Abstract

Delayed asynchronous release (AR) evoked by bursts of presynaptic action potentials (APs) occurs in certain types of hippocampal and neocortical inhibitory interneurons. Previous studies showed that AR provides long-lasting inhibition and desynchronizes the activity in postsynaptic cells. However, whether AR undergoes developmental change remains unknown. In this study, we performed whole-cell recording from fast-spiking (FS) interneurons and pyramidal cells (PCs) in prefrontal cortical slices obtained from juvenile and adult rats. In response to AP trains in FS neurons, AR occurred at their output synapses during both age periods, including FS autapses and FS-PC synapses; however, the AR strength was significantly weaker in adults than that in juveniles. Further experiments suggested that the reduction of AR in adult animals could be attributable to the rapid clearance of residual Ca2+ from presynaptic terminals. Together, our results revealed that the AR strength was stronger at juvenile but weaker in adult, possibly resulting from changes in presynaptic Ca2+ dynamics. AR changes may meet the needs of the neural network to generate different types of oscillations for cortical processing at distinct behavioral states.

Introduction

The neocortical network consists of interweaved excitatory and inhibitory neurons (Markram et al. 2004). Activation of inhibitory interneurons mainly causes membrane potential (Vm) hyperpolarization and thus inhibition in their target cells (Glickfeld et al. 2009). Among different types of interneurons, the parvalbumin-containing fast-spiking (FS) neuron that can discharge nonadapting high-frequency action potentials (APs) is the most prevalent interneuronal cell type in the neocortex (Kawaguchi and Kubota 1997; Markram et al. 2004). FS neurons mainly innervate the perisomatic region of pyramidal cells (PCs) and control the timing of postsynaptic APs (Miles et al. 1996). Moreover, activation of these interneurons plays important roles in maintaining rhythmic oscillations (Cardin et al. 2009; Sohal et al. 2009) and regulating behavior states (Fuchs et al. 2007; Lee et al. 2012).

Previous studies showed that neurotransmitter release from FS neurons is both synchronous and asynchronous. Synchronous release of GABA occurs at FS neuron terminals within a narrow time window (∼1 ms) after each AP (Sabatini and Regehr 1999); however, asynchronous release (AR) takes place in a wide time window (up to hundreds of milliseconds) after AP generation, particularly in response to high-frequency discharges (Hefft and Jonas 2005; Daw et al. 2009; Manseau et al. 2010; Jiang et al. 2012). In hippocampus, in comparison with parvalbumin-containing FS interneurons, cholecystokinin-expressing non-FS interneurons show much stronger asynchronous GABA release in response to a burst of APs (Hefft and Jonas 2005; Daw et al. 2009). In neocortex, however, recent studies revealed that after a burst of APs neocortical FS neurons show prolonged AR at their output synapses, including FS autapses (synapses onto themselves), FS-to-FS and FS-to-PC synapses (Manseau et al. 2010; Jiang et al. 2012; Ma and Prince 2012). This prolonged asynchronous GABA release causes long-lasting inhibition and reduction of the spiking precision in postsynaptic neurons (Hefft and Jonas 2005; Manseau et al. 2010). A recent study showed that AR occurs in both human and rat neocortical FS neurons, and its strength is upregulated in epileptic tissues (Jiang et al. 2012). Considering the key role of GABAergic transmission mediated by FS interneurons in shaping network activity and regulating network function (Tamas et al. 2000; Freund 2003; Bartos et al. 2007), one can speculate that changes in the strength of asynchronous GABA release could also contribute to the regulation of network activities.

At early postnatal developmental stages, FS neurons undergo morphological changes, including extension of neurites, proliferation of boutons, and formation of synapses (Chattopadhyaya et al. 2004; Huang et al. 2007). This maturation process could be controlled by neuronal activity and sensory inputs (Chattopadhyaya et al. 2004). Furthermore, synaptic properties were refined to ensure a fast and precise synaptic transmission, by showing shortened synaptic delay and increased synaptic efficiency (Doischer et al. 2008; Pangratz-Fuehrer and Hestrin 2011; Sauer and Bartos 2011). These studies mainly focused on developmental changes in synchronous GABA release from FS neurons; however, it remains unclear whether AR is subjected to a developmental regulation.

In this study, we compared the strength of AR and synchronous release from neocortical FS neurons in juvenile and adult animals. We found that, in both age periods, AR occurred at output synapses of FS neurons. Interestingly, AR strength was substantially weaker in adult in comparison with juvenile animals. This developmental reduction of AR strength may result from a rapid recovery of Ca2+ transients in axonal terminals in adults.

Materials and Methods

Ethics Statement

The use and care of animals complied with the guidelines of the Animal Advisory Committee at the Shanghai Institutes for Biological Sciences.

Slice Preparation

Most of the experiments were performed in 2 groups of SD rats, the juvenile group (P15–19, 45–51 g) and adult group (P50–55, 265–305 g). In a subset of experiments, we also performed recordings from P32 to P34 rats for AR strength comparison (Fig. 5). Cortical slices were obtained from these animals with standard slicing procedures. In brief, the animal was first anesthetized with sodium pentobarbital (40 mg/kg, i.p.) and sacrificed with decapitation. The brain was then dissected out and immersed into ice-cold oxygenated sucrose ACSF in which the NaCl was substituted with equiosmolar sucrose and dextrose was reduced to 10 mM. Coronal slices (350 μm in thickness) from prefrontal cortices were cut in this ice-cold sucrose ACSF and maintained in an incubation chamber at 35.5 °C. After incubation for ∼1 h, the slices were then transferred into a submerge-style recording chamber and perfused with regular ACSF containing (in mM): 126 NaCl, 2.5 KCl, 2 MgSO4, 2 CaCl2, 26 NaHCO3, 1.25 NaH2PO4, 25 dextrose (315 mOsm, pH 7.4). Neurons in the slice were visualized with an infrared-differential interference contrast (IR-DIC) microscope. The temperature during the electrophysiological recording was maintained stable around 35.5 °C.

In some experiments (Supplementary Fig. 4), we prepared prefrontal slices from B13 parvalbumin (PV)-EGFP transgenic mice (juvenile group: P15–21; adult group: P55–57), in which PV-containing neurons could be visualized under fluorescent microscope due to the expression of EGFP (Dumitriu et al. 2007). We compared the autaptic AR frequencies in PV-positive FS cells at the 2 ages.

Paired Recording

We performed whole-cell recording from cortical neurons with a Multiclamp 700B amplifier (Molecular Devices). Usually, FS-PC paired recordings were obtained in adjacent neurons with intersomatic distance of ∼20 μm, no larger than 50 μm. Voltage and current signals were filtered at 10 kHz and sampled at 20 kHz with Micro 1401 mkII data acquisition system and Spike 2 software (Cambridge Electronic Design, UK). The impedance of patch pipettes was about 5–7 MΩ. Internal solution for whole-cell recording contained (in mM): 71 KCl, 72 Kgluconate, 2 MgCl2, 10 HEPES, 0.025 BAPTA, and 2 Na2ATP (pH 7.2 with KOH, 288 mOsm). With this internal solution, the calculated reversal potential for Cl was −15 mV, and thus, the IPSPs were depolarizing potentials and IPSCs were inward currents at Vm levels more negative than −15 mV. Alexa Fluor 488 (50 μM) and biocytin (0.2%) were added to the internal solution for visualizing the morphology of recorded neurons. In these experiments, trains of APs with varying numbers (2–40 APs) and frequencies (50–200 Hz) were evoked by trains of step-current injections (in current-clamp mode) or voltage commands (in voltage-clamp mode). The occurrence of AR was examined by 40-AP train stimulation (200 Hz). Unless otherwise stated, properties of PT-AR induced by 20 APs at 200 Hz were analyzed and illustrated in figures. Liquid junction potential (∼9 mV) has not been corrected for the Vm shown in the text and figures.

Picrotoxin (PTX, GABAA receptor antagonist, Tocris), EGTA-AM (membrane permeable Ca2+ chelator, Invitrogen) were applied through bath perfusion.

Two-Photon Ca2+ Imaging

Calcium imaging was performed with a 40× (NA 0.8) water immersion objective on an upright laser-scanning microscope (BX-61WI, Olympus). Two-photon excitation was achieved with a mode-locked Ti:S laser (Chameleon Vision II, Coherent) running at a wavelength of 810 nm (repetition rate: 80 MHz; pulse width: 140 fs). We recorded FS neurons with a pipette solution similar to that of paired recording but without any additional Ca2+ buffer. We utilized Fluo-5F (200 μM) as a Ca2+ indicator to monitor Ca2+ transients. Alexa Fluor 594 (50 μM) was added into the internal solution for visualization of the cell morphology. To minimize the bleaching effect, a low laser power (typically between 2 and 7 mW under the objective) was used. The loading of Ca2+ indicator took about 20 min; Ca2+ transients could be then measured. We identified the axons of FS neurons by their small diameter, randomly curved collaterals with densely distributed boutons. Axonal terminals were imaged with 4 or 5× digital zoom. APs were elicited by somatic injection of current pulses (1 ms, 1.5 nA) at 200 Hz. The number of APs varied from 2 to 60. Fluorescence changes were monitored in line scan mode. An individual sweep of line scan took 2.5 ms, and a trail of scan contained 1000 repeats and took 2500 ms. Trials with severe bleaching (>20% rundown in red fluorescence intensity during scanning) were discarded. For each bouton and stimulus intensity, the fluorescent signals from 5 trials were averaged to increase the signal/noise ratio. Imaging data were acquired with Fluoview FV1000 (Olympus) and further analyzed by using ImageJ and MATLAB (MathWorks, Bethesda, MD, USA).

To measure Ca2+ changes induced by somatic stimulation, we subtracted the mean baseline intensity (0–200 ms, Fluo-5F channel, green) from the fluorescence signals to obtain the ΔG trace, which was then divided by the averaged intensity R of the red fluorescence (0–1000 ms, Alexa Fluor 594 channel) to obtain the ratio of ΔG/R. Because fluorescence changes of Ca2+ indicator were normalized to the red fluorescence that was insensitive to Ca2+, this ratio could be used for comparison between different age groups. Single exponential fits (decay phase: from 90% to 10% of peak amplitude) were performed to obtain the decay time constant of Ca2+ transients. Quality of these fits was evaluated by r values (Supplementary Fig. 5).

Data Analysis

In order to quantitatively analyze the strength of post-train AR (PT-AR), PT-AR duration and events were calculated as reported previously (Jiang et al. 2012). Because individual synaptic currents have more rapid rising slope than baseline fluctuations, current traces were first transformed into slope for detection of spontaneous event. Synaptic events were identified if their rising slopes were greater than a slope threshold. We then obtained the PT-AR frequency with a bin size of 50 ms within a time window of 2 s immediately after the AP train. Because stochastic synaptic events (EPSCs/IPSCs) from other presynaptic cells could also occur during recording, we subtracted the mean frequency of these spontaneous events occurred during a period of 4 s before the train stimulation from the frequency of PT-AR events to minimize the potential effects of spontaneous events on the count of AR events. To measure the PT-AR duration, we first identified the time of the last synaptic event after which PT-AR frequency decreased to the baseline frequency. The PT-AR duration was then measured as the time period from AP train to the last synaptic event. We also obtained the total number of PT-AR events occurred during this period. These parameters (PT-AR duration, frequency and total number) were used for measurement of AR strength in most of the figures (Figs 1, 2, 4, and 5 and Supplementary Fig. 4).

Figure 1.

Asynchronous release (AR) at FS autapses in juvenile prefrontal cortex. (A) Both current-clamp (left) and voltage-clamp recording (right) showed the occurrence of spontaneous AR events (arrowheads) after a train of APs in FS neurons. (B) Autaptic IPSC could be blocked by 50 μM PTX. Trace subtraction (control-PTX) reveals unitary IPSCs (bottom). (C) Group data show the dependence of PT-AR duration (top) and event number (bottom) on AP number and frequency.

Figure 1.

Asynchronous release (AR) at FS autapses in juvenile prefrontal cortex. (A) Both current-clamp (left) and voltage-clamp recording (right) showed the occurrence of spontaneous AR events (arrowheads) after a train of APs in FS neurons. (B) Autaptic IPSC could be blocked by 50 μM PTX. Trace subtraction (control-PTX) reveals unitary IPSCs (bottom). (C) Group data show the dependence of PT-AR duration (top) and event number (bottom) on AP number and frequency.

Figure 2.

Occurrence of AR at FS-PC synaptic connections in juvenile rats. (A) Example FS-PC pair recording. Note that the FS neuron received depressing EPSPs. Traces from FS (black) and PC (red) are color-coded. (B) AR occurred when the FS neuron fired a train of APs. Part of the PT-AR was expanded for clarity. Note the individual postsynaptic events (black arrows). (C) Example recording from an FS-PC pair showing the occurrence of AR in both FS autaptic (arrowheads) and FS-PC synaptic connections (arrows). (D) Group data shows the dependence of AR strength (PT-AR duration and event number) on the number and frequency of APs in presynaptic FS neurons. (E) Cumulative frequency distribution indicates no significant difference in AR strength between FS autaptic (black) and FS-PC synaptic connections (red). P = 0.28 and 0.06 for PT-AR duration and total events, respectively. (F) No significant difference in PT-AR frequency between the 2 types of connections. (G) Group data from FS-PC pairs that had both autaptic and synaptic connections. No difference was observed (P = 0.44 and 0.35 for PT-AR duration and total events, respectively). (H) Plot of the ratio of PT-AR duration (and event number) in FS-PC synaptic connections by that in autaptic connections. Student's t-test, P = 0.86 and 0.98 for duration and events, respectively.

Figure 2.

Occurrence of AR at FS-PC synaptic connections in juvenile rats. (A) Example FS-PC pair recording. Note that the FS neuron received depressing EPSPs. Traces from FS (black) and PC (red) are color-coded. (B) AR occurred when the FS neuron fired a train of APs. Part of the PT-AR was expanded for clarity. Note the individual postsynaptic events (black arrows). (C) Example recording from an FS-PC pair showing the occurrence of AR in both FS autaptic (arrowheads) and FS-PC synaptic connections (arrows). (D) Group data shows the dependence of AR strength (PT-AR duration and event number) on the number and frequency of APs in presynaptic FS neurons. (E) Cumulative frequency distribution indicates no significant difference in AR strength between FS autaptic (black) and FS-PC synaptic connections (red). P = 0.28 and 0.06 for PT-AR duration and total events, respectively. (F) No significant difference in PT-AR frequency between the 2 types of connections. (G) Group data from FS-PC pairs that had both autaptic and synaptic connections. No difference was observed (P = 0.44 and 0.35 for PT-AR duration and total events, respectively). (H) Plot of the ratio of PT-AR duration (and event number) in FS-PC synaptic connections by that in autaptic connections. Student's t-test, P = 0.86 and 0.98 for duration and events, respectively.

Deconvolution analysis was also used to evaluate the amount of synchronous and AR, as previously described (Hefft and Jonas 2005; Jiang et al. 2012). In brief, we first used an IPSC template to identify individual IPSCs occurred during the PT-AR period, and then chose the smallest IPSC and fitted its rising and decay phase with linear and exponential functions, respectively. This resultant trace was considered as quanta for following deconvolution: release rate = F−1[F (IPSC trace)/F (quanta)], in which F is the discrete Fourier transform (Hefft and Jonas 2005; Daw et al. 2009; Jiang et al. 2012). The release rate was then filtered (5–10 trials) using Gaussian-window FIR filter. The amount of release during (Train) and after the train (PT-AR) was measured as the integrated area of the release rate. Based on these analyses, the sensitivity of synchronous release and AR to EGTA was evaluated in Figure 3 and Supplementary Figure 1. In the EGTA-AM experiment (Fig. 3D), the progressive changes of Train and PT-AR were illustrated to show the time course of drug effect. Plots of quanta number (bin size: 5 ms) as a function of time was utilized to show the decay process of AR following the train. The time constant of this decay phase was estimated by exponential fitting, and the changes in time constant during drug application were plotted in Figure 3F. To minimize the contribution of residual GABA in the synaptic cleft to PT-AR, we used the decay time constant of unitary IPSC to predict the decay of summated train IPSCs and subtracted the train-IPSC decay component from the total post-train charge (Supplementary Fig. 3). The AP threshold was defined as the voltage when the derivation of Vm (dV/dt) reaches 20 V/s. The AP amplitude was measured as the voltage difference between AP threshold and its peak. The AP half-width was the duration measured at half amplitude. The AP area was the integrated voltage above the AP threshold.

Figure 3.

AR is Ca2+ dependent. (A) Bath application of EGTA-AM (100 μM) could block the occurrence of PT-AR in FS autaptic connections. Black, control; red, EGTA-AM. (B) Group data showing the time course of the effect of EGTA-AM on PT-AR in FS autaptic connections. (C and D) EGTA-AM blocked PT-AR in FS-PC pairs. Note that the peak amplitude of IPSC1 and the total release during the train (Train) were also reduced. (E) Plots of the number of quanta (bin size: 5 ms) as a function of time since the end of AP train at different time point (0, 2, 5, 10 min) after EGTA-AM application. (F) Plot of the time constant as a function of time after the drug application in FS-PC pairs (n = 8).

Figure 3.

AR is Ca2+ dependent. (A) Bath application of EGTA-AM (100 μM) could block the occurrence of PT-AR in FS autaptic connections. Black, control; red, EGTA-AM. (B) Group data showing the time course of the effect of EGTA-AM on PT-AR in FS autaptic connections. (C and D) EGTA-AM blocked PT-AR in FS-PC pairs. Note that the peak amplitude of IPSC1 and the total release during the train (Train) were also reduced. (E) Plots of the number of quanta (bin size: 5 ms) as a function of time since the end of AP train at different time point (0, 2, 5, 10 min) after EGTA-AM application. (F) Plot of the time constant as a function of time after the drug application in FS-PC pairs (n = 8).

Values are given as mean ± SEM and error bars in figures also indicate SEM. Significance of differences was assessed by student's t-test or paired t-test at the significance level (P) indicated. Two-sample Kolmogorov–Smirnov (K-S) test were performed if the data were not normally distributed. In order to examine the stimulation intensity-dependent properties of synchronous release and AR, 2-way ANOVA analysis was utilized for comparison.

Results

AR at FS Autapse in Juvenile Rats

We performed whole-cell recordings in prefrontal cortical slices from juvenile rats (P15–19) to examine the AR properties, which were compared later with those from adult rats (P50–55). Recordings were obtained from layer-5 FS-PC pairs or single FS neurons that could be identified by their nonadapting FS firing pattern and morphological properties (Jiang et al. 2012).

First, we examined the occurrence of AR in FS autaptic connections. In 160 of 574 (27.9%) FS neurons tested, we observed barrages of spontaneous events immediately after high-frequency discharges (Fig. 1A). These events were inhibitory postsynaptic responses (IPSPs or IPSCs) because they could be blocked by the bath application of 50 μM picrotoxin (PTX, n = 14). Individual events could be frequently distinguished from the outlasting IPSP (or IPSC) barrages. Consistent with previous studies (Bacci et al. 2003; Jiang et al. 2012), the tightly coupled IPSCs, which reflected the synchronous GABA release in autaptic connections, could be observed after single APs and blocked by PTX (Fig. 1B). We subtracted the current traces in the presence of PTX from those in control condition (control – PTX, Fig. 1B) and then measured the kinetics of individual autaptic IPSCs (Table 1).

Table 1

Synaptic kinetics in FS autapses and FS-PC synapses

IPSC kinetics FS autaptic IPSC
 
FS-PC IPSC
 
 Juvenile (n = 14) Adult (n = 12) Juvenile (n = 39) Adult (n = 24) 
Failure rate (%) 0 ± 0 0.2 ± 0.1 0.5 ± 0.3 0.1 ± 0.1 
Latency (ms) 1.30 ± 0.15 0.89 ± 0.07* 0.81 ± 0.04 0.82 ± 0.04 
SD of latency 0.12 ± 0.05 0.16 ± 0.03 0.21 ± 0.03 0.23 ± 0.03 
Amplitude (pA) 256 ± 40 178 ± 30 118 ± 16.9 61.4 ± 13.2* 
Conductance (nS) 4.0 ± 0.6 2.8 ± 0.5 1.8 ± 0.3 1.0 ± 0.2* 
CV of amplitude 0.25 ± 0.04 0.41 ± 0.04** 0.35 ± 0.03 0.50 ± 0.04** 
Rising (ms) 0.64 ± 0.08 0.43 ± 0.06* 1.06 ± 0.09 0.86 ± 0.04 
Decay (ms) 3.5 ± 0.4 2.0 ± 0.2** 9.9 ± 0.6 7.3 ± 0.3** 
PPR 1.08 ± 0.09 1.05 ± 0.05 1.04 ± 0.07 1.31 ± 0.13* 
IPSC kinetics FS autaptic IPSC
 
FS-PC IPSC
 
 Juvenile (n = 14) Adult (n = 12) Juvenile (n = 39) Adult (n = 24) 
Failure rate (%) 0 ± 0 0.2 ± 0.1 0.5 ± 0.3 0.1 ± 0.1 
Latency (ms) 1.30 ± 0.15 0.89 ± 0.07* 0.81 ± 0.04 0.82 ± 0.04 
SD of latency 0.12 ± 0.05 0.16 ± 0.03 0.21 ± 0.03 0.23 ± 0.03 
Amplitude (pA) 256 ± 40 178 ± 30 118 ± 16.9 61.4 ± 13.2* 
Conductance (nS) 4.0 ± 0.6 2.8 ± 0.5 1.8 ± 0.3 1.0 ± 0.2* 
CV of amplitude 0.25 ± 0.04 0.41 ± 0.04** 0.35 ± 0.03 0.50 ± 0.04** 
Rising (ms) 0.64 ± 0.08 0.43 ± 0.06* 1.06 ± 0.09 0.86 ± 0.04 
Decay (ms) 3.5 ± 0.4 2.0 ± 0.2** 9.9 ± 0.6 7.3 ± 0.3** 
PPR 1.08 ± 0.09 1.05 ± 0.05 1.04 ± 0.07 1.31 ± 0.13* 

Note: In order to accurately estimate the kinetics of autaptic IPSCs, we selectively isolate the autaptic IPSCs with obvious autaptic current after the action current; therefore, the measured average amplitude of autaptic IPSCs could be overestimated. An outlier in the group of juvenile FS-PC IPSC was not analyzed because its amplitude was larger than 1 nA. Conductances were calculated as IPSC amplitude/driving force. PPR was the paired pulse ratio (IPSC2/IPSC1, interstimulus interval: 20 ms). Statistical comparisons of these parameters were performed using 2-sample student's t-test. *P < 0.05, **P < 0.01.

To examine the dependence of AR strength on stimulation intensity, we varied the AP numbers and frequencies. Due to the difficulties in distinguishing AR events during the high-frequency AP train, we only examined the properties of post-train AR (PT-AR) in this study. As shown in Figure 1C, the PT-AR duration and total number of events progressively increased with increasing AP numbers or frequencies (P < 0.05 for both, 2-way ANOVA). Taken together, we observed robust AR in FS autapses in juvenile rats, similar to that in human cortex (Jiang et al. 2012).

AR in FS-PC Synapse in Juvenile Rats

We next investigated whether postsynaptic excitatory neurons receive AR from FS neurons. Among 451 FS-PC pairs tested, we found 103 FS-to-PC (22.8%) and 18 PC-to-FS (4.0%) connected pairs, and 12 (2.7%) bidirectionally connected pairs. In 98 of 121 FS-PC connections (81.0%), trains of APs in the presynaptic FS neurons could trigger both synchronous and asynchronous IPSCs in postsynaptic PCs (Fig. 2AC).

Further analysis demonstrated that the AR strength depended on the stimulation intensity. Group data from 37 FS-PC pairs showed that the PT-AR duration and total number of events increased with increasing number and frequency of APs (P < 0.001, 2-way ANOVA, Fig. 2D). Cumulative frequency distribution of FS autaptic and FS-PC synaptic connections showed no significant difference in PT-AR strength between the 2 types of connections (P = 0.28 and 0.06 for PT-AR duration and events, respectively, Fig. 2E). Similarly, the PT-AR frequency showed no significant difference either (Fig. 2F). Consistently, in 17 FS-PC pairs that exhibited both autaptic and FS-PC synaptic AR, no significant differences in the duration and total events of PT-AR were observed in the 2 types of connections (Fig. 2G,H). These results differ from a recent finding in human (5–69 years old) and adult rat (P52–62) cortical tissue that demonstrated substantially stronger AR in FS autapses as compared with FS-PC synapses (Jiang et al. 2012), possibly due to the difference in age periods. Together, our results indicate that, in juvenile rats, AR also occurred in FS-to-PC connections, and the AR strength was similar to that in FS autaptic connections.

Ca2+ Dependence of AR

Calcium accumulation in presynaptic terminals during AP trains is responsible for AR in hippocampus (Hefft and Jonas 2005) and neocortex (Manseau et al. 2010; Jiang et al. 2012); we therefore investigated the role of background or residual Ca2+ in mediating the occurrence of AR in prefrontal cortex of juvenile rats. In the presence of EGTA-AM (100 μM), a membrane permeable Ca2+ chelator, FS autaptic AR, and FS-PC synaptic AR were significantly reduced after the drug application (Fig. 3AD). Release rate was calculated by deconvolution as described in previous reports (Hefft and Jonas 2005; Jiang et al. 2012). The integrated areas of release rate during and after AP train were used to measure the quanta released during (Train) and after (PT-AR) AP train (also see Materials and Methods section). The PT-AR in FS autaptic connections was decreased to 19.0 ± 11.1% of control (P < 0.01, n = 5, Fig. 3A,B). In FS-PC synaptic connections, the peak amplitude of the first IPSC (IPSC1), the total release during the train (Train) and the PT-AR were decreased to 58.9 ± 9.6, 40.3 ± 6.9, and 9.4 ± 3.4% of control, respectively (P < 0.001, n = 8, Fig. 3C,D).

In order to characterize the time course of PT-AR, the release rate was binned (bin size: 5 ms) and fitted with a single exponential function. The time course of PT-AR was progressively shortened after the application of EGTA-AM (τ = 81.3, 70.0, 44.5, 30.5 ms at 0, 2, 5, 10 min after EGTA-AM application, Fig. 3E). The normalized group data also showed that the PT-AR decayed faster after EGTA-AM application (Fig. 3F).

As the actual concentration of EGTA inside the presynaptic terminals was unknown with EGTA-AM perfusion, we then performed experiments using internal solutions containing 2 mM EGTA (n = 8). Similar to the effect of bath EGTA-AM (Fig. 3C,D), the presence of EGTA in patch pipettes also substantially decreased the strength of both synchronous and asynchronous GABA release (Supplementary Fig. 1). These results were consistent with the finding that EGTA accelerates the initial decay phase of Ca2+ transients (Helmchen and Tank 2005). Together, our results showed that the occurrence of AR in juvenile prefrontal cortex was dependent on the residual Ca2+ level in presynaptic terminals after high-frequency firing.

Comparing AR Strength in Juvenile and Adult Rats

We next examined the difference of AR strength in juvenile and adult rats. In adult rats (P50–55), among 160 FS-PC pairs tested, we found 31 FS-to-PC (19.4%) and 4 PC-to-FS (6.7%) connected pairs, and 2 (1.3%) bidirectionally connected pairs. The kinetics of unitary IPSCs in adult rats were measured and compared with those from juvenile animals. As shown in Table 1, the CV and paired pulse ratio (PPR) of IPSC amplitude was significantly increased, while the peak amplitude and decay time constant of IPSCs were decreased in adult animals. The changes in CV and PPR reflected a reduced release probability in presynaptic vesicles (Sola et al. 2004; Ma and Prince 2012). The reduced synaptic strength might result from increased complexity of circuit wiring (Chattopadhyaya et al. 2004) and homeostatic regulation of synaptic transmission (Turrigiano and Nelson 2004). Although there was a reduction in unitary strength (first IPSCs), we found no significant change in peak amplitudes of 2nd–20th IPSCs evoked by APs at 200 Hz (Supplementary Fig. 2). Furthermore, we observed that the integrated charge of PT-AR was significantly decreased in adult tissue (P < 0.01), while the integrated charge of train IPSCs only showed a slight decrease at a significance level of P = 0.08 (Supplementary Fig. 3). These results indicate that the overall synaptic efficiency in response to multiple APs was not significantly changed, and the decrease in the unitary strength may result from the reduced release probability. The acceleration of decay time constant of IPSCs in adult was consistent with previous reports (Galarreta and Hestrin 2002; Doischer et al. 2008; Sauer and Bartos 2011), presumably resulting from changes in membrane properties (Doischer et al. 2008), expression of presynaptic Ca2+ channels (Iwasaki and Takahashi 1998; Iwasaki et al. 2000), and expression patterns of various GABAA receptors (Dunning et al. 1999; Hutcheon et al. 2000; Doischer et al. 2008).

Among 148 FS neurons tested, 53 (35.8%) showed autaptic AR; and 28 of 31 (90.3%) FS-PC pairs exhibited obvious synaptic AR in response to high-frequency firing in FS neurons (40 APs at 200 Hz). Consistent with recent findings in human and adult rat cortical tissues, the AR in FS autaptic connections was significantly stronger than that in FS-PC synaptic connections (P = 0.02 and 0.001 for PT-AR duration and events, respectively; K-S test; n = 33 FS neurons with autapses and 24 FS-PC pairs, Fig. 4A,B). However, this result was different from that in juvenile rats (Fig. 2), suggesting a differential developmental modification between the 2 types of connections.

Figure 4.

Comparison of AR strength in FS neurons of juvenile and adult rats. (A) Example recordings of AR in FS autaptic and FS-PC synaptic connections in juvenile and adult rats. (B) Cumulative frequency distribution of the tested FS autaptic and FS-PC synaptic connections by the PT-AR duration (top) and number of events (bottom) in response to 20 APs at 200 Hz. Juvenile: n = 63 FS neurons with autaptic connections and 37 FS-PC pairs with synaptic connections; Adult: n = 33 and 24, respectively. (C) Plots of the PT-AR frequency (bin size: 50 ms) as a function of time since the end of the train stimulation. Left, FS autaptic connections; right, FS-PC synaptic connections. Note that the PT-AR frequency in adult rats was significantly lower than that in juvenile rats. (D) Bar plot of the synaptic strength (the average peak amplitude of unitary FS-PC IPSCs) in FS-PC pairs. P = 0.03. (E) Normalized PT-AR duration by the synaptic strength from the same group of FS-PC pairs as shown in panel D. P = 0.02. *P < 0.05; **P < 0.01; ***P < 0.001.

Figure 4.

Comparison of AR strength in FS neurons of juvenile and adult rats. (A) Example recordings of AR in FS autaptic and FS-PC synaptic connections in juvenile and adult rats. (B) Cumulative frequency distribution of the tested FS autaptic and FS-PC synaptic connections by the PT-AR duration (top) and number of events (bottom) in response to 20 APs at 200 Hz. Juvenile: n = 63 FS neurons with autaptic connections and 37 FS-PC pairs with synaptic connections; Adult: n = 33 and 24, respectively. (C) Plots of the PT-AR frequency (bin size: 50 ms) as a function of time since the end of the train stimulation. Left, FS autaptic connections; right, FS-PC synaptic connections. Note that the PT-AR frequency in adult rats was significantly lower than that in juvenile rats. (D) Bar plot of the synaptic strength (the average peak amplitude of unitary FS-PC IPSCs) in FS-PC pairs. P = 0.03. (E) Normalized PT-AR duration by the synaptic strength from the same group of FS-PC pairs as shown in panel D. P = 0.02. *P < 0.05; **P < 0.01; ***P < 0.001.

The dependence of AR strength on AP number and frequency was preserved in adult animals (Fig. 4A). However, we found that AR strength in both FS autapses and FS-PC synapses in adult rats was dramatically decreased as compared with juvenile rats (Fig. 4A,B). For FS autaptic connections, the PT-AR duration and number of events in adults were 58.4 ± 5.9 ms and 4.0 ± 0.5, respectively, significantly less than those in juveniles (161 ± 11 ms and 11.9 ± 1.2; P < 0.001 for both parameters, K-S test). Similarly, for FS-PC synaptic connections, significantly weaker AR was observed in adults. The PT-AR duration and number of events in adults were 40.7 ± 5.9 ms and 2.1 ± 0.4, respectively, whereas in juveniles, these values were 141 ± 14 ms and 10.4 ± 1.9 (P < 0.001 for both parameters, K-S test). Similar results were obtained when the number of APs was increased to 40. The PT-AR frequency in FS autapses and FS-PC synapses also showed dramatic difference between juvenile and adult animals (Fig. 4C). To further narrow down the time window during which the developmental reduction of AR occurred, we performed similar recording from FS cells in an intermediate age group (P32–34). We found that the PT-AR frequency in this group of rats was significantly lower than that in juveniles (P15–19), but similar to that in adults (P50–55) (Fig. 5). These results suggest that a great reduction of AR occurred during an age period between P19 and P32, a critical period for the development of medial prefrontal cortex (Makinodan et al. 2012).

Figure 5.

Comparison of AR strength in FS autapses from 3 ages. (A) Example recordings showing FS autaptic AR from P15–19 (juvenile), P32–34 (Intermediate), and P50–55 (Adult) rats. (B) PT-AR frequencies showed significant difference between P32–34 and P15–19 rats (P < 0.001 within 200 ms after 20-AP trains and 50–200 ms after 40-AP trains), but no significant difference between P32–34 and P50–55 rats (except the data point of 100–150 ms after 40 APs). For 20-AP trains, plots for juvenile and adult animals (same as in Fig. 4C) are shown for comparison.

Figure 5.

Comparison of AR strength in FS autapses from 3 ages. (A) Example recordings showing FS autaptic AR from P15–19 (juvenile), P32–34 (Intermediate), and P50–55 (Adult) rats. (B) PT-AR frequencies showed significant difference between P32–34 and P15–19 rats (P < 0.001 within 200 ms after 20-AP trains and 50–200 ms after 40-AP trains), but no significant difference between P32–34 and P50–55 rats (except the data point of 100–150 ms after 40 APs). For 20-AP trains, plots for juvenile and adult animals (same as in Fig. 4C) are shown for comparison.

Previous studies demonstrated that the amount of AR correlates with the synaptic strength (Manseau et al. 2010; Jiang et al. 2012); therefore, AR differences between the 2 age periods may reflect changes in synaptic strength. Indeed, we found a significant reduction in the average amplitude of unitary IPSCs, that is, synaptic strength (Fig. 4D). However, when we normalized the PT-AR by synaptic strength, the AR differences between adults and juveniles were still observed (Fig. 4E). In addition, the PT-AR duration showed a weak but significant correlation with the synaptic strength in juvenile (r = 0.38, P = 0.03) but not in adult rats (r = 0.06, P = 0.78). The inconsistent correlation may result from developmental changes in presynaptic release machinery, including the size of readily releasable pool of vesicles, the expression pattern of Ca2+ channels and sensors.

To exclude the possibility that we sampled different subgroups of FS cells at different ages, we next performed recordings only from PV-expressing FS neurons in B13 PV-EGFP transgenic mice (Dumitriu et al. 2007). The strength of autaptic AR in these GFP-positive neurons was examined in juvenile (P15–21) and adult mice (P55–57). Similarly, we observed a dramatic decrease in PT-AR frequency (Supplementary Fig. 4), suggesting that the developmental AR reduction is a common phenomenon in FS cells, including PV-positive cells.

Together, we found a substantial reduction in AR in adult animals as compared with juveniles. As AR in FS output synapses provides long-lasting inhibition (Hefft and Jonas 2005) and regulates the probability and precision of spiking in postsynaptic cells (Manseau et al. 2010), the differences in AR strength in juvenile and adult FS neurons may reflect the differential contribution of asynchronous GABA release to the control of firing probability in their target cells and the distinct network dynamics at different age periods.

Differences in Presynaptic Ca2+ Dynamics

Previous reports showed that AR may share the same Ca2+ source with synchronous release, and require a high level of background Ca2+ after AP bursts (Hefft and Jonas 2005; Jiang et al. 2012). We therefore sought to investigate the differences in Ca2+ dynamics at presynaptic terminals in juvenile and adult animals.

We combined whole-cell recording and two-photon Ca2+ imaging in these experiments. We loaded the FS neurons with Ca2+ indicator Fluo-5F and fluorescent dye Alexa Fluor 594 through patch pipettes. The presynaptic boutons were identified according to previous reports (Koester and Sakmann 2000; Goldberg et al. 2005). Fluorescence signals at these boutons were acquired while AP trains (200 Hz) with varying numbers of APs (2–60) were elicited by somatic current injection (Fig. 6AC). The Ca2+ transients in adult rats showed substantially faster decay than those in juvenile rats (Fig. 6C). Group data (n = 61 boutons from 15 juvenile FS neurons and 73 boutons from 11 adult FS neurons) showed that the peak amplitude and the decay time constant progressively increased with increasing number of APs (Fig. 6D,E and Supplementary Fig. 5), indicating that the residual Ca2+ levels were elevated and prolonged. This could explain the dependence of AR strength on the number of presynaptic APs (Figs 1C, 2D, and 4A).

Figure 6.

Different kinetics of Ca2+ transients in FS terminals in juvenile and adult rats. (A) Calcium imaging from boutons of a FS neuron in a juvenile rat. (Aa) projection of the recorded FS neurons (red, Alexa Fluor 594). (Ab) the boxed area in (Aa). The dashed arrow indicates the path of laser scanning. Note the 2 boutons that were scanned. (Ac) Vm responses to step-current injections (−100 and 400 pA, 500 ms). (B) Similar to (A) but for a FS neuron in an adult rat. Arrowheads indicate the axon initial segment. (C) Calcium transients (Ca2+ indicator: Fluo-5F, 200 μM in patch pipette) in boutons indicated in (A) and (B). Top, raw data for 20 APs at 200 Hz. The dashed line indicates the onset of the AP train (20 APs, bottom). Middle, the corresponding Ca2+ signals. The exponential fits (red lines) and the decay time constants (τ) were indicated. Bottom, APs evoked by a train of brief current injections. (D) Plot of the peak amplitudes of Ca2+ transients (ΔG/R) in FS neurons from juvenile and adult rats. (E) Plot of the decay time constant of Ca2+ transients (P < 0.001 for all comparisons). (F) Distribution of the decay time constants of Ca2+ transients evoked by 2, 20, and 40 APs (black: juvenile; red: adult). Group data (DF) were obtained from 61 boutons of 15 FS neurons (juvenile) and 73 boutons from 11 FS neurons (adult). *P < 0.05; **P < 0.01; ***P < 0.001.

Figure 6.

Different kinetics of Ca2+ transients in FS terminals in juvenile and adult rats. (A) Calcium imaging from boutons of a FS neuron in a juvenile rat. (Aa) projection of the recorded FS neurons (red, Alexa Fluor 594). (Ab) the boxed area in (Aa). The dashed arrow indicates the path of laser scanning. Note the 2 boutons that were scanned. (Ac) Vm responses to step-current injections (−100 and 400 pA, 500 ms). (B) Similar to (A) but for a FS neuron in an adult rat. Arrowheads indicate the axon initial segment. (C) Calcium transients (Ca2+ indicator: Fluo-5F, 200 μM in patch pipette) in boutons indicated in (A) and (B). Top, raw data for 20 APs at 200 Hz. The dashed line indicates the onset of the AP train (20 APs, bottom). Middle, the corresponding Ca2+ signals. The exponential fits (red lines) and the decay time constants (τ) were indicated. Bottom, APs evoked by a train of brief current injections. (D) Plot of the peak amplitudes of Ca2+ transients (ΔG/R) in FS neurons from juvenile and adult rats. (E) Plot of the decay time constant of Ca2+ transients (P < 0.001 for all comparisons). (F) Distribution of the decay time constants of Ca2+ transients evoked by 2, 20, and 40 APs (black: juvenile; red: adult). Group data (DF) were obtained from 61 boutons of 15 FS neurons (juvenile) and 73 boutons from 11 FS neurons (adult). *P < 0.05; **P < 0.01; ***P < 0.001.

In comparison with juvenile rats, the peak amplitudes of Ca2+ transients were significantly greater (Fig. 6D) and the decay time constants were dramatically shorter in adult rats (P < 0.001 for all comparisons; Fig. 6E and Supplementary Fig. 5). The distribution of decay time constant also confirmed this dramatic change (Fig. 6F). The decay time constant of Ca2+ transients reflects the time needed for Ca2+ clearance; a longer time constant, therefore, indicates that the Ca2+ concentration remains high for a longer period of time. The slow Ca2+ dynamics in juvenile rats may lead to an elevated and prolonged level of background Ca2+ in presynaptic terminals of FS neurons and thus result in stronger asynchronous GABA release.

Taken together, our results indicate that AR occurs at all output synapses of FS neurons at the ages we examined, and interestingly the AR strength is more robust at juvenile ages but weaker in adults, possibly due to changes in the clearance of accumulated Ca2+ in presynaptic boutons. Considering that asynchronous GABA release causes long-lasting inhibition and decrease in spiking precision in postsynaptic neurons (Hefft and Jonas 2005; Manseau et al. 2010), we speculate that the occurrence of AR at neocortical inhibitory synapses and the developmental reduction of AR may have distinct functional impacts on neuronal signaling and contribute to cortical processing.

Discussion

Our results indicate that AR occurs at output synapses of FS neurons including FS autapses and FS-PC synapses, and continues to function in adulthood. Interestingly, we found a dramatic decrease in AR strength from juvenile to adult ages. Further Ca2+ imaging experiments suggested that this decrease may result from rapid Ca2+ kinetics in the presynaptic terminal after spiking. Stronger AR observed at developmental stages implicated that AR may play more important roles in shaping the neuronal and network activities during cortical development.

Developmental Changes of Synapses From FS Neurons

During early developmental stages, the excitatory and inhibitory synapses undergo continuous modification by the neuronal activity (Bear and Malenka 1994; Zucker and Regehr 2002; Dan and Poo 2004; Turrigiano and Nelson 2004) and sensory inputs (Yashiro et al. 2005; Tyler et al. 2007). Although most of studies were performed in young animals, it was believed that synapses could be modified across the whole lifespan (Creutzfeldt and Heggelund 1975). Because of the uniformity in morphology and microcircuits, the maturation and refinement of excitatory synapses between PCs were intensively studied (Hsia et al. 1998; Wasling et al. 2004; Sametsky et al. 2010). However, only a few studies focused on the development of inhibitory synapses (Huang et al. 2007). Previous studies (Chattopadhyaya et al. 2004; Doischer et al. 2008; Pangratz-Fuehrer and Hestrin 2011; Sauer and Bartos 2011) revealed dramatic changes in synchronous release from inhibitory synapses. FS neurons developed abundant neurites during the first 3 weeks (Chattopadhyaya et al. 2004; Doischer et al. 2008), leading to increased probability of synaptic connection in FS-FS pairs and FS-PC pairs (Doischer et al. 2008; Pangratz-Fuehrer and Hestrin 2011). The active and passive membrane properties also undergo substantial alterations, allowing faster transduction of electrical signals (Doischer et al. 2008; Pangratz-Fuehrer and Hestrin 2011). For synchronous release, the increase in synaptic strength and decrease in synaptic failure can ensure high precision and efficiency of synaptic transmission in FS output synapses (Doischer et al. 2008; Pangratz-Fuehrer and Hestrin 2011). However, it remains unclear whether AR is subjected to developmental changes.

In this study, we compared the strength of both synchronous release and AR from FS neurons between juvenile and adult brain. We found that both types of neurotransmitter release were significantly decreased in adult animals (Fig. 4AD). These changes may reflect the modification of synaptic properties during later development. A recent study showed a positive correlation between the synaptic strength and the AR strength (Jiang et al. 2012). However, here we only found a weak linear correlation between PT-AR duration and synaptic strength in FS-PC pairs in juvenile animals. Normalization of the PT-AR duration by synaptic strength in FS-PC pairs still revealed a significant difference between juvenile and adult animals (Fig. 4E), suggesting that the mechanism for alterations of AR may differ from that of synchronous release.

Previous studies showed that both synchronous release and AR depend on Ca2+ influx through presynaptic Ca2+ channels (Hefft and Jonas 2005). Prolonged elevation of Ca2+ concentration at presynaptic terminals is required for the occurrence of AR, possibly resulting from a long perfusion distance of Ca2+ from the location of activated channels to the active zone (Hefft and Jonas 2005; Hestrin and Galarreta 2005). The reduced AR in adult animals could be attributable to shortened perfusion distance. Indeed, the coupling between Ca2+ channels and synaptic vesicles is tighter in mature calyx of held (Wang et al. 2008). In hippocampus, pharmacological experiments revealed that P/Q-type Ca2+ channels were localized closely (nanodomain coupling) with Ca2+ sensors in a type of inhibitory synapse with weak AR, whereas N-type Ca2+ channels were localized distantly (microdomain coupling) from Ca2+ sensors in another type of inhibitory synapse with strong AR (Hefft and Jonas 2005; Eggermann et al. 2011). A longer diffusion distance of Ca2+ allows a buildup of Ca2+ concentration in the presynaptic terminal, and leads to slower decay phase of Ca2+ transients and thus stronger AR. In contrast, tight coupling between Ca2+ channels and sensors results in weaker AR. Therefore, the diffusion distance can determine the AR strength. Previous studies reported that the N-type Ca2+ channels at presynaptic terminals were gradually replaced by P/Q-type Ca2+ channels during development (Iwasaki and Takahashi 1998; Iwasaki et al. 2000). The increased contribution of P/Q-type Ca2+ channels and the associated nanodomain coupling between the Ca2+ source and sensor might cause a decrease in the AR strength in adult FS neurons. However, whether the Ca2+ diffusion distance is subjected to developmental changes remains to be further examined.

Mechanisms Underlying Developmental AR Reduction

In our experiments, we found that in comparison with juveniles the decay time constant of Ca2+ transients was significantly shorter in adult animals (Fig. 6C,E,F), consistent with the decrease in AR strength in adults. These results may reflect a change in Ca2+ diffusion distance at different age periods. Moreover, the reduction in the decay time constant of Ca2+ transients can result from changes in the efficiency of Ca2+ clearance (Yasuda et al. 2004), a process dependent on Ca2+ extrusion via membrane pumps (e.g., Ca2+-ATPase and Na+/Ca2+ exchanger) and Ca2+ uptake by presynaptic endoplasmic reticulum and mitochondria. Indeed, an upregulation of Ca2+ extrusion systems has been observed during development and neuronal maturation (Li et al. 2000; Kip et al. 2006). In addition to the briefer calcium transients in the presynaptic terminals, a reduction in the number of synapses and release probability may also contribute to the reduced AR strength in adult tissue. Analysis on the unitary IPSCs revealed a decrease in peak amplitude and an increase in CV (see Table 1), suggesting a reduction in number of functional synapses in adult tissue. This reduction may not result in shorter PT-AR duration because each synaptic contact may have similar time course of AR after the train of spikes; however, less synaptic contacts may cause a decrease in the frequency of PT-AR events, similar to the effect on spontaneous “mini” PSCs. We also observed an increase in CV and PPR in adult tissue when compared with juvenile tissue (see Table 1), suggesting a reduction in release probability. Previous findings in nucleus accumbens slices indicated that the synchronous release and AR may share the same readily releasable pool of vesicles in presynaptic terminals (Hjelmstad 2006; but see Hagler and Goda 2001 and Otsu et al. 2004 in cultured hippocampal cells). Lowering the release probability could substantially reduce AR to a degree similar to the change in overall release probability. The amount of AR correlated well with the size of initial response (unitary IPSC) during the spike train (Hjelmstad 2006). Similar correlation was observed in output synapses of FS neurons in neocortical slices (Manseau et al. 2010; Jiang et al. 2012). Together, the reduction in number of functional synapses and release probability may also contribute to the reduced AR strength in adult tissue.

Calcium buffers inside a neuron play important roles in regulating Ca2+ dynamics. Interneurons express various types of Ca2+ binding proteins (Markram et al. 2004), including parvalbumin (PV), calbindin and calretinin. Previous findings showed that the decay kinetics of Ca2+ transients in cerebellar stellate cells in P21 mice are faster than that in P11 mice, and this change is accompanied with an increase in the proportion of PV-positive cells (Collin et al. 2005), indicating a role of endogenous PV in regulating Ca2+ dynamics. Interestingly, this study also suggested that PV might accelerate the initial decay phase but cause an additional slow phase in Ca2+ transients in response to AP trains, thus enhancing AR in inhibitory synapses. However, a recent finding demonstrated an opposite effect of PV on AR, by showing that PV-negative cells exhibited stronger AR (Manseau et al. 2010). These inconsistent results may result from differences in PV concentration in individual cells (Eggermann and Jonas 2012). Whether alterations in the expression level of Ca2+-binding proteins could play a role in regulating the AR strength requires further examination.

The reduced AR strength in adults could also result from changes in the expression pattern of various Ca2+ sensors. Manipulating the expression level of various Ca2+ sensors could differentially regulate the strength of synchronous release and AR (Sun et al. 2007; Yao et al. 2011). Importantly, the expression levels of synaptotagmin isoforms undergo changes during development (Berton et al. 1997), possibly contributing to the reduced AR strength in adult animals. In addition, developmental alteration of ion channel expression can lead to changes in AP waveforms, which may determine the amount of Ca2+ influx during APs and regulate synaptic transmission.(Geiger and Jonas 2000; Shu et al. 2006). Here, we also observed a developmental alteration of AP waveform, that is, decrease in AP half-width and integrated voltage area (Supplementary Fig. 6), suggesting a role of AP waveform in regulating presynaptic Ca2+ dynamics and asynchronous GABA release.

Physiological Significance

In early developmental stages, GABAergic transmission plays important roles in regulating synapse formation and structural refinement of neural network (Ben-Ari 2002). For instance, reduction of GABAergic transmission could significantly delay the critical period of ocular dominance (Hensch et al. 1998; Hensch 2005), which peaks at third postnatal week and ends by fifth week. Accelerated GABAergic innervation and inhibition caused early termination of the critical period for ocular dominance plasticity (Huang et al. 1999). More investigations suggest that different inhibitory circuits contribute differentially to this cortical plasticity, by showing a crucial role of perisomatic inhibition during the critical period (Fagiolini et al. 2004; Katagiri et al. 2007). The AR changes in prefrontal cortex during development may also contribute to the regulation of neuronal wiring and the development of functional circuits. Interestingly, our results indicate that the greatest reduction of AR strength occurred between P20 and P32, the critical period for oligodendrocyte maturation in prefrontal cortex and related behaviors including sociability and working memory (Makinodan et al. 2012). However, it remains unclear whether asynchronous GABA release at critical period has impacts on cortical plasticity.

FS neurons form intensive electrical and chemical connections with other FS neurons, and also innervate a large population of PCs (Fino and Yuste 2011; Packer and Yuste 2011; Pangratz-Fuehrer and Hestrin 2011). Because of this high connectivity with neighboring cells, FS neurons can synchronize the cortical network and generate rhythmic oscillations (Tamas et al. 2000; Freund 2003; Bartos et al. 2007). Manipulating the activity of FS neurons could selectively regulate the gamma oscillations in the barrel cortex (Cardin et al. 2009) and alter the theta phase of spikes in hippocampal place cells (Royer et al. 2012). As shown previously (Miles et al. 1996), precise neurotransmitter release from FS neurons determines the spike timing of their target cells; however, prolonged asynchronous GABA release from FS neurons could decrease the probability and precision of postsynaptic spikes, and thus desynchronize the neuronal network (Manseau et al. 2010). Previous modeling studies (Voegtlin and Martinez 2007; Volman et al. 2011) showed that the synchronization of neural network was decreased when the inhibitory transmission became asynchronous. Taken together, we speculate that the reduction of AR in adult animals may promote the synchronization of cortical network at appropriate frequencies and contribute to the proper functioning of the cortex; changes in AR strength may cause an alteration of rhythmic oscillations and consequently affect the behavioral performance. We reported recently that the AR strength is substantially enhanced in both human and rat epileptic neocortical tissues (Jiang et al. 2012). This upregulation of AR may play a role in regulating epileptiform activity but, on the other hand, it can disrupt normal cortical oscillations that are critical for cognitive functions. The reduction of AR in adult animals is therefore important for normal brain functioning; in contrast, stronger AR at early developmental stages may prevent hypersynchronization of the cortical network.

It has been reported that the neuronal network showed spontaneous or sensory-driven oscillations in anesthetized animals (Ylinen et al. 1995; Jones et al. 2000) and epileptic human patients (Staba et al. 2002). These oscillations could reach a frequency of 100–600 Hz; importantly, FS cells can generate APs with frequencies up to 500 Hz and follow these fast oscillations (Ylinen et al. 1995; Jones et al. 2000), indicating a role of FS cells in shaping these high-frequency oscillations. Considering that AR could be evoked when FS cells discharge at frequencies larger than 100 Hz, we speculate that under certain physiological and pathological conditions asynchronous GABA release may regulate the firing behavior of their postsynaptic targets and network function.

Taken together, we demonstrated a developmental decrease in AR strength of FS neurons from juvenile to adult rats, and this reduction could result from the acceleration of Ca2+ decay phase in adult FS neurons. Given the key role of GABAergic inhibition induced by FS neurons in regulating circuit maturation and neuronal synchronization, the reduction in asynchronous GABA release from FS neurons during development may differentially modulate the functional organization and performance of neural networks at distinct behavioral states.

Supplementary Material

Supplementary material can be found at: http://www.cercor.oxfordjournals.org/.

Funding

This work was supported by the 973 Program (2011CBA00400), the National Natural Science Foundation of China Project (31025012), the Hundreds of Talents Program, the Knowledge Innovation Project from Chinese Academy of Sciences (KSCX2-YW-R-102), and SA-SIBS Scholarship Program.

Notes

We thank Z. Josh Huang for providing B13 PV-EGFP transgenic mice. Conflict of Interest: None declared.

References

Bacci
A
Huguenard
JR
Prince
DA
Functional autaptic neurotransmission in fast-spiking interneurons: a novel form of feedback inhibition in the neocortex
J Neurosci
 , 
2003
, vol. 
23
 (pg. 
859
-
866
)
Bartos
M
Vida
I
Jonas
P
Synaptic mechanisms of synchronized gamma oscillations in inhibitory interneuron networks
Nat Rev Neurosci
 , 
2007
, vol. 
8
 (pg. 
45
-
56
)
Bear
MF
Malenka
RC
Synaptic plasticity: LTP and LTD
Curr Opin Neurobiol
 , 
1994
, vol. 
4
 (pg. 
389
-
399
)
Ben-Ari
Y
Excitatory actions of gaba during development: the nature of the nurture
Nat Rev Neurosci
 , 
2002
, vol. 
3
 (pg. 
728
-
739
)
Berton
F
Iborra
C
Boudier
JA
Seagar
MJ
Marqueze
B
Developmental regulation of synaptotagmin I, II, III, and IV mRNAs in the rat CNS
J Neurosci
 , 
1997
, vol. 
17
 (pg. 
1206
-
1216
)
Cardin
JA
Carlen
M
Meletis
K
Knoblich
U
Zhang
F
Deisseroth
K
Tsai
LH
Moore
CI
Driving fast-spiking cells induces gamma rhythm and controls sensory responses
Nature
 , 
2009
, vol. 
459
 (pg. 
663
-
667
)
Chattopadhyaya
B
Di Cristo
G
Higashiyama
H
Knott
GW
Kuhlman
SJ
Welker
E
Huang
ZJ
Experience and activity-dependent maturation of perisomatic GABAergic innervation in primary visual cortex during a postnatal critical period
J Neurosci
 , 
2004
, vol. 
24
 (pg. 
9598
-
9611
)
Collin
T
Chat
M
Lucas
MG
Moreno
H
Racay
P
Schwaller
B
Marty
A
Llano
I
Developmental changes in parvalbumin regulate presynaptic Ca2+ signaling
J Neurosci
 , 
2005
, vol. 
25
 (pg. 
96
-
107
)
Creutzfeldt
OD
Heggelund
P
Neural plasticity in visual cortex of adult cats after exposure to visual patterns
Science
 , 
1975
, vol. 
188
 (pg. 
1025
-
1027
)
Dan
Y
Poo
MM
Spike timing-dependent plasticity of neural circuits
Neuron
 , 
2004
, vol. 
44
 (pg. 
23
-
30
)
Daw
MI
Tricoire
L
Erdelyi
F
Szabo
G
McBain
CJ
Asynchronous transmitter release from cholecystokinin-containing inhibitory interneurons is widespread and target-cell independent
J Neurosci
 , 
2009
, vol. 
29
 (pg. 
11112
-
11122
)
Doischer
D
Hosp
JA
Yanagawa
Y
Obata
K
Jonas
P
Vida
I
Bartos
M
Postnatal differentiation of basket cells from slow to fast signaling devices
J Neurosci
 , 
2008
, vol. 
28
 (pg. 
12956
-
12968
)
Dumitriu
D
Cossart
R
Huang
J
Yuste
R
Correlation between axonal morphologies and synaptic input kinetics of interneurons from mouse visual cortex
Cereb Cortex
 , 
2007
, vol. 
17
 (pg. 
81
-
91
)
Dunning
DD
Hoover
CL
Soltesz
I
Smith
MA
O'Dowd
DK
GABAA receptor-mediated miniature postsynaptic currents and alpha-subunit expression in developing cortical neurons
J Neurophysiol
 , 
1999
, vol. 
82
 (pg. 
3286
-
3297
)
Eggermann
E
Bucurenciu
I
Goswami
SP
Jonas
P
Nanodomain coupling between Ca2+ channels and sensors of exocytosis at fast mammalian synapses
Nat Rev Neurosci
 , 
2011
, vol. 
13
 (pg. 
7
-
21
)
Eggermann
E
Jonas
P
How the “slow” Ca2+ buffer parvalbumin affects transmitter release in nanodomain-coupling regimes
Nat Neurosci
 , 
2012
, vol. 
15
 (pg. 
20
-
22
)
Fagiolini
M
Fritschy
JM
Low
K
Mohler
H
Rudolph
U
Hensch
TK
Specific GABAA circuits for visual cortical plasticity
Science
 , 
2004
, vol. 
303
 (pg. 
1681
-
1683
)
Fino
E
Yuste
R
Dense inhibitory connectivity in neocortex
Neuron
 , 
2011
, vol. 
69
 (pg. 
1188
-
1203
)
Freund
TF
Interneuron diversity series: Rhythm and mood in perisomatic inhibition
Trends Neurosci
 , 
2003
, vol. 
26
 (pg. 
489
-
495
)
Fuchs
EC
Zivkovic
AR
Cunningham
MO
Middleton
S
Lebeau
FE
Bannerman
DM
Rozov
A
Whittington
MA
Traub
RD
Rawlins
JN
, et al.  . 
Recruitment of parvalbumin-positive interneurons determines hippocampal function and associated behavior
Neuron
 , 
2007
, vol. 
53
 (pg. 
591
-
604
)
Galarreta
M
Hestrin
S
Electrical and chemical synapses among parvalbumin fast-spiking GABAergic interneurons in adult mouse neocortex
P Natl Acad Sci USA
 , 
2002
, vol. 
99
 (pg. 
12438
-
12443
)
Geiger
JR
Jonas
P
Dynamic control of presynaptic Ca2+ inflow by fast-inactivating K+ channels in hippocampal mossy fiber boutons
Neuron
 , 
2000
, vol. 
28
 (pg. 
927
-
939
)
Glickfeld
LL
Roberts
JD
Somogyi
P
Scanziani
M
Interneurons hyperpolarize pyramidal cells along their entire somatodendritic axis
Nat Neurosci
 , 
2009
, vol. 
12
 (pg. 
21
-
23
)
Goldberg
EM
Watanabe
S
Chang
SY
Joho
RH
Huang
ZJ
Leonard
CS
Rudy
B
Specific functions of synaptically localized potassium channels in synaptic transmission at the neocortical GABAergic fast-spiking cell synapse
J Neurosci
 , 
2005
, vol. 
25
 (pg. 
5230
-
5235
)
Hagler
DJ
Jr
Goda
Y
Properties of synchronous and asynchronous release during pulse train depression in cultured hippocampal neurons
J Neurophysiol
 , 
2001
, vol. 
85
 (pg. 
2324
-
2334
)
Hefft
S
Jonas
P
Asynchronous GABA release generates long-lasting inhibition at a hippocampal interneuron-principal neuron synapse
Nat Neurosci
 , 
2005
, vol. 
8
 (pg. 
1319
-
1328
)
Helmchen
F
Tank
DW
Yuste
R
Konnerth
A
A single-compartment model of calcium dynamics in nerve terminals and dendrites
Imaging in neuroscience and development
 , 
2005
1st ed
New York, NY
Cold Spring Harbor Laboratory Press
(pg. 
265
-
275
)
Hensch
TK
Critical period plasticity in local cortical circuits
Nat Rev Neurosci
 , 
2005
, vol. 
6
 (pg. 
877
-
888
)
Hensch
TK
Fagiolini
M
Mataga
N
Stryker
MP
Baekkeskov
S
Kash
SF
Local GABA circuit control of experience-dependent plasticity in developing visual cortex
Science
 , 
1998
, vol. 
282
 (pg. 
1504
-
1508
)
Hestrin
S
Galarreta
M
Synchronous versus asynchronous transmitter release: a tale of two types of inhibitory neurons
Nat Neurosci
 , 
2005
, vol. 
8
 (pg. 
1283
-
1284
)
Hjelmstad
GO
Interactions between asynchronous release and short-term plasticity in the nucleus accumbens slice
J Neurophysiol
 , 
2006
, vol. 
95
 (pg. 
2020
-
2023
)
Hsia
AY
Malenka
RC
Nicoll
RA
Development of excitatory circuitry in the hippocampus
J Neurophysiol
 , 
1998
, vol. 
79
 (pg. 
2013
-
2024
)
Huang
ZJ
Di Cristo
G
Ango
F
Development of GABA innervation in the cerebral and cerebellar cortices
Nat Rev Neurosci
 , 
2007
, vol. 
8
 (pg. 
673
-
686
)
Huang
ZJ
Kirkwood
A
Pizzorusso
T
Porciatti
V
Morales
B
Bear
MF
Maffei
L
Tonegawa
S
BDNF regulates the maturation of inhibition and the critical period of plasticity in mouse visual cortex
Cell
 , 
1999
, vol. 
98
 (pg. 
739
-
755
)
Hutcheon
B
Morley
P
Poulter
MO
Developmental change in GABAA receptor desensitization kinetics and its role in synapse function in rat cortical neurons
J Physiol
 , 
2000
, vol. 
522
 
Pt 1
(pg. 
3
-
17
)
Iwasaki
S
Momiyama
A
Uchitel
OD
Takahashi
T
Developmental changes in calcium channel types mediating central synaptic transmission
J Neurosci
 , 
2000
, vol. 
20
 (pg. 
59
-
65
)
Iwasaki
S
Takahashi
T
Developmental changes in calcium channel types mediating synaptic transmission in rat auditory brainstem
J Physiol
 , 
1998
, vol. 
509
 
Pt 2
(pg. 
419
-
423
)
Jiang
M
Zhu
J
Liu
Y
Yang
M
Tian
C
Jiang
S
Wang
Y
Guo
H
Wang
K
Shu
Y
Enhancement of asynchronous release from fast-spiking interneuron in human and rat epileptic neocortex
PLoS Biol
 , 
2012
, vol. 
10
 
5
pg. 
e1001324
 
Jones
MS
MacDonald
KD
Choi
B
Dudek
FE
Barth
DS
Intracellular correlates of fast (>200 Hz) electrical oscillations in rat somatosensory cortex
J Neurophysiol
 , 
2000
, vol. 
84
 (pg. 
1505
-
1518
)
Katagiri
H
Fagiolini
M
Hensch
TK
Optimization of somatic inhibition at critical period onset in mouse visual cortex
Neuron
 , 
2007
, vol. 
53
 (pg. 
805
-
812
)
Kawaguchi
Y
Kubota
Y
GABAergic cell subtypes and their synaptic connections in rat frontal cortex
Cereb Cortex
 , 
1997
, vol. 
7
 (pg. 
476
-
486
)
Kip
SN
Gray
NW
Burette
A
Canbay
A
Weinberg
RJ
Strehler
EE
Changes in the expression of plasma membrane calcium extrusion systems during the maturation of hippocampal neurons
Hippocampus
 , 
2006
, vol. 
16
 (pg. 
20
-
34
)
Koester
HJ
Sakmann
B
Calcium dynamics associated with action potentials in single nerve terminals of pyramidal cells in layer 2/3 of the young rat neocortex
J Physiol
 , 
2000
, vol. 
529
 
Pt 3
(pg. 
625
-
646
)
Lee
SH
Kwan
AC
Zhang
S
Phoumthipphavong
V
Flannery
JG
Masmanidis
SC
Taniguchi
H
Huang
ZJ
Zhang
F
Boyden
ES
, et al.  . 
Activation of specific interneurons improves V1 feature selectivity and visual perception
Nature
 , 
2012
, vol. 
488
 (pg. 
379
-
383
)
Li
L
Guerini
D
Carafoli
E
Calcineurin controls the transcription of Na+/Ca2+ exchanger isoforms in developing cerebellar neurons
J Biol Chem
 , 
2000
, vol. 
275
 (pg. 
20903
-
20910
)
Ma
Y
Prince
DA
Functional alterations in GABAergic fast-spiking interneurons in chronically injured epileptogenic neocortex
Neurobiol Dis
 , 
2012
, vol. 
47
 (pg. 
102
-
113
)
Makinodan
M
Rosen
KM
Ito
S
Corfas
G
A critical period for social experience-dependent oligodendrocyte maturation and myelination
Science
 , 
2012
, vol. 
337
 (pg. 
1357
-
1360
)
Manseau
F
Marinelli
S
Mendez
P
Schwaller
B
Prince
DA
Huguenard
JR
Bacci
A
Desynchronization of neocortical networks by asynchronous release of GABA at autaptic and synaptic contacts from fast-spiking interneurons
PLoS Biol
 , 
2010
, vol. 
8
 
9
pg. 
e1000492
 
Markram
H
Toledo-Rodriguez
M
Wang
Y
Gupta
A
Silberberg
G
Wu
C
Interneurons of the neocortical inhibitory system
Nat Rev Neurosci
 , 
2004
, vol. 
5
 (pg. 
793
-
807
)
Miles
R
Toth
K
Gulyas
AI
Hajos
N
Freund
TF
Differences between somatic and dendritic inhibition in the hippocampus
Neuron
 , 
1996
, vol. 
16
 (pg. 
815
-
823
)
Otsu
Y
Shahrezaei
V
Li
B
Raymond
LA
Delaney
KR
Murphy
TH
Competition between phasic and asynchronous release for recovered synaptic vesicles at developing hippocampal autaptic synapses
J Neurosci
 , 
2004
, vol. 
24
 (pg. 
420
-
433
)
Packer
AM
Yuste
R
Dense, unspecific connectivity of neocortical parvalbumin-positive interneurons: a canonical microcircuit for inhibition?
J Neurosci
 , 
2011
, vol. 
31
 (pg. 
13260
-
13271
)
Pangratz-Fuehrer
S
Hestrin
S
Synaptogenesis of electrical and GABAergic synapses of fast-spiking inhibitory neurons in the neocortex
J Neurosci
 , 
2011
, vol. 
31
 (pg. 
10767
-
10775
)
Royer
S
Zemelman
BV
Losonczy
A
Kim
J
Chance
F
Magee
JC
Buzsaki
G
Control of timing, rate and bursts of hippocampal place cells by dendritic and somatic inhibition
Nat Neurosci
 , 
2012
, vol. 
15
 (pg. 
769
-
775
)
Sabatini
BL
Regehr
WG
Timing of synaptic transmission
Annu Rev Physiol
 , 
1999
, vol. 
61
 (pg. 
521
-
542
)
Sametsky
EA
Disterhoft
JF
Geinisman
Y
Nicholson
DA
Synaptic strength and postsynaptically silent synapses through advanced aging in rat hippocampal CA1 pyramidal neurons
Neurobiol Aging
 , 
2010
, vol. 
31
 (pg. 
813
-
825
)
Sauer
JF
Bartos
M
Postnatal differentiation of cortical interneuron signalling
Eur J Neurosci
 , 
2011
, vol. 
34
 (pg. 
1687
-
1696
)
Shu
Y
Hasenstaub
A
Duque
A
Yu
Y
McCormick
DA
Modulation of intracortical synaptic potentials by presynaptic somatic membrane potential
Nature
 , 
2006
, vol. 
441
 (pg. 
761
-
765
)
Sohal
VS
Zhang
F
Yizhar
O
Deisseroth
K
Parvalbumin neurons and gamma rhythms enhance cortical circuit performance
Nature
 , 
2009
, vol. 
459
 (pg. 
698
-
702
)
Sola
E
Prestori
F
Rossi
P
Taglietti
V
D'Angelo
E
Increased neurotransmitter release during long-term potentiation at mossy fibre-granule cell synapses in rat cerebellum
J Physiol
 , 
2004
, vol. 
557
 (pg. 
843
-
861
)
Staba
RJ
Wilson
CL
Bragin
A
Fried
I
Engel
J
Jr
Quantitative analysis of high-frequency oscillations (80–500 Hz) recorded in human epileptic hippocampus and entorhinal cortex
J Neurophysiol
 , 
2002
, vol. 
88
 (pg. 
1743
-
1752
)
Sun
J
Pang
ZP
Qin
D
Fahim
AT
Adachi
R
Sudhof
TC
A dual-Ca2+-sensor model for neurotransmitter release in a central synapse
Nature
 , 
2007
, vol. 
450
 (pg. 
676
-
682
)
Tamas
G
Buhl
EH
Lorincz
A
Somogyi
P
Proximally targeted GABAergic synapses and gap junctions synchronize cortical interneurons
Nat Neurosci
 , 
2000
, vol. 
3
 (pg. 
366
-
371
)
Turrigiano
GG
Nelson
SB
Homeostatic plasticity in the developing nervous system
Nat Rev Neurosci
 , 
2004
, vol. 
5
 (pg. 
97
-
107
)
Tyler
WJ
Petzold
GC
Pal
SK
Murthy
VN
Experience-dependent modification of primary sensory synapses in the mammalian olfactory bulb
J Neurosci
 , 
2007
, vol. 
27
 (pg. 
9427
-
9438
)
Voegtlin
T
Martinez
D
Effect of asynchronous GABA release on the oscillatory dynamics of inhibitory coupled neurons
Neurocomputing
 , 
2007
, vol. 
70
 (pg. 
2079
-
2084
)
Volman
V
Behrens
MM
Sejnowski
TJ
Downregulation of parvalbumin at cortical GABA synapses reduces network gamma oscillatory activity
J Neurosci
 , 
2011
, vol. 
31
 (pg. 
18137
-
18148
)
Wang
LY
Neher
E
Taschenberger
H
Synaptic vesicles in mature calyx of held synapses sense higher nanodomain calcium concentrations during action potential-evoked glutamate release
J Neurosci
 , 
2008
, vol. 
28
 (pg. 
14450
-
14458
)
Wasling
P
Hanse
E
Gustafsson
B
Developmental changes in release properties of the CA3-CA1 glutamate synapse in rat hippocampus
J Neurophysiol
 , 
2004
, vol. 
92
 (pg. 
2714
-
2724
)
Yao
J
Gaffaney
JD
Kwon
SE
Chapman
ER
Doc2 is a Ca2+ sensor required for asynchronous neurotransmitter release
Cell
 , 
2011
, vol. 
147
 (pg. 
666
-
677
)
Yashiro
K
Corlew
R
Philpot
BD
Visual deprivation modifies both presynaptic glutamate release and the composition of perisynaptic/extrasynaptic NMDA receptors in adult visual cortex
J Neurosci
 , 
2005
, vol. 
25
 (pg. 
11684
-
11692
)
Yasuda
R
Nimchinsky
EA
Scheuss
V
Pologruto
TA
Oertner
TG
Sabatini
BL
Svoboda
K
Imaging calcium concentration dynamics in small neuronal compartments
Sci STKE
 , 
2004
, vol. 
2004
 pg. 
pl5
 
Ylinen
A
Bragin
A
Nadasdy
Z
Jando
G
Szabo
I
Sik
A
Buzsaki
G
Sharp wave-associated high-frequency oscillation (200 Hz) in the intact hippocampus: network and intracellular mechanisms
J Neurosci
 , 
1995
, vol. 
15
 (pg. 
30
-
46
)
Zucker
RS
Regehr
WG
Short-term synaptic plasticity
Annu Rev Physiol
 , 
2002
, vol. 
64
 (pg. 
355
-
405
)