Impaired synaptic scaling in mouse hippocampal neurones expressing NMDA receptors with reduced calcium permeability
1 Max-Planck-Institute for Medical Research, D-69120 Heidelberg, Germany
Abstract
NMDA receptors (NMDARs) play a crucial role for the acquisition of functional AMPARs during Hebbian synaptic plasticity at cortical and hippocampal synapses over a short timescale of seconds to minutes. In contrast, homeostatic synaptic plasticity can occur over longer timescales of hours to days. The induction mechanisms of this activity-dependent synaptic scaling are poorly understood but are assumed to be independent of NMDAR signalling in the cortex. Here we investigated in the hippocampus a potential role of NMDAR-mediated Ca2+ influx for synaptic scaling of AMPA currents by genetic means. The Ca2+ permeability of NMDARs was reduced by selective postnatal expression in principal neurones of mouse forebrain half of the NR1 subunits with an amino acid substitution at the critical channel site (N598R). This genetic manipulation did not reduce the total charge transfer via NMDARs in nucleated patches (somatic) and at synaptic sites. In contrast, the current amplitude and the charge carried through AMPARs were substantially reduced at somatic and synaptic sites in juvenile and adult mutants, indicating persistent downscaling of AMPA responses. Smaller and less frequent AMPA miniature currents in the mutant demonstrated a postsynaptic locus of this down-regulation. Afferent innervation and release probability were unchanged at CA3-to-CA1 synapses of mutants, as judged from input-output and minimal stimulation experiments. Our results indicate that NMDAR-mediated Ca2+ signalling is important for synaptic scaling of AMPA currents in the hippocampus in vivo.
, 百拇医药
Introduction
After early postnatal development, NMDA receptors (NMDARs) and AMPA receptors (AMPARs) co-localize at most excitatory glutamatergic synapses and can be up- or down-regulated over short timescales of seconds to minutes by input-specific synaptic plasticity (LTP/LTD; Malenka & Nicoll, 1999; Malinow & Malenka, 2002) or globally by neuronal activity over timescales which can be as long as hours to days (homeostatic synaptic plasticity; Burrone & Murthy, 2003; Turrigiano & Nelson, 2004).
, 百拇医药
In the cortex, NMDA and AMPA currents are scaled proportionally, even across synapses of the same neurone, to maintain a constant AMPA/NMDA current ratio (Umemiya et al. 1999; Watt et al. 2000). The induction mechanisms of this homeostatic synaptic plasticity are unknown but appear to involve postsynaptic depolarization (Turrigiano & Nelson, 2004). At cortical synapses, the induction is independent of NMDAR activation and mainly affects the quantal amplitude (Turrigiano et al. 1998; Leslie et al. 2001). In contrast in the hippocampus, the induction mechanisms could depend on NMDARs (Lissin et al. 1998; Liao et al. 1999) and are known to lead only to a modest increase in the quantal amplitude, but a large increase in miniature current frequency (Burrone et al. 2002; Thiagarajan et al. 2002).
, 百拇医药
In the present study, we used mutant mice to investigate the consequences of NMDARs with reduced Ca2+ permeability for synaptic scaling in the hippocampus. To this end, we activated by appropriate transgenic Cre recombinase expression using transgenic Camkcre4 mice (TgCre4; Mantamadiotis et al. 2002) an expression-silenced NR1(N598R) allele in a previously described gene-targeted mouse (NR1+/Rneo; Single et al. 2000), such that forebrain principal neurones express the wild-type form of NR1 alongside the NR1(R) form postnatally (NR1+/Rneo/TgCre4 mice). NR1(R) when co-expressed in heterologous systems with NR2A is known to yield NMDAR channels lacking Ca2+ permeability and Mg2+ block (Burnashev et al. 1992). In our genetic model, however, the presence of both wild-type and mutant NR1 subunits should substantially reduce but not abolish the NMDAR-mediated Ca2+ influx. Indeed, if two NR1 subunits are required for channel formation, and given functional dominance of the R-form over the N-form, the postsynaptic NMDAR-dependent Ca2+ influx is predicted to be reduced by 75% (Single et al. 2000), as it should occur only through the 25% channels that carry two wild-type NR1 subunits. Furthermore, NMDA currents will be larger under physiological conditions in NR1+/Rneo/TgCre4 compared with wild-type mice due to the reduced Mg2+ block of NMDARs containing NR1(R) (Burnashev et al. 1992).
, 百拇医药
Here, we show that in juvenile and adult NR1+/Rneo/TgCre4 mice amplitude and charge of AMPA responses were smaller in hippocampal CA1 neurones, in nucleated patches (somatic) and at synaptic sites than in wild-type. The characterization of the input–output relationship for NMDA currents as well as CV analyses and minimal stimulation experiments indicated normal presynaptic function at hippocampal CA3-to-CA1 synapses of NR1+/Rneo/TgCre4 mice. Instead, a postsynaptic change is indicated by the smaller and less frequent AMPA miniature currents in the mutant. Thus, NMDARs with reduced Ca2+ permeability scaled down postsynaptic AMPA currents in the hippocampus in vivo.
, http://www.100md.com
Methods
HEK293 cells
HEK293 cells were cotransfected with plasmids encoding NR1-1a or NR1-1a(N598R), NR2A and green fluorescent protein (GFP) (Chen & Okayama, 1987). Forty eight hours after transfection, cells were continuously perfused with (mM): 135 NaCl, 5.4 KCl, 1.8 CaCl2, 5 Hepes (pH 7.3). GFP-labelled cells were lifted from the coverslip and whole-cell currents were activated at –60 mV in the presence of 10 μM glycine by fast applying 1 mM glutamate from a Piezo-driven double-barrelled pipette using an EPC-9 amplifier (HEKA, Lambrecht, Germany). Patch electrodes (3–5 M) were filled with (mM): 140 CsCl, 2 MgATP, 10 EGTA, 10 Hepes (pH 7.3, 290–305 mosmol l–1).
, http://www.100md.com
NR1+/Rneo/TgCre4 mice
Heterozygous NR1+/Rneo mice (Single et al. 2000) were bred with transgenic Camkcre4 (TgCre4) mice (Mantamadiotis et al. 2002) to generate NR1+/Rneo/TgCre4 mice in NMRI background to obtain bigger litters compared with C57Bl/6. NR1+/Rneo/TgCre4 mice carried one wild-type and one mutated NR1 allele and showed forebrain-restricted NR1(R) expression. Cre expression in NR1+/Rneo/TgCre4 mice was investigated in fixed hippocampal sections by immunohistochemistry and revealed significant mosaic expression in the CA1 pyramidal cell layer at P3 but not at P12 (not shown), indicating that NR1(R) expression does not start in a synchronous manner in NR1+/Rneo/TgCre4 mice. However, at the time points of our analysis (P14 and P42), we determined the presence or absence of NR1(R) in every whole-cell recording (except during minimal stimulation; see below). NR1(R) expression was indicated by significant NMDA currents at negative holding potentials under physiological conditions (1 mM Mg2+; 2 mM Ca2+) due to the impaired Mg2+ block (Fig. 1), or NR1(R) expression was indicated by the slower kinetics of the NMDA currents in the absence of Mg2+ (Fig. 2). Based on these characterizations, most CA1 cells at P14 (80–90%) and all CA1 cells at P42 expressed NR1(R).
, 百拇医药
A, in 2-week-old wild-type and NR1+/Rneo/TgCre4 mice, somatic NMDARs were activated in CA1 nucleated patches (n.p.) by 1 mM glutamate and synaptic NMDARs by Schaffer collateral stimulation in 1 mM Mg2+ ( and , respectively) (somatic, n = 6–8; synaptic, both n = 13). Currents were normalized to peak responses at +40 mV and show mean ± S.E.M.B, the majority of mutant neurones investigated in A (80–90%; ) expressed NR1(R), as indicated from the significantly increased NMDA current ratios recorded at +40 mV versus –80 mV in 1 mM Mg2+. *P < 0.02 or **P < 0.001. Bars indicate the mean value for each genotype.
, 百拇医药
A, 1 mM glutamate was applied in the presence of 10 μM glycine as indicated by bars to activate NMDARs in nucleated patches (n.p.) at –60 mV in 0 mM Mg2+ (left panel). The time of synaptic stimulation is indicated by the arrow (right panel). The currents displayed smaller amplitudes and slower kinetics in NR1+/Rneo/TgCre4 () compared with wild-type mice (). To compare charge, see also Table 1. B, wild-type NR1/NR2A () and mutated NR1(R)/NR2A () receptors were activated in HEK293 cells at –60 mV in 0 mM Mg2+ by 1 mM glutamate in the presence of glycine for 600 ms to determine desensitization (left panel) and for 20 ms to determine activation and deactivation kinetics (right panel).
, 百拇医药
Brain slices and patch pipettes
All experimental procedures were in accordance with the animal welfare guidelines of the Max-Planck-Society. Deeply anesthetized (halothane) mice were decapitated and the brain was quickly removed. Transverse hippocampal slices (250 μm) were prepared and incubated for 30 min at 37°C in artificial CSF (ACSF) containing (mM): 125 NaCl, 25 NaHCO3, 2.5 KCl, 1.25 NaH2PO4, 1 MgCl2, 25 D-glucose, 2 CaCl2; bubbled with 95% O2–5% CO2 (pH 7.4). Patch pipettes were pulled from borosilicate glass capillaries and had resistances of 3–5 M when filled with (mM): 125 caesium gluconate, 20 CsCl, 10 NaCl, 10 Hepes, 0.2 EGTA, 4 MgATP, 0.3 Na3GTP (pH 7.3, 290–305 mosmol l–1). Series resistances (16–30 M) and input resistances (100–300 M) were continuously monitored by measuring the peak and steady-state currents in response to hyperpolarizing pulses (–5 mV; 20 ms).
, http://www.100md.com
Somatic and synaptic currents
Somatic currents were activated in excised CA1 whole-soma patches (i.e. nucleated patches) by fast glutamate application at –60 mV and synaptic currents were activated at –70 mV by stimulating the Schaffer collaterals 150 μm distant from the CA1 cell body with a glass electrode filled with 1 M NaCl. In nucleated patches, AMPA currents were recorded in the presence of 50 μM D-AP5 and 1 mM MgCl2, and NMDA currents were recorded in the presence of 5 μM NBQX, 10 μM glycine and 0 or 1 mM MgCl2. To record NMDA EPSCs, 10 μM bicuculline methiodide (BMI), 5 μM NBQX, 10 μM glycine and 0 or 1 mM MgCl2 were added to the ACSF (see above); to record AMPA EPSCs, 10 μM BMI and 50 μM D-AP5 were added. The AMPA EPSC coefficient of variation (CV; mean ± S.D.) was determined from AMPA EPSCs which were evoked in the absence of D-AP5 but in the presence of 4 mM MgCl2 and 4 mM CaCl2. Spontaneous AMPA mEPSCs were recorded in ACSF containing 4 mM CaCl2, in the presence of 50 μM D-AP5, 10 μM BMI and 1 μM TTX. All recordings were performed at room temperature.
, 百拇医药
Minimal stimulation
Minimal stimulation was used to putatively stimulate one axonal fibre making one synapse on a CA1 cell (Raastad et al. 1992; Allen & Stevens, 1994). The presence of NR1(R)-containing NMDARs was demonstrated in a fraction of CA1 cells from wild-type and NR1+/Rneo/TgCre4 mice (P14, Fig. 5A). Subsequently, AMPA EPSCs were evoked by Schaffer collateral stimulation at 0.25 Hz in ACSF containing 10 μM BMI and 50 μM D-AP5. Stimulus intensity was carefully adjusted so that a significant portion of failures mixed with evoked minimal AMPA EPSCs. Only low noise recordings (S.D. of noise about 1.5 pA) with stable access resistance during the experiment and large sample numbers (n = 100–200) were analysed. Further criteria for inclusion in analysis were invariance of latency and shape of the individual minimal EPSCs for repeated stimuli and constancy of EPSC amplitude and percentage of failures during data collection. For each recording, all failures and successes were averaged to show successful discrimination of failures from successes (Fig. 5B). To quantify the size of minimal EPSCs, a 6 ms time window at the EPSC peak, typically 3–5 ms after the stimulus artifact, was selected and the charge of the minimal EPSCs was calculated. Transmission failures were detected by visual discriminination. Rarely, spontaneous events or minimal EPSCs with different onset latencies occurred, which were counted as failures. Overlap between charges of thereby visually determined failures and successes are due to such single spontaneous events (Fig. 5C). Release probability at putative single synapses was estimated by calculating the ratio between successes and total number of trials. This led to heterogeneous release probabilites across different putative single synapses which is in accordance with previous work (Dobrunz & Stevens, 1997).
, 百拇医药
A, scaled mean of EPSCs in 1 mM Mg2+/2 mM Ca2+ at –70 mV holding potential before (black) and during D-AP5 (grey) perfusion, indicating presence of NR1(R)-containing NMDARs in NR1+/Rneo/TgCre4 cells that were subjected to minimal stimulation. Failures were excluded. B, representative responses to minimal stimulation at 0.25 Hz in presence of D-AP5 as in A. Failures (F) and successes (S) recorded from a wild-type and NR1+/Rneo/TgCre4 neurone (release probabilities, 0.445 versus 0.400). The boxes mark time window selected for quantification of EPSC charge. Bottom traces are averages of all detected failures (grey) and successes (black) following 200 stimuli for each cell. C, EPSC charge plotted for failures and successes (same neurones as in B). Overlaps between failures and successes are caused by spontaneous events or events with onset latencies different from B, which were classified as failures. D, release probabilities estimated for individual recordings from wild-type (n = 6) and mutant (n = 5) neurones following 100–200 stimuli. Bars indicate the mean release probability for each genotype.
, 百拇医药
Data analysis
Spontaneous recordings were analysed off-line (filtering, 2.9 kHz; sampling, 5 kHz) using an event detection program (kindly provided by Professor Misgeld, Heidelberg). Threshold for event detection was –5 pA and detected events were inspected manually as described (Steigerwald et al. 2000). Additional off-line data analysis, curve fitting and figure preparation was performed with PulseFit (HEKA, Lambrecht, Germany) and Igor Pro (WaveMetrics, Lake Oswego, OR, USA). Deactivation kinetics of averaged NMDA currents were described using a weighted decay time constant: tw = (If/(If + Is)) tf + (Is/(Is + If)) ts, where If and Is are the amplitudes of the fast and slow decay components, and tf and ts are their respective decay time constants. For analysis and illustration, 3–10 single traces were averaged. Data are presented as mean ± S.E.M. Statistical significance was evaluated by two-tailed, unpaired Student's t test or Pearson's test.
, http://www.100md.com
Results
Functional consequences of NR1(N598R) for NMDA currents
NMDARs were activated in CA1 neurones of acute hippocampal slices from 2-week-old mice, either in nucleated patches by 1 mM glutamate or by Schaffer collateral stimulation. I–V relationships from NR1+/Rneo/TgCre4 neurones compared with wild-type showed reduced negative slope conductance under physiological ionic conditions (Fig. 1A), as predicted from in vitro experiments (Burnashev et al. 1992). On average, the ratio of NMDA currents evoked at +40 and –80 mV was about 6-fold higher in nucleated patches of NR1+/Rneo/TgCre4 compared with wild-type neurones (–5.5 ± 1.3, n = 8 versus –32.3 ± 5.9, n = 6; P = 0.02) and about 2.5-fold higher at synaptic sites (–3.0 ± 0.6, n = 13 versus –7.4 ± 1.0, n = 12; P = 0.001). When comparing the +40 mV/–80 mV ratios between individual neurones of NR1+/Rneo/TgCre4 mice with wild-type (Fig. 1B), the ratios of 7 out of 8 nucleated patch recordings and the ratios of 10 out of 13 synaptic recordings were higher than all wild-type ratios recorded at both sites, indicating that NR1(R) was present in 80–90% of the investigated neurones at P14 (for P42 see below).
, 百拇医药
The reduced Ca2+ permeability of the NMDAR population in NR1+/Rneo/TgCre4 mice by about 75% (Single et al. 2000) was indicated by a shift in the reversal potentials of NMDA currents recorded in 1.8 mM [Ca2+] (3 mV at somatic sites and 6.2 mV at synaptic sites (Fig. 1A); see also Burnashev et al. 1992 and Single et al. 2000 for major reversal potential shifts when recorded in high Ca2+ solution).
NMDA current amplitudes in NR1+/Rneo/TgCre4 compared with wild-type mice were smaller at somatic and synaptic sites when activated in Mg2+-free solution at negative potentials (Fig. 2A and Table 1). The smaller NMDA currents in NR1+/Rneo/TgCre4 mice were unlikely to be caused by one expression-silenced wild-type NR1 allele, since heterozygous NR1 knockout mice display normal NMDA currents (Forrest et al. 1994). Instead, NMDA current reduction may well be caused by reduced single channel conductance imposed by NR1(R) (Behe et al. 1995). Consistently, glutamate-activated currents mediated by pure NR1(R)/NR2A receptors in HEK293 cells were also smaller than when mediated by wild-type NR1/NR2A receptors (Fig. 2B and Table 1). In addition, we observed that NMDA currents displayed slower kinetics in NR1+/Rneo/TgCre4 mice (Fig. 2A and Table 1). Importantly, as a consequence of the slower kinetics the total charge transfer via NMDAR channels was not significantly different between the genotypes at somatic and synaptic sites (Table 1).
, 百拇医药
The slower kinetics of NMDA currents in HEK293 cells expressing NR1(R)/NR2A receptors and in CA1 neurones of NR1+/Rneo/TgCre4 mice could derive from changes in agonist binding and/or channel gating. We recorded dose–response curves of wild-type and NR1(R)-containing NMDARs in HEK293 cells (not shown) and found no change in the potency for glycine (EC50s: NR1/NR2A, 2.11 μM; NR1(R)/NR2A, 2.03 μM, but a minor increase in glutamate potency (EC50s: NR1/NR2A, 5.51 μM; NR1(R)/NR2A, 1.94 μM). In spite of the minor increase in glutamate potency, the rise time of glutamate-activated currents was slower in the presence of NR1(R), indicating that NR1(R) probably did not increase the glutamate binding rate. A concurrently performed study on recombinant NR1(R)/NR2A receptors observed very similar changes on kinetics and glutamate potency (Chen et al. 2004). In this study, a kinetic model explained the leftward shift in the glutamate dose–response curve by a about 10-fold reduction in the channel closing rate, even without any modification in ligand binding and unbinding rates. Thus, the slower glutamate-activated currents we observed in neurones expressing NR1(R) are very likely to be independent of any ligand binding modification. Furthermore in our NR1+/Rneo/TgCre4 mice, we excluded the possibility that a change in NR2A expression relative to NR2B is responsible for the altered current kinetics, since NMDA currents in nucleated patches from CA1 neurones of both genotypes showed comparable sensitivity to the NR2B-specific antagonist ifenprodil (3 μM; not shown).
, http://www.100md.com
In summary, the presence of NR1(R) in hippocampal CA1 neurones was responsible for the reduced voltage-dependent Mg2+ block and the impaired Ca2+ permeability of the NMDAR population. However, the charge carried by NMDARs was not different between both genotypes at somatic and synaptic sites.
AMPA currents are reduced in CA1 neurones of NR1+/Rneo/TgCre4 mice
Next, we investigated whether NMDARs with impaired Ca2+ permeability would affect acquisition of functional AMPARs in the hippocampus of 2-week-old NR1+/Rneo/TgCre4 mice. AMPA current amplitudes were reduced by about 30% in nucleated patches and at synaptic sites in CA1 pyramidal cells of NR1+/Rneo/TgCre4 mice (Fig. 3A and Table 1). Subsequent to the recording of AMPA currents, NMDA currents were recorded for each neurone in the absence of extracellular Mg2+. Decay kinetics of NMDA currents were used to indicate the presence of NR1(R) (see above) and related to the amplitude of AMPA currents (Fig. 3B). In recordings from wild-type neurones, deactivation of NMDA currents did not correlate with the size of AMPA currents. In contrast, in synaptic recordings from NR1+/Rneo/TgCre4 neurones, NMDA decay and AMPA peak currents correlated, raising the possibility that the smaller AMPA currents at synaptic sites are a consequence of NR1(R) expression. In somatic recordings, the correlation coefficient was only close to significance in NR1+/Rneo/TgCre4 neurones (for explanation see below). The charge carried through AMPAR channels was also reduced in NR1+/Rneo/TgCre4 mice, since the kinetics of the AMPA currents remained unchanged. In these experiments, the position of the stimulation electrode relative to the CA1 cell was kept constant. However, post hoc comparison of stimulation intensities between cells revealed that higher intensities were used to activate synapses in slices of NR1+/Rneo/TgCre4 compared with wild-type mice (by 33.3%; P = 0.007; Table 1). Thus, the AMPA current reduction at synaptic sites in NR1+/Rneo/TgCre4 mice is actually even higher than 30%. This became evident when relating AMPA current amplitudes to identical stimulation intensities, which were increased in steps of 5 V, leading to 2- to 3-fold reduced AMPA currents (Fig. 4A). The fact, that AMPA currents in NR1+/Rneo/TgCre4 neurones were more drastically reduced at synaptic than at somatic sites, could explain why a correlation between the presence of NR1(R) and size of AMPA currents was found for synaptic but not for somatic recordings.
, 百拇医药
A, AMPA current amplitudes in NR1+/Rneo/TgCre4 mice (open bars) were reduced by about 30% compared with wild-type (filled bars) in nucleated patches (n.p.) and at synaptic sites. Bars are mean ± S.E.M., number of cells is indicated in parentheses, and level of significance is either **P < 0.001 or *P < 0.05. A representative current trace is given next to each bar. For NMDA currents, see Table 1. B, correlation of peak AMPA currents and decay kinetics of NMDA currents. Slowly deactivating NMDA currents indicate presence of NR1(R) in NR1+/Rneo/TgCre4 mice () and significantly correlate with AMPA peak currents in synaptic recordings from NR1+/Rneo/TgCre4 cells (correlation coefficient R = 0.72, n = 9, P < 0.05). No significant correlation was found in n.p. recordings of NR1+/Rneo/TgCre4 cells (R = 0.45, n = 14, P > 0.05). Recordings from wild-type cells () showed no correlation between size of AMPA current and NMDA decay (synaptic: R = –0.21, n = 8; n.p. R = 0.18, n = 23, P > 0.05). Significance of correlation was assessed with Pearson's test.
, 百拇医药
A, synaptic AMPA currents were evoked at –70 mV and are plotted as a function of stimulation intensity. The amplitudes were 2- to 3-fold reduced in NR1+/Rneo/TgCre4 (; n = 12) compared with wild-type mice (; n = 7). *P < 0.05. B, synaptic NMDA currents were evoked at +40 mV in solutions containing 1 mM Mg2+ extracellularly and 2 mM Mg2+ intracellularly. Current amplitudes were not significantly different between NR1+/Rneo/TgCre4 (; n = 5) and wild-type mice (; n = 5). C, paired-pulse facilitation (PPF) was determined at 50 ms intervals and was unchanged in NR1+/Rneo/TgCre4 mice. AMPA and NMDA EPSCs were evoked in 1 mM Mg2+ at –70 and –30 mV, respectively. Bars are mean ± S.E.M.
, 百拇医药
To provide additional evidence that the impaired AMPAR function resulted from expression of NR1(R), we also recorded AMPA currents in littermates of NR1+/Rneo/TgCre4 mice which lack the Cre transgene (NR1+/Rneo mice, Single et al. 2000). As a result, NR1(R) expression is not turned on by Cre-mediated recombination. AMPA current amplitudes were comparable between wild-type and NR1+/Rneo mice, in nucleated patches (WT, –2520 ± 95 pA, n = 27; NR1+/Rneo, –2442 ± 106 pA, n = 19; P = 0.59) and at synaptic sites in CA1 pyramidal cells (WT, –77.4 ± 8.1 pA, n = 15; NR1+/Rneo, –99.1 ± 13.3 pA, n = 7; P = 0.19). The stimulation intensities to evoke AMPA EPSCs in NR1+/Rneo mice were similar to those used in wild-type recordings (P = 0.17), strengthening the conclusion that AMPA currents were not downscaled in the absence of Cre.
, http://www.100md.com
In NR1+/Rneo/TgCre4 mice, the AMPA current reduction was also found in adult animals. At P42, synaptic AMPA current amplitudes were –70 ± 6 pA (n = 6) in wild-type and –42 ± 5 pA (n = 9) in mutant neurones (P = 0.006), and the stimulation intensities were 88% higher in NR1+/Rneo/TgCre4 than in wild-type recordings (11 ± 2 V versus 20 ± 1 V; P = 0.04). At this age, every neurone expressed NR1(R), as indicated by prominent and slow NMDA components of composite EPSCs (data not shown).
, http://www.100md.com
To investigate whether afferent innervation was still intact in NR1+/Rneo/TgCre4 mice, synaptic NMDA currents were evoked using increasing stimulation intensities at +40 mV in 2-week-old mice (Fig. 4B). At this positive potential, the 2 mM Mg2+ present in the intracellular solution blocked wild-type NMDARs to a greater extent than NR1(R)-containing NMDARs in NR1+/Rneo/TgCre4 mice. This is because NR1(R) also reduces the intracellular Mg2+ block (Wollmuth et al. 1998). As a consequence, NMDA current amplitudes were indistinguishable at +40 mV between genotypes (160 ± 21 pA; n = 13 versus 177 ± 29 pA; n = 13). When stimulation intensities were increased, the amplitudes of NMDA currents similarly increased in both genotypes and were not significantly different at any tested stimulation intensity, indicating normal afferent innervation of CA1 neurones in NR1+/Rneo/TgCre4 mice.
, http://www.100md.com
To investigate presynaptic function at CA3-to-CA1 synapses of NR1+/Rneo/TgCre4 mice, we first determined paired-pulse facilitation (PPF) for AMPA and NMDA currents delivering two stimuli to the Schaffer collaterals at intervals of 50 ms (Fig. 4C). The PPF ratio of AMPA and NMDA currents was comparable between wild-type and NR1+/Rneo/TgCre4 mice. Next, we estimated the AMPA EPSC coefficient of variation (CV; S.D. mean–1), which could be affected by changes in release probability. When calculating the CV from 60 EPSCs, which had peak amplitudes between 50 and 100 pA and were evoked at –70 mV in each neurone, the CV was found to be comparable (wild-type, 0.32 ± 0.03; n = 15 and NR1+/Rneo/TgCre4 mice, 0.29 ± 0.02; n = 22). Lastly, we performed minimal stimulation experiments to compare the release probability between genotypes (Fig. 5). First, we evoked EPSCs in both genotypes in 1 mM Mg2+/2 mM Ca2+ at –70 mV to reassure NR1(R) expression in the CA1 neurones of NR1+/Rneo/TgCre4 mice (see prominent NMDA component in composite EPSC of mutant neurone; Fig. 5A). Following blockade of the NMDA component by D-AP5, failures or successes (minimal AMPA EPSCs) were evoked during 100–200 minimal stimulations in neurones of both genotypes (see example traces in Fig. 5B). The analysis of all individual failures and successes of these two neurones is shown in Fig. 5C. In Fig. 5D the release probabilities of the individual recordings from wild-type and mutant cells are plotted, showing that release probability at CA3-to-CA1 synapses of NR1+/Rneo/TgCre4 mice is unchanged (wild-type, 0.31 ± 0.06; n = 6 and NR1+/Rneo/TgCre4 mice, 0.29 ± 0.05; n = 5).
, 百拇医药
To test for postsynaptic changes, AMPA miniature EPSCs (AMPA mEPSCs) were recorded in tetrodotoxin (TTX) and D-AP5. Under these conditions the S.D. of the noise was comparable between wild-type and NR1+/Rneo/TgCre4 mice (1.70 ± 0.07 pA versus 1.64 ± 0.03 pA; n = 10 each; Fig. 6A) and was favourable in detecting miniature AMPA responses. Most (24 out of 34) CA1 cells in NR1+/Rneo/TgCre4 mice in contrast to 1 out of 10 cells in wild-type displayed less than 1 event per recording minute. From cells with frequencies above 1 event min–1, around 100 events were averaged. Averaged AMPA mEPSC rise and decay times were unchanged; their amplitudes, however, were decreased by 21% in NR1+/Rneo/TgCre4 mice (Fig. 6B and Table 1). Consistently, when constructing amplitude histograms which contained all analysed events (WT, n = 1594; NR1+/Rneo/TgCre4, n = 1225; Fig. 6C), both histograms could be fitted to a lognormal function and confirmed the 21% smaller mEPSC amplitudes in NR1+/Rneo/TgCre4 mice (10.9 ± 0.1 pA versus 8.6 ± 0.1 pA; Fig. 6C) found arithmetically (Table 1).
, 百拇医药
A, representative current traces of AMPA mEPSCs recorded from a wild-type (WT) and a NR1+/Rneo/TgCre4 neurone at –70 mV in presence of TTX and D-AP5. B, about 100 events per neurone were averaged (left). The bars indicate the 21% reduced amplitudes of AMPA mEPSCs in NR1+/Rneo/TgCre4 neurones (*P = 0.004), whereas the rise and decay times were unchanged relative to wild-type (= 100%; see bars). (WT, filled symbols and NR1+/Rneo/TgCre4, open symbols.) See also Table 1. C, histograms, which include all events of all measured cells (WT, n = 1594; NR1+/Rneo/TgCre4, n = 1225), were fitted by the lognormal function yo + A exp (– (ln(x/xo)/peak width)2), where yo sets the baseline, A sets the amplitude and xo sets the peak position in x. The frequencies of AMPA miniature EPSCs were determined in a different set of experiments and the reduced frequencies in NR1+/Rneo/TgCre4 mice were illustrated in a distribution plot (P = 0.003). Bars indicate the average miniature frequency for each genotype.
, 百拇医药
The reduction in mEPSC amplitudes was too small to explain the 2- to 3-fold reduced synaptic AMPA currents
(see above). Therefore, we considered the possibility of an increased number of synapses devoid of functional AMPARs in NR1+/Rneo/TgCre4 mice, since we noted fewer miniature events in NR1+/Rneo/TgCre4 compared with wild-type mice (see above). In an additional set of experiments, AMPA mEPSC frequencies were recorded from 10 cells of both genotypes and 9 min traces were analysed each. Again consistent with the observation that about 80–90% of the CA1 neurones of 2-week-old NR1+/Rneo/TgCre4 mice express NR1(R), 9 out of 10 neurones were similar with respect to very low mEPSC frequencies. On average, mEPSC frequencies were reduced to 1.2 ± 0.8 events min–1 in NR1+/Rneo/TgCre4 neurones compared with 5.4 ± 0.9 events min–1 in wild-type (P = 0.003; Fig. 6C; both n = 10). In agreement with the initial mEPSC experiments, the amplitudes of AMPA mEPSCs were decreased by 19% (P = 0.007). This amplitude reduction together with the concurrent frequency reduction of AMPA mEPSCs reduced the spontaneous AMPA currents during the 9 min recordings 4- to 5-fold in CA1 neurones of NR1+/Rneo/TgCre4 compared with wild-type mice (1215.0 pA from 112 events versus 5435.5 pA from 441 events). This 4- to 5-fold reduction is close to the 2- to 3-fold reduced amplitudes of evoked AMPA currents (Fig. 4A).
, http://www.100md.com
Collectively, afferent innervation and release probability remained unchanged at CA3-to-CA1 synapses in NR1+/Rneo/TgCre4 mice. Conversely, AMPA EPSCs were smaller and mEPSCs in addition less frequent in the mutant, indicating postsynaptic downscaling of AMPA responses.
Discussion
Juvenile and adult mice expressing NMDARs with reduced Ca2+ permeability in forebrain principal neurones (NR1+/Rneo/TgCre4) displayed smaller AMPA current amplitudes and charges at hippocampal CA3-to-CA1 synapses. Paired-pulse facilitation, input–output relationships for NMDA EPSCs and release probability examined by minimal stimulation were unchanged, implying a postsynaptic locus for this down-regulation of AMPA responses. In agreement with a postsynaptic locus, AMPA mEPSC amplitudes were slightly reduced and the frequencies of AMPA mEPSCs were severely decreased.
, http://www.100md.com
Studies in cultured neurones indicated a differential contribution of amplitude and/or frequency changes of AMPA mEPSCs to homeostatic synaptic plasticity and revealed differences between hippocampal and cortical cultures (see introduction and Turrigiano & Nelson, 2004). These differences could arise from the fact that the number of AMPAR-free synapses is higher in cultured hippocampal (Liao et al. 1999) than in cortical neurones (Pratt et al. 2003). It has been suggested that the functionalization of AMPAR-free synapses will increase the mEPSC frequency with only a modest enhancement of amplitude (Burrone & Murthy, 2003), and it has been shown that reduced AMPAR surface expression decreased the frequency but not the amplitude of AMPA mEPSCs (Noel et al. 1999). Based on these observations, the moderate decrease in AMPA mEPSC amplitude in hippocampal CA1 neurones of NR1+/Rneo/TgCre4 mice indicate synapses with reduced AMPAR content and the drastic decrease in mEPSC frequency is consistent with a lower number of AMPAR-containing synapses. The recordings of AMPA responses by stimulating the Schaffer collaterals indicated a 2- to 3-fold reduction in AMPA current. When we reduce all our mEPSC events recorded in wild-type neurones by 60%, 40% of all mEPSCs would fall below our detection limit of –5 pA. In our recordings, we found a reduction in mEPSC frequency by 80% in NR1+/Rneo/TgCre4 mice. This is a good indication that actually a higher number of AMPA-free synapses exist in NR1+/Rneo/TgCre4 mice. The equally possible mechanism, that AMPAR content has been scaled down homogeneously across all synapses in NR1+/Rneo/TgCre4 mice would also generate synapses for which AMPARs can no longer be detected, since the postsynaptic AMPAR content, different from the NMDAR content, is highly variable at CA3-to-CA1 synapses, mainly containing a very small AMPAR number (Takumi et al. 1999; Racca et al. 2000).
, http://www.100md.com
In NR1+/Rneo/TgCre4 mice, fewer AMPARs might be needed in synapses, since, due to their reduced Mg2+ block, NR1(R)-containing NMDARs are activated when glutamate is released under physiological ionic conditions. The observed unchanged release probability and normal input–output relationship for NMDA EPSCs, as opposed to the altered input–output relationship for AMPA EPSCs in CA1 neurones of NR1+/Rneo/TgCre4 mice, are consistent with this hypothesis. The time required for the AMPA current reduction following onset of NR1(R) expression cannot be determined in our current mouse model, since Cre expression in CA1 neurones does not start in a synchronous manner, and the time from Cre expression until synaptic insertion of NR1(R)-containing NMDARs in CA1 neurones is unknown. Virus-driven Cre expression could be a first step to achieve at least synchrony of Cre expression.
, 百拇医药
Several putative mechanisms for synaptic scaling of AMPA responses have been discussed, including changes in the level of postsynaptic depolarization and/or NMDAR activation and subsequently activated Ca2+-dependent mechanisms. The level of postsynaptic depolarization in response to synaptic glutamate receptor activation could be enhanced in NR1+/Rneo/TgCre4 mice in 1 mM Mg2+, since NR1(R)-containing NMDARs mediate postsynaptic depolarization close to the resting membrane potential. However, in a similar mouse mutant expressing NMDARs with strongly reduced Mg2+ block but largely unaltered Ca2+ permeability the synaptic AMPA currents remained unchanged (Single et al. 2002), excluding the possibility that AMPA currents in NR1+/Rneo/TgCre4 mice were down-regulated by enhanced NMDAR-mediated postsynaptic depolarization. Here, we show in NR1+/Rneo mice that AMPAR expression is not impaired before NR1(R) expression gets turned on and that due to the slower kinetics of NMDA currents the charge transfer via NMDAR channels in NR1+/Rneo/TgCre4 mice is unaltered. This leaves the reduced Ca2+ influx via NMDARs as the most likely explanation for the downscaling of postsynaptic AMPA responses, consistent with a Ca2+-dependent homeostatic plasticity model (Yeung et al. 2004). Smaller postsynaptic Ca2+ elevations via NMDARs could impair a shift of AMPARs from a degradative to a recycling pathway (Ehlers, 2000) and/or could impair the Ca2+/CaM sensitivity of CaMKII which may be important for activity-dependent synaptic homeostasis (Thiagarajan et al. 2002). Our finding that reduced NMDAR-mediated Ca2+ influx scaled down AMPA responses also at somatic sites (in nucleated patches) might be explained by the possibility that AMPARs are first inserted at extrasynaptic sites before moving laterally to synaptic sites (Passafaro et al. 2001).
, 百拇医药
In contrast to cortical neurones, NMDAR-mediated Ca2+ influx appears to regulate the AMPA currents in hippocampal CA1 neurones in vivo. In addition, NMDARs with reduced Ca2+ permeability may also regulate postsynaptic AMPA responses during synaptic plasticity, which will be addressed in the future.
References
Allen C & Stevens CF (1994). An evaluation of causes for unreliability of synaptic transmission. Proc Natl Acad Sci U S A 91, 10380–10383.
, 百拇医药
Behe P, Stern P, Wyllie DJ, Nassar M, Schoepfer R & Colquhoun D (1995). Determination of NMDA NR1 subunit copy number in recombinant NMDA receptors. Proc R Soc Lond B Biol Sci 262, 205–213.
Burnashev N, Schoepfer R, Monyer H, Ruppersberg JP, Gunther W, Seeburg PH & Sakmann B (1992). Control by asparagine residues of calcium permeability and magnesium blockade in the NMDA receptor. Science 257, 1415–1419.
Burrone J & Murthy VN (2003). Synaptic gain control and homeostasis. Curr Opin Neurobiol 13, 560–567.
, 百拇医药
Burrone J, O'Byrne M & Murthy VN (2002). Multiple forms of synaptic plasticity triggered by selective suppression of activity in individual neurons. Nature 420, 414–418.
Chen N, Li B, Murphy TH & Raymond LA (2004). Site within N-methyl-D-aspartate receptor pore modulates channel gating. Mol Pharmacol 65, 157–164.
Chen C & Okayama H (1987). High-efficiency transformation of mammalian cells by plasmid DNA. Mol Cell Biol 7, 2745–2752.
, 百拇医药
Dobrunz LE & Stevens CF (1997). Heterogeneity of release probability, facilitation, and depletion at central synapses. Neuron 18, 995–1008.
Ehlers MD (2000). Reinsertion or degradation of AMPA receptors determined by activity-dependent endocytic sorting. Neuron 28, 511–525.
Forrest D, Yuzaki M, Soares HD, Ng L, Luk DC, Sheng M, Stewart CL, Morgan JI, Connor JA & Curran T (1994). Targeted disruption of NMDA receptor 1 gene abolishes NMDA response and results in neonatal death. Neuron 13, 325–338.
, 百拇医药
Leslie KR, Nelson SB & Turrigiano GG (2001). Postsynaptic depolarization scales quantal amplitude in cortical pyramidal neurons. J Neurosci 21, RC170.
Liao D, Zhang X, O'Brien R, Ehlers MD & Huganir RL (1999). Regulation of morphological postsynaptic silent synapses in developing hippocampal neurons. Nat Neurosci 2, 37–43.
Lissin DV, Gomperts SN, Carroll RC, Christine CW, Kalman D, Kitamura M, Hardy S, Nicoll RA, Malenka RC & Von Zastrow M (1998). Activity differentially regulates the surface expression of synaptic AMPA and NMDA glutamate receptors. Proc Natl Acad Sci U S A 95, 7097–7102.
, 百拇医药
Malenka RC & Nicoll RA (1999). Long-term potentiation – a decade of progress Science 285, 1870–1874.
Malinow R & Malenka RC (2002). AMPA receptor trafficking and synaptic plasticity. Annu Rev Neurosci 25, 103–126.
Mantamadiotis T, Lemberger T, Bleckmann SC, Kern H, Kretz O, Martin Villalba A, Tronche F, Kellendonk C, Gau D, Kapfhammer J, Otto C, Schmid W & Schutz G (2002). Disruption of CREB function in brain leads to neurodegeneration. Nat Genet 31, 47–54.
, 百拇医药
Noel J, Ralph GS, Pickard L, Williams J, Molnar E, Uney JB, Collingridge GL & Henley JM (1999). Surface expression of AMPA receptors in hippocampal neurons is regulated by an NSF-dependent mechanism. Neuron 23, 365–376.
Passafaro M, Piech V & Sheng M (2001). Subunit-specific temporal and spatial patterns of AMPA receptor exocytosis in hippocampal neurons. Nat Neurosci 4, 917–926.
Pratt KG, Watt AJ, Griffith LC, Nelson SB & Turrigiano GG (2003). Activity-dependent remodeling of presynaptic inputs by postsynaptic expression of activated CaMKII. Neuron 39, 269–281.
, http://www.100md.com
Raastad M, Storm JF & Andersen P (1992). Putative single quantum and single fibre excitatory postsynaptic currents show similar amplitude range and variability in rat hippocampal slices. Eur J Neurosci 4, 113–117.
Racca C, Stephenson FA, Streit P, Roberts JD & Somogyi P (2000). NMDA receptor content of synapses in stratum radiatum of the hippocampal CA1 area. J Neurosci 20, 2512–2522.
Single FN, Rozov A, Burnashev N, Zimmermann F, Hanley DF, Forrest D, Curran T, Jensen V, Hvalby O, Sprengel R & Seeburg PH (2000). Dysfunctions in mice by NMDA receptor point mutations NR1 (N598Q) and NR1 (N598R). J Neurosci 20, 2558–2566.
, 百拇医药
Single FN, Rozov A, Burnashsev N, Wollmuth LP, Kuner T, Jensen V, Hvalby O, Sprengel R & Seeburg PH (2002). Gene targeted mice with NMDA receptor subunit NR2A point mutation N596S show elevated levels of LTP. Program No. 150.6 2002 Abstract Viewer and Itinerary Planner. Society for Neuroscience Abstracts. Online.
Steigerwald F, Schulz TW, Schenker LT, Kennedy MB, Seeburg PH & Khr G (2000). C-Terminal truncation of NR2A subunits impairs synaptic but not extrasynaptic localization of NMDA receptors. J Neurosci 20, 4573–4581.
, http://www.100md.com
Takumi Y, Ramirez-Leon V, Laake P, Rinvik E & Ottersen OP (1999). Different modes of expression of AMPA and NMDA receptors in hippocampal synapses. Nat Neurosci 2, 618–624.
Thiagarajan TC, Piedras-Renteria ES & Tsien RW (2002). alpha- and betaCaMKII. Inverse regulation by neuronal activity and opposing effects on synaptic strength. Neuron 36, 1103–1114.
Turrigiano GG, Leslie KR, Desai NS, Rutherford LC & Nelson SB (1998). Activity-dependent scaling of quantal amplitude in neocortical neurons. Nature 391, 892–896.
, 百拇医药
Turrigiano GG & Nelson SB (2004). Homeostatic plasticity in the developing nervous system. Nat Rev Neurosci 5, 97–107.
Umemiya M, Senda M & Murphy TH (1999). Behaviour of NMDA and AMPA receptor-mediated miniature EPSCs at rat cortical neuron synapses identified by calcium imaging. J Physiol 521, 113–122.
Watt AJ, Van Rossum MC, MacLeod KM, Nelson SB & Turrigiano GG (2000). Activity coregulates quantal AMPA and NMDA currents at neocortical synapses. Neuron 26, 659–670.
, 百拇医药
Wollmuth LP, Kuner T & Sakmann B (1998). Intracellular Mg2+ interacts with structural determinants of the narrow constriction contributed by the NR1-subunit in the NMDA receptor channel. J Physiol 506, 33–52.
Yeung LC, Shouval HZ, Blais BS & Cooper LN (2004). Synaptic homeostasis and input selectivity follow from a calcium dependent plasticity model. Proc Natl Acad Sci U S A 101, 14943–14948., 百拇医药(Verena Pawlak, Bettina J )
Abstract
NMDA receptors (NMDARs) play a crucial role for the acquisition of functional AMPARs during Hebbian synaptic plasticity at cortical and hippocampal synapses over a short timescale of seconds to minutes. In contrast, homeostatic synaptic plasticity can occur over longer timescales of hours to days. The induction mechanisms of this activity-dependent synaptic scaling are poorly understood but are assumed to be independent of NMDAR signalling in the cortex. Here we investigated in the hippocampus a potential role of NMDAR-mediated Ca2+ influx for synaptic scaling of AMPA currents by genetic means. The Ca2+ permeability of NMDARs was reduced by selective postnatal expression in principal neurones of mouse forebrain half of the NR1 subunits with an amino acid substitution at the critical channel site (N598R). This genetic manipulation did not reduce the total charge transfer via NMDARs in nucleated patches (somatic) and at synaptic sites. In contrast, the current amplitude and the charge carried through AMPARs were substantially reduced at somatic and synaptic sites in juvenile and adult mutants, indicating persistent downscaling of AMPA responses. Smaller and less frequent AMPA miniature currents in the mutant demonstrated a postsynaptic locus of this down-regulation. Afferent innervation and release probability were unchanged at CA3-to-CA1 synapses of mutants, as judged from input-output and minimal stimulation experiments. Our results indicate that NMDAR-mediated Ca2+ signalling is important for synaptic scaling of AMPA currents in the hippocampus in vivo.
, 百拇医药
Introduction
After early postnatal development, NMDA receptors (NMDARs) and AMPA receptors (AMPARs) co-localize at most excitatory glutamatergic synapses and can be up- or down-regulated over short timescales of seconds to minutes by input-specific synaptic plasticity (LTP/LTD; Malenka & Nicoll, 1999; Malinow & Malenka, 2002) or globally by neuronal activity over timescales which can be as long as hours to days (homeostatic synaptic plasticity; Burrone & Murthy, 2003; Turrigiano & Nelson, 2004).
, 百拇医药
In the cortex, NMDA and AMPA currents are scaled proportionally, even across synapses of the same neurone, to maintain a constant AMPA/NMDA current ratio (Umemiya et al. 1999; Watt et al. 2000). The induction mechanisms of this homeostatic synaptic plasticity are unknown but appear to involve postsynaptic depolarization (Turrigiano & Nelson, 2004). At cortical synapses, the induction is independent of NMDAR activation and mainly affects the quantal amplitude (Turrigiano et al. 1998; Leslie et al. 2001). In contrast in the hippocampus, the induction mechanisms could depend on NMDARs (Lissin et al. 1998; Liao et al. 1999) and are known to lead only to a modest increase in the quantal amplitude, but a large increase in miniature current frequency (Burrone et al. 2002; Thiagarajan et al. 2002).
, 百拇医药
In the present study, we used mutant mice to investigate the consequences of NMDARs with reduced Ca2+ permeability for synaptic scaling in the hippocampus. To this end, we activated by appropriate transgenic Cre recombinase expression using transgenic Camkcre4 mice (TgCre4; Mantamadiotis et al. 2002) an expression-silenced NR1(N598R) allele in a previously described gene-targeted mouse (NR1+/Rneo; Single et al. 2000), such that forebrain principal neurones express the wild-type form of NR1 alongside the NR1(R) form postnatally (NR1+/Rneo/TgCre4 mice). NR1(R) when co-expressed in heterologous systems with NR2A is known to yield NMDAR channels lacking Ca2+ permeability and Mg2+ block (Burnashev et al. 1992). In our genetic model, however, the presence of both wild-type and mutant NR1 subunits should substantially reduce but not abolish the NMDAR-mediated Ca2+ influx. Indeed, if two NR1 subunits are required for channel formation, and given functional dominance of the R-form over the N-form, the postsynaptic NMDAR-dependent Ca2+ influx is predicted to be reduced by 75% (Single et al. 2000), as it should occur only through the 25% channels that carry two wild-type NR1 subunits. Furthermore, NMDA currents will be larger under physiological conditions in NR1+/Rneo/TgCre4 compared with wild-type mice due to the reduced Mg2+ block of NMDARs containing NR1(R) (Burnashev et al. 1992).
, 百拇医药
Here, we show that in juvenile and adult NR1+/Rneo/TgCre4 mice amplitude and charge of AMPA responses were smaller in hippocampal CA1 neurones, in nucleated patches (somatic) and at synaptic sites than in wild-type. The characterization of the input–output relationship for NMDA currents as well as CV analyses and minimal stimulation experiments indicated normal presynaptic function at hippocampal CA3-to-CA1 synapses of NR1+/Rneo/TgCre4 mice. Instead, a postsynaptic change is indicated by the smaller and less frequent AMPA miniature currents in the mutant. Thus, NMDARs with reduced Ca2+ permeability scaled down postsynaptic AMPA currents in the hippocampus in vivo.
, http://www.100md.com
Methods
HEK293 cells
HEK293 cells were cotransfected with plasmids encoding NR1-1a or NR1-1a(N598R), NR2A and green fluorescent protein (GFP) (Chen & Okayama, 1987). Forty eight hours after transfection, cells were continuously perfused with (mM): 135 NaCl, 5.4 KCl, 1.8 CaCl2, 5 Hepes (pH 7.3). GFP-labelled cells were lifted from the coverslip and whole-cell currents were activated at –60 mV in the presence of 10 μM glycine by fast applying 1 mM glutamate from a Piezo-driven double-barrelled pipette using an EPC-9 amplifier (HEKA, Lambrecht, Germany). Patch electrodes (3–5 M) were filled with (mM): 140 CsCl, 2 MgATP, 10 EGTA, 10 Hepes (pH 7.3, 290–305 mosmol l–1).
, http://www.100md.com
NR1+/Rneo/TgCre4 mice
Heterozygous NR1+/Rneo mice (Single et al. 2000) were bred with transgenic Camkcre4 (TgCre4) mice (Mantamadiotis et al. 2002) to generate NR1+/Rneo/TgCre4 mice in NMRI background to obtain bigger litters compared with C57Bl/6. NR1+/Rneo/TgCre4 mice carried one wild-type and one mutated NR1 allele and showed forebrain-restricted NR1(R) expression. Cre expression in NR1+/Rneo/TgCre4 mice was investigated in fixed hippocampal sections by immunohistochemistry and revealed significant mosaic expression in the CA1 pyramidal cell layer at P3 but not at P12 (not shown), indicating that NR1(R) expression does not start in a synchronous manner in NR1+/Rneo/TgCre4 mice. However, at the time points of our analysis (P14 and P42), we determined the presence or absence of NR1(R) in every whole-cell recording (except during minimal stimulation; see below). NR1(R) expression was indicated by significant NMDA currents at negative holding potentials under physiological conditions (1 mM Mg2+; 2 mM Ca2+) due to the impaired Mg2+ block (Fig. 1), or NR1(R) expression was indicated by the slower kinetics of the NMDA currents in the absence of Mg2+ (Fig. 2). Based on these characterizations, most CA1 cells at P14 (80–90%) and all CA1 cells at P42 expressed NR1(R).
, 百拇医药
A, in 2-week-old wild-type and NR1+/Rneo/TgCre4 mice, somatic NMDARs were activated in CA1 nucleated patches (n.p.) by 1 mM glutamate and synaptic NMDARs by Schaffer collateral stimulation in 1 mM Mg2+ ( and , respectively) (somatic, n = 6–8; synaptic, both n = 13). Currents were normalized to peak responses at +40 mV and show mean ± S.E.M.B, the majority of mutant neurones investigated in A (80–90%; ) expressed NR1(R), as indicated from the significantly increased NMDA current ratios recorded at +40 mV versus –80 mV in 1 mM Mg2+. *P < 0.02 or **P < 0.001. Bars indicate the mean value for each genotype.
, 百拇医药
A, 1 mM glutamate was applied in the presence of 10 μM glycine as indicated by bars to activate NMDARs in nucleated patches (n.p.) at –60 mV in 0 mM Mg2+ (left panel). The time of synaptic stimulation is indicated by the arrow (right panel). The currents displayed smaller amplitudes and slower kinetics in NR1+/Rneo/TgCre4 () compared with wild-type mice (). To compare charge, see also Table 1. B, wild-type NR1/NR2A () and mutated NR1(R)/NR2A () receptors were activated in HEK293 cells at –60 mV in 0 mM Mg2+ by 1 mM glutamate in the presence of glycine for 600 ms to determine desensitization (left panel) and for 20 ms to determine activation and deactivation kinetics (right panel).
, 百拇医药
Brain slices and patch pipettes
All experimental procedures were in accordance with the animal welfare guidelines of the Max-Planck-Society. Deeply anesthetized (halothane) mice were decapitated and the brain was quickly removed. Transverse hippocampal slices (250 μm) were prepared and incubated for 30 min at 37°C in artificial CSF (ACSF) containing (mM): 125 NaCl, 25 NaHCO3, 2.5 KCl, 1.25 NaH2PO4, 1 MgCl2, 25 D-glucose, 2 CaCl2; bubbled with 95% O2–5% CO2 (pH 7.4). Patch pipettes were pulled from borosilicate glass capillaries and had resistances of 3–5 M when filled with (mM): 125 caesium gluconate, 20 CsCl, 10 NaCl, 10 Hepes, 0.2 EGTA, 4 MgATP, 0.3 Na3GTP (pH 7.3, 290–305 mosmol l–1). Series resistances (16–30 M) and input resistances (100–300 M) were continuously monitored by measuring the peak and steady-state currents in response to hyperpolarizing pulses (–5 mV; 20 ms).
, http://www.100md.com
Somatic and synaptic currents
Somatic currents were activated in excised CA1 whole-soma patches (i.e. nucleated patches) by fast glutamate application at –60 mV and synaptic currents were activated at –70 mV by stimulating the Schaffer collaterals 150 μm distant from the CA1 cell body with a glass electrode filled with 1 M NaCl. In nucleated patches, AMPA currents were recorded in the presence of 50 μM D-AP5 and 1 mM MgCl2, and NMDA currents were recorded in the presence of 5 μM NBQX, 10 μM glycine and 0 or 1 mM MgCl2. To record NMDA EPSCs, 10 μM bicuculline methiodide (BMI), 5 μM NBQX, 10 μM glycine and 0 or 1 mM MgCl2 were added to the ACSF (see above); to record AMPA EPSCs, 10 μM BMI and 50 μM D-AP5 were added. The AMPA EPSC coefficient of variation (CV; mean ± S.D.) was determined from AMPA EPSCs which were evoked in the absence of D-AP5 but in the presence of 4 mM MgCl2 and 4 mM CaCl2. Spontaneous AMPA mEPSCs were recorded in ACSF containing 4 mM CaCl2, in the presence of 50 μM D-AP5, 10 μM BMI and 1 μM TTX. All recordings were performed at room temperature.
, 百拇医药
Minimal stimulation
Minimal stimulation was used to putatively stimulate one axonal fibre making one synapse on a CA1 cell (Raastad et al. 1992; Allen & Stevens, 1994). The presence of NR1(R)-containing NMDARs was demonstrated in a fraction of CA1 cells from wild-type and NR1+/Rneo/TgCre4 mice (P14, Fig. 5A). Subsequently, AMPA EPSCs were evoked by Schaffer collateral stimulation at 0.25 Hz in ACSF containing 10 μM BMI and 50 μM D-AP5. Stimulus intensity was carefully adjusted so that a significant portion of failures mixed with evoked minimal AMPA EPSCs. Only low noise recordings (S.D. of noise about 1.5 pA) with stable access resistance during the experiment and large sample numbers (n = 100–200) were analysed. Further criteria for inclusion in analysis were invariance of latency and shape of the individual minimal EPSCs for repeated stimuli and constancy of EPSC amplitude and percentage of failures during data collection. For each recording, all failures and successes were averaged to show successful discrimination of failures from successes (Fig. 5B). To quantify the size of minimal EPSCs, a 6 ms time window at the EPSC peak, typically 3–5 ms after the stimulus artifact, was selected and the charge of the minimal EPSCs was calculated. Transmission failures were detected by visual discriminination. Rarely, spontaneous events or minimal EPSCs with different onset latencies occurred, which were counted as failures. Overlap between charges of thereby visually determined failures and successes are due to such single spontaneous events (Fig. 5C). Release probability at putative single synapses was estimated by calculating the ratio between successes and total number of trials. This led to heterogeneous release probabilites across different putative single synapses which is in accordance with previous work (Dobrunz & Stevens, 1997).
, 百拇医药
A, scaled mean of EPSCs in 1 mM Mg2+/2 mM Ca2+ at –70 mV holding potential before (black) and during D-AP5 (grey) perfusion, indicating presence of NR1(R)-containing NMDARs in NR1+/Rneo/TgCre4 cells that were subjected to minimal stimulation. Failures were excluded. B, representative responses to minimal stimulation at 0.25 Hz in presence of D-AP5 as in A. Failures (F) and successes (S) recorded from a wild-type and NR1+/Rneo/TgCre4 neurone (release probabilities, 0.445 versus 0.400). The boxes mark time window selected for quantification of EPSC charge. Bottom traces are averages of all detected failures (grey) and successes (black) following 200 stimuli for each cell. C, EPSC charge plotted for failures and successes (same neurones as in B). Overlaps between failures and successes are caused by spontaneous events or events with onset latencies different from B, which were classified as failures. D, release probabilities estimated for individual recordings from wild-type (n = 6) and mutant (n = 5) neurones following 100–200 stimuli. Bars indicate the mean release probability for each genotype.
, 百拇医药
Data analysis
Spontaneous recordings were analysed off-line (filtering, 2.9 kHz; sampling, 5 kHz) using an event detection program (kindly provided by Professor Misgeld, Heidelberg). Threshold for event detection was –5 pA and detected events were inspected manually as described (Steigerwald et al. 2000). Additional off-line data analysis, curve fitting and figure preparation was performed with PulseFit (HEKA, Lambrecht, Germany) and Igor Pro (WaveMetrics, Lake Oswego, OR, USA). Deactivation kinetics of averaged NMDA currents were described using a weighted decay time constant: tw = (If/(If + Is)) tf + (Is/(Is + If)) ts, where If and Is are the amplitudes of the fast and slow decay components, and tf and ts are their respective decay time constants. For analysis and illustration, 3–10 single traces were averaged. Data are presented as mean ± S.E.M. Statistical significance was evaluated by two-tailed, unpaired Student's t test or Pearson's test.
, http://www.100md.com
Results
Functional consequences of NR1(N598R) for NMDA currents
NMDARs were activated in CA1 neurones of acute hippocampal slices from 2-week-old mice, either in nucleated patches by 1 mM glutamate or by Schaffer collateral stimulation. I–V relationships from NR1+/Rneo/TgCre4 neurones compared with wild-type showed reduced negative slope conductance under physiological ionic conditions (Fig. 1A), as predicted from in vitro experiments (Burnashev et al. 1992). On average, the ratio of NMDA currents evoked at +40 and –80 mV was about 6-fold higher in nucleated patches of NR1+/Rneo/TgCre4 compared with wild-type neurones (–5.5 ± 1.3, n = 8 versus –32.3 ± 5.9, n = 6; P = 0.02) and about 2.5-fold higher at synaptic sites (–3.0 ± 0.6, n = 13 versus –7.4 ± 1.0, n = 12; P = 0.001). When comparing the +40 mV/–80 mV ratios between individual neurones of NR1+/Rneo/TgCre4 mice with wild-type (Fig. 1B), the ratios of 7 out of 8 nucleated patch recordings and the ratios of 10 out of 13 synaptic recordings were higher than all wild-type ratios recorded at both sites, indicating that NR1(R) was present in 80–90% of the investigated neurones at P14 (for P42 see below).
, 百拇医药
The reduced Ca2+ permeability of the NMDAR population in NR1+/Rneo/TgCre4 mice by about 75% (Single et al. 2000) was indicated by a shift in the reversal potentials of NMDA currents recorded in 1.8 mM [Ca2+] (3 mV at somatic sites and 6.2 mV at synaptic sites (Fig. 1A); see also Burnashev et al. 1992 and Single et al. 2000 for major reversal potential shifts when recorded in high Ca2+ solution).
NMDA current amplitudes in NR1+/Rneo/TgCre4 compared with wild-type mice were smaller at somatic and synaptic sites when activated in Mg2+-free solution at negative potentials (Fig. 2A and Table 1). The smaller NMDA currents in NR1+/Rneo/TgCre4 mice were unlikely to be caused by one expression-silenced wild-type NR1 allele, since heterozygous NR1 knockout mice display normal NMDA currents (Forrest et al. 1994). Instead, NMDA current reduction may well be caused by reduced single channel conductance imposed by NR1(R) (Behe et al. 1995). Consistently, glutamate-activated currents mediated by pure NR1(R)/NR2A receptors in HEK293 cells were also smaller than when mediated by wild-type NR1/NR2A receptors (Fig. 2B and Table 1). In addition, we observed that NMDA currents displayed slower kinetics in NR1+/Rneo/TgCre4 mice (Fig. 2A and Table 1). Importantly, as a consequence of the slower kinetics the total charge transfer via NMDAR channels was not significantly different between the genotypes at somatic and synaptic sites (Table 1).
, 百拇医药
The slower kinetics of NMDA currents in HEK293 cells expressing NR1(R)/NR2A receptors and in CA1 neurones of NR1+/Rneo/TgCre4 mice could derive from changes in agonist binding and/or channel gating. We recorded dose–response curves of wild-type and NR1(R)-containing NMDARs in HEK293 cells (not shown) and found no change in the potency for glycine (EC50s: NR1/NR2A, 2.11 μM; NR1(R)/NR2A, 2.03 μM, but a minor increase in glutamate potency (EC50s: NR1/NR2A, 5.51 μM; NR1(R)/NR2A, 1.94 μM). In spite of the minor increase in glutamate potency, the rise time of glutamate-activated currents was slower in the presence of NR1(R), indicating that NR1(R) probably did not increase the glutamate binding rate. A concurrently performed study on recombinant NR1(R)/NR2A receptors observed very similar changes on kinetics and glutamate potency (Chen et al. 2004). In this study, a kinetic model explained the leftward shift in the glutamate dose–response curve by a about 10-fold reduction in the channel closing rate, even without any modification in ligand binding and unbinding rates. Thus, the slower glutamate-activated currents we observed in neurones expressing NR1(R) are very likely to be independent of any ligand binding modification. Furthermore in our NR1+/Rneo/TgCre4 mice, we excluded the possibility that a change in NR2A expression relative to NR2B is responsible for the altered current kinetics, since NMDA currents in nucleated patches from CA1 neurones of both genotypes showed comparable sensitivity to the NR2B-specific antagonist ifenprodil (3 μM; not shown).
, http://www.100md.com
In summary, the presence of NR1(R) in hippocampal CA1 neurones was responsible for the reduced voltage-dependent Mg2+ block and the impaired Ca2+ permeability of the NMDAR population. However, the charge carried by NMDARs was not different between both genotypes at somatic and synaptic sites.
AMPA currents are reduced in CA1 neurones of NR1+/Rneo/TgCre4 mice
Next, we investigated whether NMDARs with impaired Ca2+ permeability would affect acquisition of functional AMPARs in the hippocampus of 2-week-old NR1+/Rneo/TgCre4 mice. AMPA current amplitudes were reduced by about 30% in nucleated patches and at synaptic sites in CA1 pyramidal cells of NR1+/Rneo/TgCre4 mice (Fig. 3A and Table 1). Subsequent to the recording of AMPA currents, NMDA currents were recorded for each neurone in the absence of extracellular Mg2+. Decay kinetics of NMDA currents were used to indicate the presence of NR1(R) (see above) and related to the amplitude of AMPA currents (Fig. 3B). In recordings from wild-type neurones, deactivation of NMDA currents did not correlate with the size of AMPA currents. In contrast, in synaptic recordings from NR1+/Rneo/TgCre4 neurones, NMDA decay and AMPA peak currents correlated, raising the possibility that the smaller AMPA currents at synaptic sites are a consequence of NR1(R) expression. In somatic recordings, the correlation coefficient was only close to significance in NR1+/Rneo/TgCre4 neurones (for explanation see below). The charge carried through AMPAR channels was also reduced in NR1+/Rneo/TgCre4 mice, since the kinetics of the AMPA currents remained unchanged. In these experiments, the position of the stimulation electrode relative to the CA1 cell was kept constant. However, post hoc comparison of stimulation intensities between cells revealed that higher intensities were used to activate synapses in slices of NR1+/Rneo/TgCre4 compared with wild-type mice (by 33.3%; P = 0.007; Table 1). Thus, the AMPA current reduction at synaptic sites in NR1+/Rneo/TgCre4 mice is actually even higher than 30%. This became evident when relating AMPA current amplitudes to identical stimulation intensities, which were increased in steps of 5 V, leading to 2- to 3-fold reduced AMPA currents (Fig. 4A). The fact, that AMPA currents in NR1+/Rneo/TgCre4 neurones were more drastically reduced at synaptic than at somatic sites, could explain why a correlation between the presence of NR1(R) and size of AMPA currents was found for synaptic but not for somatic recordings.
, 百拇医药
A, AMPA current amplitudes in NR1+/Rneo/TgCre4 mice (open bars) were reduced by about 30% compared with wild-type (filled bars) in nucleated patches (n.p.) and at synaptic sites. Bars are mean ± S.E.M., number of cells is indicated in parentheses, and level of significance is either **P < 0.001 or *P < 0.05. A representative current trace is given next to each bar. For NMDA currents, see Table 1. B, correlation of peak AMPA currents and decay kinetics of NMDA currents. Slowly deactivating NMDA currents indicate presence of NR1(R) in NR1+/Rneo/TgCre4 mice () and significantly correlate with AMPA peak currents in synaptic recordings from NR1+/Rneo/TgCre4 cells (correlation coefficient R = 0.72, n = 9, P < 0.05). No significant correlation was found in n.p. recordings of NR1+/Rneo/TgCre4 cells (R = 0.45, n = 14, P > 0.05). Recordings from wild-type cells () showed no correlation between size of AMPA current and NMDA decay (synaptic: R = –0.21, n = 8; n.p. R = 0.18, n = 23, P > 0.05). Significance of correlation was assessed with Pearson's test.
, 百拇医药
A, synaptic AMPA currents were evoked at –70 mV and are plotted as a function of stimulation intensity. The amplitudes were 2- to 3-fold reduced in NR1+/Rneo/TgCre4 (; n = 12) compared with wild-type mice (; n = 7). *P < 0.05. B, synaptic NMDA currents were evoked at +40 mV in solutions containing 1 mM Mg2+ extracellularly and 2 mM Mg2+ intracellularly. Current amplitudes were not significantly different between NR1+/Rneo/TgCre4 (; n = 5) and wild-type mice (; n = 5). C, paired-pulse facilitation (PPF) was determined at 50 ms intervals and was unchanged in NR1+/Rneo/TgCre4 mice. AMPA and NMDA EPSCs were evoked in 1 mM Mg2+ at –70 and –30 mV, respectively. Bars are mean ± S.E.M.
, 百拇医药
To provide additional evidence that the impaired AMPAR function resulted from expression of NR1(R), we also recorded AMPA currents in littermates of NR1+/Rneo/TgCre4 mice which lack the Cre transgene (NR1+/Rneo mice, Single et al. 2000). As a result, NR1(R) expression is not turned on by Cre-mediated recombination. AMPA current amplitudes were comparable between wild-type and NR1+/Rneo mice, in nucleated patches (WT, –2520 ± 95 pA, n = 27; NR1+/Rneo, –2442 ± 106 pA, n = 19; P = 0.59) and at synaptic sites in CA1 pyramidal cells (WT, –77.4 ± 8.1 pA, n = 15; NR1+/Rneo, –99.1 ± 13.3 pA, n = 7; P = 0.19). The stimulation intensities to evoke AMPA EPSCs in NR1+/Rneo mice were similar to those used in wild-type recordings (P = 0.17), strengthening the conclusion that AMPA currents were not downscaled in the absence of Cre.
, http://www.100md.com
In NR1+/Rneo/TgCre4 mice, the AMPA current reduction was also found in adult animals. At P42, synaptic AMPA current amplitudes were –70 ± 6 pA (n = 6) in wild-type and –42 ± 5 pA (n = 9) in mutant neurones (P = 0.006), and the stimulation intensities were 88% higher in NR1+/Rneo/TgCre4 than in wild-type recordings (11 ± 2 V versus 20 ± 1 V; P = 0.04). At this age, every neurone expressed NR1(R), as indicated by prominent and slow NMDA components of composite EPSCs (data not shown).
, http://www.100md.com
To investigate whether afferent innervation was still intact in NR1+/Rneo/TgCre4 mice, synaptic NMDA currents were evoked using increasing stimulation intensities at +40 mV in 2-week-old mice (Fig. 4B). At this positive potential, the 2 mM Mg2+ present in the intracellular solution blocked wild-type NMDARs to a greater extent than NR1(R)-containing NMDARs in NR1+/Rneo/TgCre4 mice. This is because NR1(R) also reduces the intracellular Mg2+ block (Wollmuth et al. 1998). As a consequence, NMDA current amplitudes were indistinguishable at +40 mV between genotypes (160 ± 21 pA; n = 13 versus 177 ± 29 pA; n = 13). When stimulation intensities were increased, the amplitudes of NMDA currents similarly increased in both genotypes and were not significantly different at any tested stimulation intensity, indicating normal afferent innervation of CA1 neurones in NR1+/Rneo/TgCre4 mice.
, http://www.100md.com
To investigate presynaptic function at CA3-to-CA1 synapses of NR1+/Rneo/TgCre4 mice, we first determined paired-pulse facilitation (PPF) for AMPA and NMDA currents delivering two stimuli to the Schaffer collaterals at intervals of 50 ms (Fig. 4C). The PPF ratio of AMPA and NMDA currents was comparable between wild-type and NR1+/Rneo/TgCre4 mice. Next, we estimated the AMPA EPSC coefficient of variation (CV; S.D. mean–1), which could be affected by changes in release probability. When calculating the CV from 60 EPSCs, which had peak amplitudes between 50 and 100 pA and were evoked at –70 mV in each neurone, the CV was found to be comparable (wild-type, 0.32 ± 0.03; n = 15 and NR1+/Rneo/TgCre4 mice, 0.29 ± 0.02; n = 22). Lastly, we performed minimal stimulation experiments to compare the release probability between genotypes (Fig. 5). First, we evoked EPSCs in both genotypes in 1 mM Mg2+/2 mM Ca2+ at –70 mV to reassure NR1(R) expression in the CA1 neurones of NR1+/Rneo/TgCre4 mice (see prominent NMDA component in composite EPSC of mutant neurone; Fig. 5A). Following blockade of the NMDA component by D-AP5, failures or successes (minimal AMPA EPSCs) were evoked during 100–200 minimal stimulations in neurones of both genotypes (see example traces in Fig. 5B). The analysis of all individual failures and successes of these two neurones is shown in Fig. 5C. In Fig. 5D the release probabilities of the individual recordings from wild-type and mutant cells are plotted, showing that release probability at CA3-to-CA1 synapses of NR1+/Rneo/TgCre4 mice is unchanged (wild-type, 0.31 ± 0.06; n = 6 and NR1+/Rneo/TgCre4 mice, 0.29 ± 0.05; n = 5).
, 百拇医药
To test for postsynaptic changes, AMPA miniature EPSCs (AMPA mEPSCs) were recorded in tetrodotoxin (TTX) and D-AP5. Under these conditions the S.D. of the noise was comparable between wild-type and NR1+/Rneo/TgCre4 mice (1.70 ± 0.07 pA versus 1.64 ± 0.03 pA; n = 10 each; Fig. 6A) and was favourable in detecting miniature AMPA responses. Most (24 out of 34) CA1 cells in NR1+/Rneo/TgCre4 mice in contrast to 1 out of 10 cells in wild-type displayed less than 1 event per recording minute. From cells with frequencies above 1 event min–1, around 100 events were averaged. Averaged AMPA mEPSC rise and decay times were unchanged; their amplitudes, however, were decreased by 21% in NR1+/Rneo/TgCre4 mice (Fig. 6B and Table 1). Consistently, when constructing amplitude histograms which contained all analysed events (WT, n = 1594; NR1+/Rneo/TgCre4, n = 1225; Fig. 6C), both histograms could be fitted to a lognormal function and confirmed the 21% smaller mEPSC amplitudes in NR1+/Rneo/TgCre4 mice (10.9 ± 0.1 pA versus 8.6 ± 0.1 pA; Fig. 6C) found arithmetically (Table 1).
, 百拇医药
A, representative current traces of AMPA mEPSCs recorded from a wild-type (WT) and a NR1+/Rneo/TgCre4 neurone at –70 mV in presence of TTX and D-AP5. B, about 100 events per neurone were averaged (left). The bars indicate the 21% reduced amplitudes of AMPA mEPSCs in NR1+/Rneo/TgCre4 neurones (*P = 0.004), whereas the rise and decay times were unchanged relative to wild-type (= 100%; see bars). (WT, filled symbols and NR1+/Rneo/TgCre4, open symbols.) See also Table 1. C, histograms, which include all events of all measured cells (WT, n = 1594; NR1+/Rneo/TgCre4, n = 1225), were fitted by the lognormal function yo + A exp (– (ln(x/xo)/peak width)2), where yo sets the baseline, A sets the amplitude and xo sets the peak position in x. The frequencies of AMPA miniature EPSCs were determined in a different set of experiments and the reduced frequencies in NR1+/Rneo/TgCre4 mice were illustrated in a distribution plot (P = 0.003). Bars indicate the average miniature frequency for each genotype.
, 百拇医药
The reduction in mEPSC amplitudes was too small to explain the 2- to 3-fold reduced synaptic AMPA currents
(see above). Therefore, we considered the possibility of an increased number of synapses devoid of functional AMPARs in NR1+/Rneo/TgCre4 mice, since we noted fewer miniature events in NR1+/Rneo/TgCre4 compared with wild-type mice (see above). In an additional set of experiments, AMPA mEPSC frequencies were recorded from 10 cells of both genotypes and 9 min traces were analysed each. Again consistent with the observation that about 80–90% of the CA1 neurones of 2-week-old NR1+/Rneo/TgCre4 mice express NR1(R), 9 out of 10 neurones were similar with respect to very low mEPSC frequencies. On average, mEPSC frequencies were reduced to 1.2 ± 0.8 events min–1 in NR1+/Rneo/TgCre4 neurones compared with 5.4 ± 0.9 events min–1 in wild-type (P = 0.003; Fig. 6C; both n = 10). In agreement with the initial mEPSC experiments, the amplitudes of AMPA mEPSCs were decreased by 19% (P = 0.007). This amplitude reduction together with the concurrent frequency reduction of AMPA mEPSCs reduced the spontaneous AMPA currents during the 9 min recordings 4- to 5-fold in CA1 neurones of NR1+/Rneo/TgCre4 compared with wild-type mice (1215.0 pA from 112 events versus 5435.5 pA from 441 events). This 4- to 5-fold reduction is close to the 2- to 3-fold reduced amplitudes of evoked AMPA currents (Fig. 4A).
, http://www.100md.com
Collectively, afferent innervation and release probability remained unchanged at CA3-to-CA1 synapses in NR1+/Rneo/TgCre4 mice. Conversely, AMPA EPSCs were smaller and mEPSCs in addition less frequent in the mutant, indicating postsynaptic downscaling of AMPA responses.
Discussion
Juvenile and adult mice expressing NMDARs with reduced Ca2+ permeability in forebrain principal neurones (NR1+/Rneo/TgCre4) displayed smaller AMPA current amplitudes and charges at hippocampal CA3-to-CA1 synapses. Paired-pulse facilitation, input–output relationships for NMDA EPSCs and release probability examined by minimal stimulation were unchanged, implying a postsynaptic locus for this down-regulation of AMPA responses. In agreement with a postsynaptic locus, AMPA mEPSC amplitudes were slightly reduced and the frequencies of AMPA mEPSCs were severely decreased.
, http://www.100md.com
Studies in cultured neurones indicated a differential contribution of amplitude and/or frequency changes of AMPA mEPSCs to homeostatic synaptic plasticity and revealed differences between hippocampal and cortical cultures (see introduction and Turrigiano & Nelson, 2004). These differences could arise from the fact that the number of AMPAR-free synapses is higher in cultured hippocampal (Liao et al. 1999) than in cortical neurones (Pratt et al. 2003). It has been suggested that the functionalization of AMPAR-free synapses will increase the mEPSC frequency with only a modest enhancement of amplitude (Burrone & Murthy, 2003), and it has been shown that reduced AMPAR surface expression decreased the frequency but not the amplitude of AMPA mEPSCs (Noel et al. 1999). Based on these observations, the moderate decrease in AMPA mEPSC amplitude in hippocampal CA1 neurones of NR1+/Rneo/TgCre4 mice indicate synapses with reduced AMPAR content and the drastic decrease in mEPSC frequency is consistent with a lower number of AMPAR-containing synapses. The recordings of AMPA responses by stimulating the Schaffer collaterals indicated a 2- to 3-fold reduction in AMPA current. When we reduce all our mEPSC events recorded in wild-type neurones by 60%, 40% of all mEPSCs would fall below our detection limit of –5 pA. In our recordings, we found a reduction in mEPSC frequency by 80% in NR1+/Rneo/TgCre4 mice. This is a good indication that actually a higher number of AMPA-free synapses exist in NR1+/Rneo/TgCre4 mice. The equally possible mechanism, that AMPAR content has been scaled down homogeneously across all synapses in NR1+/Rneo/TgCre4 mice would also generate synapses for which AMPARs can no longer be detected, since the postsynaptic AMPAR content, different from the NMDAR content, is highly variable at CA3-to-CA1 synapses, mainly containing a very small AMPAR number (Takumi et al. 1999; Racca et al. 2000).
, http://www.100md.com
In NR1+/Rneo/TgCre4 mice, fewer AMPARs might be needed in synapses, since, due to their reduced Mg2+ block, NR1(R)-containing NMDARs are activated when glutamate is released under physiological ionic conditions. The observed unchanged release probability and normal input–output relationship for NMDA EPSCs, as opposed to the altered input–output relationship for AMPA EPSCs in CA1 neurones of NR1+/Rneo/TgCre4 mice, are consistent with this hypothesis. The time required for the AMPA current reduction following onset of NR1(R) expression cannot be determined in our current mouse model, since Cre expression in CA1 neurones does not start in a synchronous manner, and the time from Cre expression until synaptic insertion of NR1(R)-containing NMDARs in CA1 neurones is unknown. Virus-driven Cre expression could be a first step to achieve at least synchrony of Cre expression.
, 百拇医药
Several putative mechanisms for synaptic scaling of AMPA responses have been discussed, including changes in the level of postsynaptic depolarization and/or NMDAR activation and subsequently activated Ca2+-dependent mechanisms. The level of postsynaptic depolarization in response to synaptic glutamate receptor activation could be enhanced in NR1+/Rneo/TgCre4 mice in 1 mM Mg2+, since NR1(R)-containing NMDARs mediate postsynaptic depolarization close to the resting membrane potential. However, in a similar mouse mutant expressing NMDARs with strongly reduced Mg2+ block but largely unaltered Ca2+ permeability the synaptic AMPA currents remained unchanged (Single et al. 2002), excluding the possibility that AMPA currents in NR1+/Rneo/TgCre4 mice were down-regulated by enhanced NMDAR-mediated postsynaptic depolarization. Here, we show in NR1+/Rneo mice that AMPAR expression is not impaired before NR1(R) expression gets turned on and that due to the slower kinetics of NMDA currents the charge transfer via NMDAR channels in NR1+/Rneo/TgCre4 mice is unaltered. This leaves the reduced Ca2+ influx via NMDARs as the most likely explanation for the downscaling of postsynaptic AMPA responses, consistent with a Ca2+-dependent homeostatic plasticity model (Yeung et al. 2004). Smaller postsynaptic Ca2+ elevations via NMDARs could impair a shift of AMPARs from a degradative to a recycling pathway (Ehlers, 2000) and/or could impair the Ca2+/CaM sensitivity of CaMKII which may be important for activity-dependent synaptic homeostasis (Thiagarajan et al. 2002). Our finding that reduced NMDAR-mediated Ca2+ influx scaled down AMPA responses also at somatic sites (in nucleated patches) might be explained by the possibility that AMPARs are first inserted at extrasynaptic sites before moving laterally to synaptic sites (Passafaro et al. 2001).
, 百拇医药
In contrast to cortical neurones, NMDAR-mediated Ca2+ influx appears to regulate the AMPA currents in hippocampal CA1 neurones in vivo. In addition, NMDARs with reduced Ca2+ permeability may also regulate postsynaptic AMPA responses during synaptic plasticity, which will be addressed in the future.
References
Allen C & Stevens CF (1994). An evaluation of causes for unreliability of synaptic transmission. Proc Natl Acad Sci U S A 91, 10380–10383.
, 百拇医药
Behe P, Stern P, Wyllie DJ, Nassar M, Schoepfer R & Colquhoun D (1995). Determination of NMDA NR1 subunit copy number in recombinant NMDA receptors. Proc R Soc Lond B Biol Sci 262, 205–213.
Burnashev N, Schoepfer R, Monyer H, Ruppersberg JP, Gunther W, Seeburg PH & Sakmann B (1992). Control by asparagine residues of calcium permeability and magnesium blockade in the NMDA receptor. Science 257, 1415–1419.
Burrone J & Murthy VN (2003). Synaptic gain control and homeostasis. Curr Opin Neurobiol 13, 560–567.
, 百拇医药
Burrone J, O'Byrne M & Murthy VN (2002). Multiple forms of synaptic plasticity triggered by selective suppression of activity in individual neurons. Nature 420, 414–418.
Chen N, Li B, Murphy TH & Raymond LA (2004). Site within N-methyl-D-aspartate receptor pore modulates channel gating. Mol Pharmacol 65, 157–164.
Chen C & Okayama H (1987). High-efficiency transformation of mammalian cells by plasmid DNA. Mol Cell Biol 7, 2745–2752.
, 百拇医药
Dobrunz LE & Stevens CF (1997). Heterogeneity of release probability, facilitation, and depletion at central synapses. Neuron 18, 995–1008.
Ehlers MD (2000). Reinsertion or degradation of AMPA receptors determined by activity-dependent endocytic sorting. Neuron 28, 511–525.
Forrest D, Yuzaki M, Soares HD, Ng L, Luk DC, Sheng M, Stewart CL, Morgan JI, Connor JA & Curran T (1994). Targeted disruption of NMDA receptor 1 gene abolishes NMDA response and results in neonatal death. Neuron 13, 325–338.
, 百拇医药
Leslie KR, Nelson SB & Turrigiano GG (2001). Postsynaptic depolarization scales quantal amplitude in cortical pyramidal neurons. J Neurosci 21, RC170.
Liao D, Zhang X, O'Brien R, Ehlers MD & Huganir RL (1999). Regulation of morphological postsynaptic silent synapses in developing hippocampal neurons. Nat Neurosci 2, 37–43.
Lissin DV, Gomperts SN, Carroll RC, Christine CW, Kalman D, Kitamura M, Hardy S, Nicoll RA, Malenka RC & Von Zastrow M (1998). Activity differentially regulates the surface expression of synaptic AMPA and NMDA glutamate receptors. Proc Natl Acad Sci U S A 95, 7097–7102.
, 百拇医药
Malenka RC & Nicoll RA (1999). Long-term potentiation – a decade of progress Science 285, 1870–1874.
Malinow R & Malenka RC (2002). AMPA receptor trafficking and synaptic plasticity. Annu Rev Neurosci 25, 103–126.
Mantamadiotis T, Lemberger T, Bleckmann SC, Kern H, Kretz O, Martin Villalba A, Tronche F, Kellendonk C, Gau D, Kapfhammer J, Otto C, Schmid W & Schutz G (2002). Disruption of CREB function in brain leads to neurodegeneration. Nat Genet 31, 47–54.
, 百拇医药
Noel J, Ralph GS, Pickard L, Williams J, Molnar E, Uney JB, Collingridge GL & Henley JM (1999). Surface expression of AMPA receptors in hippocampal neurons is regulated by an NSF-dependent mechanism. Neuron 23, 365–376.
Passafaro M, Piech V & Sheng M (2001). Subunit-specific temporal and spatial patterns of AMPA receptor exocytosis in hippocampal neurons. Nat Neurosci 4, 917–926.
Pratt KG, Watt AJ, Griffith LC, Nelson SB & Turrigiano GG (2003). Activity-dependent remodeling of presynaptic inputs by postsynaptic expression of activated CaMKII. Neuron 39, 269–281.
, http://www.100md.com
Raastad M, Storm JF & Andersen P (1992). Putative single quantum and single fibre excitatory postsynaptic currents show similar amplitude range and variability in rat hippocampal slices. Eur J Neurosci 4, 113–117.
Racca C, Stephenson FA, Streit P, Roberts JD & Somogyi P (2000). NMDA receptor content of synapses in stratum radiatum of the hippocampal CA1 area. J Neurosci 20, 2512–2522.
Single FN, Rozov A, Burnashev N, Zimmermann F, Hanley DF, Forrest D, Curran T, Jensen V, Hvalby O, Sprengel R & Seeburg PH (2000). Dysfunctions in mice by NMDA receptor point mutations NR1 (N598Q) and NR1 (N598R). J Neurosci 20, 2558–2566.
, 百拇医药
Single FN, Rozov A, Burnashsev N, Wollmuth LP, Kuner T, Jensen V, Hvalby O, Sprengel R & Seeburg PH (2002). Gene targeted mice with NMDA receptor subunit NR2A point mutation N596S show elevated levels of LTP. Program No. 150.6 2002 Abstract Viewer and Itinerary Planner. Society for Neuroscience Abstracts. Online.
Steigerwald F, Schulz TW, Schenker LT, Kennedy MB, Seeburg PH & Khr G (2000). C-Terminal truncation of NR2A subunits impairs synaptic but not extrasynaptic localization of NMDA receptors. J Neurosci 20, 4573–4581.
, http://www.100md.com
Takumi Y, Ramirez-Leon V, Laake P, Rinvik E & Ottersen OP (1999). Different modes of expression of AMPA and NMDA receptors in hippocampal synapses. Nat Neurosci 2, 618–624.
Thiagarajan TC, Piedras-Renteria ES & Tsien RW (2002). alpha- and betaCaMKII. Inverse regulation by neuronal activity and opposing effects on synaptic strength. Neuron 36, 1103–1114.
Turrigiano GG, Leslie KR, Desai NS, Rutherford LC & Nelson SB (1998). Activity-dependent scaling of quantal amplitude in neocortical neurons. Nature 391, 892–896.
, 百拇医药
Turrigiano GG & Nelson SB (2004). Homeostatic plasticity in the developing nervous system. Nat Rev Neurosci 5, 97–107.
Umemiya M, Senda M & Murphy TH (1999). Behaviour of NMDA and AMPA receptor-mediated miniature EPSCs at rat cortical neuron synapses identified by calcium imaging. J Physiol 521, 113–122.
Watt AJ, Van Rossum MC, MacLeod KM, Nelson SB & Turrigiano GG (2000). Activity coregulates quantal AMPA and NMDA currents at neocortical synapses. Neuron 26, 659–670.
, 百拇医药
Wollmuth LP, Kuner T & Sakmann B (1998). Intracellular Mg2+ interacts with structural determinants of the narrow constriction contributed by the NR1-subunit in the NMDA receptor channel. J Physiol 506, 33–52.
Yeung LC, Shouval HZ, Blais BS & Cooper LN (2004). Synaptic homeostasis and input selectivity follow from a calcium dependent plasticity model. Proc Natl Acad Sci U S A 101, 14943–14948., 百拇医药(Verena Pawlak, Bettina J )