Delayed development of adult-generated granule cells in dentate gyrus. J. Neurosci. 26, 2326–2334. Poulsen, F.R., Blaabjerg, M., Montero, M., Zimmer, J., 2005. Glutamate receptor antagonists and growth factors modulate dentate granule cell neurogenesis in organotypic, rat hippocampal slice cultures. Brain Res. 1051, 35–49. Raineteau, O., Rietschin, L., Gradwohl, G., Guillemot, F., Gahwiler, B.H., 2004. Neurogenesis in hippocampal slice cultures. Mol. Cell. Neurosci. 26, 241–250. Raineteau, O., Hugel, S., Ozen, I., Rietschin, L., Sigrist, M., Arber, S., Gahwiler, B.H., 2006. Conditional labeling of newborn granule cells to visualize their integration into established circuits in hippocampal slice cultures. Mol. Cell. Neurosci. 32, 344–355. Rami, A., Brehier, A., Thomasset, M., Rabie, A., 1987. Cholecalcin (28-kDa calciumbinding protein) in the rat hippocampus: development in normal animals and in altered thyroid states. An immunocytochemical study. Dev. Biol. 124, 228–238. Ramirez-Amaya, V., Marrone, D.F., Gage, F.H., Worley, P.F., Barnes, C.A., 2006. Integration of new neurons into functional neural networks. J. Neurosci. 26, 12237–12241. Rickmann, M., Amaral, D.G., Cowan, W.M., 1987. Organization of radial glial cells during the development of the rat dentate gyrus. J. Comp. Neurol. 264, 449–479. Schmidt-Hieber, C., Jonas, P., Bischofberger, J., 2004. Enhanced synaptic plasticity in newly generated granule cells of the adult hippocampus. Nature 429, 184–187. Seri, B., Garcia-Verdugo, J.M., McEwen, B.S., Alvarez-Buylla, A., 2001. Astrocytes give rise to new neurons in the adult mammalian hippocampus. J. Neurosci. 21, 7153–7160.

214

J.E. Cheyne et al. / Molecular and Cellular Neuroscience 47 (2011) 203–214

Shapira, M., Zhai, R.G., Dresbach, T., Bresler, T., Torres, V.I., Gundelfinger, E.D., Ziv, N.E., Garner, C.C., 2003. Unitary assembly of presynaptic active zones from PiccoloBassoon transport vesicles. Neuron 38, 237–252. Sun, W., Winseck, A., Vinsant, S., Park, O.H., Kim, H., Oppenheim, R.W., 2004. Programmed cell death of adult-generated hippocampal neurons is mediated by the proapoptotic gene Bax. J. Neurosci. 24, 11205–11213. Swanwick, C.C., Murthy, N.R., Mtchedlishvili, Z., Sieghart, W., Kapur, J., 2006. Development of gamma-aminobutyric acidergic synapses in cultured hippocampal neurons. J. Comp. Neurol. 495, 497–510. Tashiro, A., Sandler, V.M., Toni, N., Zhao, C., Gage, F.H., 2006. NMDA-receptor-mediated, cell-specific integration of new neurons in adult dentate gyrus. Nature 442, 929–933. Toni, N., Teng, E.M., Bushong, E.A., Aimone, J.B., Zhao, C., Consiglio, A., van Praag, H., Martone, M.E., Ellisman, M.H., Gage, F.H., 2007. Synapse formation on neurons born in the adult hippocampus. Nat. Neurosci. 10, 727–734. Toni, N., Laplagne, D.A., Zhao, C., Lombardi, G., Ribak, C.E., Gage, F.H., Schinder, A.F., 2008. Neurons born in the adult dentate gyrus form functional synapses with target cells. Nat. Neurosci. 11, 901–907. van Praag, H., Schinder, A.F., Christie, B.R., Toni, N., Palmer, T.D., Gage, F.H., 2002. Functional neurogenesis in the adult hippocampus. Nature 415, 1030–1034.

Varju, P., Schlett, K., Eisel, U., Madarasz, E., 2001. Schedule of NMDA receptor subunit expression and functional channel formation in the course of in vitro-induced neurogenesis. J. Neurochem. 77, 1444–1456. Waites, C.L., Specht, C.G., Hartel, K., Leal-Ortiz, S., Genoux, D., Li, D., Drisdel, R.C., Jeyifous, O., Cheyne, J.E., Green, W.N., et al., 2009. Synaptic SAP97 isoforms regulate AMPA receptor dynamics and access to presynaptic glutamate. J. Neurosci. 29, 4332–4345. Wyszynski, M., Kharazia, V., Shanghvi, R., Rao, A., Beggs, A.H., Craig, A.M., Weinberg, R., Sheng, M., 1998. Differential regional expression and ultrastructural localization of alpha-actinin-2, a putative NMDA receptor-anchoring protein, in rat brain. J. Neurosci. 18, 1383–1392. Zhai, R., Olias, G., Chung, W.J., Lester, R.A., tom Dieck, S., Langnaese, K., Kreutz, M.R., Kindler, S., Gundelfinger, E.D., Garner, C.C., 2000. Temporal appearance of the presynaptic cytomatrix protein bassoon during synaptogenesis. Mol. Cell. Neurosci. 15, 417–428. Zhao, C., Teng, E.M., Summers Jr., R.G., Ming, G.L., Gage, F.H., 2006. Distinct morphological stages of dentate granule neuron maturation in the adult mouse hippocampus. J. Neurosci. 26, 3–11. Ziv, N.E., Garner, C.C., 2004. Cellular and molecular mechanisms of presynaptic assembly. Nat. Rev. Neurosci. 5, 385–399.