Optogenetic Examination of Prefrontal-Amygdala Synaptic Development

Abstract

A brain network comprising the medial prefrontal cortex (mPFC) and amygdala plays important roles in developmentally regulated cognitive and emotional processes. However, very little is known about the maturation of mPFC-amygdala circuitry. We conducted anatomical tracing of mPFC projections and optogenetic interrogation of their synaptic connections with neurons in the basolateral amygdala (BLA) at neonatal to adult developmental stages in mice. Results indicate that mPFC-BLA projections exhibit delayed emergence relative to other mPFC pathways and establish synaptic transmission with BLA excitatory and inhibitory neurons in late infancy, events that coincide with a massive increase in overall synaptic drive. During subsequent adolescence, mPFC-BLA circuits are further modified by excitatory synaptic strengthening as well as a transient surge in feedforward inhibition. The latter was correlated with increased spontaneous inhibitory currents in excitatory neurons, suggesting that mPFC-BLA circuit maturation culminates in a period of exuberant GABAergic transmission. These findings establish a time course for the onset and refinement of mPFC-BLA transmission and point to potential sensitive periods in the development of this critical network.SIGNIFICANCE STATEMENT Human mPFC-amygdala functional connectivity is developmentally regulated and figures prominently in numerous psychiatric disorders with a high incidence of adolescent onset. However, it remains unclear when synaptic connections between these structures emerge or how their properties change with age. Our work establishes developmental windows and cellular substrates for synapse maturation in this pathway involving both excitatory and inhibitory circuits. The engagement of these substrates by early life experience may support the ontogeny of fundamental behaviors but could also lead to inappropriate circuit refinement and psychopathology in adverse situations.

Keywords: adolescence; amygdala; prefrontal development.

Figure 1.

Volume estimation of prefrontal channelrhodopsin expression. A, Representative confocal images of prefrontal sections containing target site at postnatal age of death. Subjects received stereotaxic injections of AAV₁-CaMKII-ChR2(H134R)-EYFP 7 d before perfusion and brain sectioning. Sections were stained with Hoechst nuclear dye before imaging EYFP intrinsic fluorescence. Scale bar, 500 μm. Cavalieri volume estimation was based on the area of EYFP and mPFC (B) contained in every fifth section at 50 μm thickness as described in Materials and Methods. mPFC was defined as Cg1, PL, IL, and DP extending from the caudal edge of the main olfactory bulb to the genu of corpus callosum. As expected, both EYFP and mPFC volume exhibited age-dependent increase. Tukey's post hoc test was conducted on all pairwise comparisons following one-way ANOVA. *p < 0.05. C, Percentage infection of each mPFC subregion was calculated by dividing EYFP volume by the volume of the structure in which it was contained. P10, n = 9 mice; P15, n = 11 mice; P21, n = 11 mice; P30, n = 11 mice; P45, n = 10 mice; P60, n = 9 mice.

Figure 2.

Delayed emergence of mPFC-BLA projections. Representative images containing EYFP-positive projections were collected by confocal scanning microscopy. At P10, mPFC projections were observed in caudate-putamen (CPu) (A), claustrum (Cl) (A, B), mediodorsal (MD), and ventromedial (VM) thalamus (B, C), and nucleus reuniens (Re) (C), but BLA fluorescence was indiscernible in the same animals (B, C). At P15, EYFP-positive projections could be seen in the BLA in addition to the above structures (D–F). *Previously identified mPFC fiber tract (Xu and Südhof, 2013). Scale bar, 400 μm.

Figure 3.

Age-dependent proliferation of mPFC axons in the BLA. A–F, High-power confocal images were obtained across ages to confirm the presence of EYFP-positive fibers. Within the BLA, fluorescence was mostly restricted to the anterior basal amygdala (BA), and was hardly discernible in the lateral (LA), central, and basomedial nuclei at any age. Immunofluorescent labeling of PSD95 (red) is used strictly as contrast. Scale bars: Top, 400 μm; Bottom, 50 μm. BLA was almost entirely devoid of EYFP-positive fibers at P10 (A). However, by P15, mPFC projections had densely ramified within the anterior BA (B). G, A comparison of integrated fluorescence intensity for all animals confirmed that EYFP labeling had an abrupt onset between P10 and P15 and continued to increase until up to P30 in the anterior BA. Background fluorescence was measured in the LA and subtracted from BA fluorescence for each brain section. Tukey's post hoc test was conducted on all pairwise comparisons following one-way ANOVA. *p < 0.05. ***p < 0.001. Analysis of mPFC infection spread and BLA fluorescence was performed on the same animals, with n values reported in Figure 1.

Figure 4.

Development of spontaneous synaptic transmission in the anterior basal nucleus. A, sEPSCs and sIPSCs were obtained from principal excitatory neurons in the anterior basal nucleus of the BLA in acute brain slices. To enable within-cell comparisons of excitatory and inhibitory transmission, sEPSCs and sIPSCs were sampled from the same neuron by clamping the cell at reversal potential for IPSCs (−60 mV) or EPSCs (0 mV), respectively, in a low-chloride internal solution. B, C, Increase in the frequency (B) but not amplitude (C) of sEPSCs between P10 and later ages. Tukey's post hoc test was conducted on all pairwise comparisons following one-way ANOVA. ^#p < 0.10. *p < 0.05. **p < 0.01. P10, n = 14 cells (2 mice); P15, n = 15 cells (7 mice); P21, n = 9 cells (4 mice); P30, n = 13 cells (8 mice); P45, n = 8 cells (4 mice); P60, n = 8 cells (6 mice). D, E, Increase in the frequency (D) and amplitude (E) between P10 and later ages, as well as further increase in frequency between P15 and P30. Tukey's post hoc test was conducted on all pairwise comparisons following one-way ANOVA. *p < 0.05. ***p < 0.001. F, Decrease in IPSC decay time constant between P10 and later ages. Tukey's post hoc test was conducted on all pairwise comparisons following one-way ANOVA. ****p < 0.0001. P10, n = 15 cells (2 mice); P15, n = 16 cells (7 mice); P21, n = 9 cells (4 mice); P30, n = 14 cells (8 mice); P45, n = 10 cells (5 mice); P60, n = 9 cells (7 mice). G, Cell-by-cell comparison of sEPSC (E) and sIPSC frequency (I) illustrating developmental shift to dominance by spontaneous inhibition. P10, n = 11 cells (2 mice); P15, n = 12 cells (7 mice); P21, n = 7 cells (4 mice); P30, n = 12 cells (8 mice); P45, n = 8 cells (4 mice); P60, n = 7 cells (5 mice). H, Normalized frequency of inhibition (calculated by subtracting sEPSC from sIPSC frequency for individual neurons) illustrating a surge in relative inhibitory drive at P30. Tukey's post hoc comparison was conducted on all pairwise comparisons following one-way ANOVA. **p < 0.01. ***p < 0.001. P10, n = 11 cells (2 mice); P15, n = 12 cells (7 mice); P21, n = 7 cells (4 mice); P30, n = 12 cells (8 mice); P45, n = 8 cells (4 mice); P60, n = 7 cells (7 mice).

Figure 5.

Developmental strengthening of BLA excitatory synapses formed by mPFC projections. A, Summary of experimental procedure. At 7 d after stereotaxic viral infusions, acute brain slices containing the amygdala were obtained for synaptic physiology. Prefrontal tissue was retained and subsectioned to confirm target expression. Monosynaptic EPSCs were recorded from principal excitatory neurons in the anterior basal nucleus during stimulation of mPFC projections by microscope-objective coupled LEDs (460 nm). B, Response amplitude during paired-pulse stimulation was examined as a proxy for glutamate release probability of mPFC terminals. Scale bars: P15, 150 pA; P30, 80 pA; P45, 500 pA; P60, 150 pA × 100 ms. C, Comparison of paired-pulse ratio (second/first response) revealed a decrease between P21 and P30. Bonferroni post hoc test was conducted on all pairwise comparisons following two-way repeated-measures ANOVA. *[black]p < 0.05, P15 versus P60. *[purple]p < 0.01, P21 versus P30, P45, and P60. **[black]p < 0.01, P15 versus P30, P45, and P60. **[purple]p < 0.01, P21 versus P30, P45, and P60. P15, n = 8 cells (3 mice); P21, n = 11 cells (4 mice); P30, n = 10 cells (5 mice); P45, n = 6 cells (5 mice); P60, n = 22 cells (10 mice). D, Representative EPSCs resulting from optic stimulation at −70 mV and 40 mV during blockade of GABAergic transmission (100 μm picrotoxin). For calculation of AMPA/NMDA ratio, AMPAR-EPSC amplitude was defined as the peak amplitude at −70 mV, a potential at which NMDA receptors contribute negligible current. Conversely, NMDAR-EPSCs were measured at 40 mV at 100 ms after LED stimulation, a time point when AMPA receptor currents have fully decayed and responses are dominated by NMDA currents. E, Developmental increase in AMPA/NMDA ratio. P15, n = 11 cells (3 mice); P21, n = 13 cells (5 mice); P30, n = 12 cells (6 mice); P45, n = 13 cells (5 mice); P60, n = 8 cells (5 mice). Tukey's post hoc test was conducted on all pairwise comparisons following one-way ANOVA. **p < 0.01. F, Developmental increase in maximal amplitude of light-evoked EPSCs at −70 mV. Tukey's post hoc test was conducted on all pairwise comparisons following one-way ANOVA. *p < 0.05. G, No change in decay time constant of responses at 40 mV. Inset, Overlay of group-averaged peak-scaled currents at 40 mV, color coded according to scheme in B. P15, n = 10 cells (3 mice); P21, n = 12 cells (4 mice); P30, n = 11 cells (6 mice); P45, n = 13 cells (5 mice); P60, n = 8 cells (5 mice).

Figure 6.

Transient potentiation of feedforward inhibition in the mPFC-BLA pathway. A, Diagram of monosynaptic and disynaptic circuit interrogation by optic stimulation (460 nm). To examine the relative potency of feedforward inhibition, IPSCs were first collected at 0 mV before measuring EPSCs at −70 mV in the presence of a GABA_A-receptor antagonist (100 μm picrotoxin). B, Confirmation of disynaptic mechanism for IPSCs, which were abolished by both GABA_A receptor antagonism (100 μm pictrotoxin) or AMPA/kainate receptor blockade (10 μm CNQX). C, Representative overlays of feedforward IPSCs with monosynaptic EPSCs at each age. Traces were normalized to peak amplitude of EPSCs to illustrate shift in feedforward inhibition. Scale bars: P15, 200 pA; P21, 200 pA; P30, 100 pA; P45, 200 pA; P60, 200 pA × 50 ms. D, Onset latencies of EPSCs and IPSCs at each age were consistent with a monosynaptic and disynaptic mechanism, respectively. Bonferroni post hoc test was conducted on all pairwise comparisons following two-way repeated-measures ANOVA. **p < 0.01. ***p < 0.001. E, IPSC/EPSC amplitude ratio was transiently elevated at P30. Tukey's post hoc test was conducted on all pairwise comparisons following one-way ANOVA. *p < 0.05. P15, n = 9 cells (5 mice); P21, n = 8 cells (5 mice); P30, n = 12 cells (6 mice); P45, n = 9 cells (5 mice); P60, n = 12 cells (10 mice).