Cell-type-specific recruitment of amygdala interneurons to hippocampal theta rhythm and noxious stimuli in vivo

Abstract

Neuronal synchrony in the basolateral amygdala (BLA) is critical for emotional behavior. Coordinated theta-frequency oscillations between the BLA and the hippocampus and precisely timed integration of salient sensory stimuli in the BLA are involved in fear conditioning. We characterized GABAergic interneuron types of the BLA and determined their contribution to shaping these network activities. Using in vivo recordings in rats combined with the anatomical identification of neurons, we found that the firing of BLA interneurons associated with network activities was cell type specific. The firing of calbindin-positive interneurons targeting dendrites was precisely theta-modulated, but other cell types were heterogeneously modulated, including parvalbumin-positive basket cells. Salient sensory stimuli selectively triggered axo-axonic cells firing and inhibited firing of a disctinct projecting interneuron type. Thus, GABA is released onto BLA principal neurons in a time-, domain-, and sensory-specific manner. These specific synaptic actions likely cooperate to promote amygdalo-hippocampal synchrony involved in emotional memory formation.

Figure 1

Axo-axonic Cells: Firing In Vivo and Anatomical Characterization (A) In vivo, the neuron tjx20f fired non θ-modulated spike trains. (B) tjx20f increased its firing rate in response to hindpaw pinches. The electrocorticogram (ECoG) shows stable global activation. (C) Another representative axo-axonic cell (tjx56b), dramatically increased firing upon hindpaw electrical shocks. (D) Close appositions between tjx63a axon varicosities and ankyrin G-expressing axon initial segments (AIS). Low (top) and high magnification (of area delineated, bottom) of projection of a confocal z stack of 4.42 μm thickness. (E) Electron micrograph of a DAB-labeled axon bouton (tjx20f) forming two synaptic junctions (arrows) with a single AIS. ^∗undercoating. (F) tjx20f is immunopositive for parvalbumin (PV; single confocal optical section). (G) Reconstruction of tjx20f. Soma and entire dendritic tree (red) are drawn from 5 sections of 60 μm thickness. Axon (blue, main axon is purple) is drawn from 2 sections, for clarity. Inset: position of tjx20f dendrites (in red) in the BLA and axonal field extent (gray area) estimated from the two drawn and surrounding sections. Boundary colors apply to the main panel. LA: lateral amygdala, BA: basal amygdala, CeA: central amygdala, ITC: intercalated cells cluster. Orientation: top: dorsal, right: medial. Scale bars: (A) LFP raw and filtered: 0.4 mV, unit: 1 mV, time: 400 ms; (B) unit: 1 mV, ECoG: 0.25 mV, time: 4 s; (C) unit: 1 mV, ECoG: 0.25 mV, time: 1 s; (D) top: 20 μm, bottom: 5 μm; (E) 500 nm; (F) 10 μm; (G) 100 μm, inset 500 μm. See also Figures S1–S5 and Tables S1–S3.

Figure 2

PV Basket Cells: Firing In Vivo and Anatomical Characterization All panels show data from the same neuron (tjx48a). (A) Tonic, weakly modulated firing during hippocampal θ oscillations. (B) Moderate firing increases in response to hindpaw electrical shocks, not noticeable in the raw data. (C) Axon varicosities making close appositions with the soma of a CaMKIIα⁺ principal neuron (projection of a confocal z stack of 3.29 μm thickness). (D) Low (top) and high magnification (bottom) electron micrographs of a synapse (arrow) made by a labeled axon bouton with a soma. (E) Immunopositivity for parvalbumin and calbindin. Immunofluorescence confocal images (single optical section). (F) Accumulation of GABA_AR-α1 (α1) at the dendritic membrane of the Neurobiotin (NB)-filled cell (single confocal optical sections). (G) Reconstruction of tjx48a. Soma and entire dendritic tree are drawn from 12 sections of 60 μm thickness. Axon is drawn from 2 sections for clarity. Inset: position of tjx48a dendrites in the BLA and axonal field extent estimated from the two drawn and surrounding sections (color code as in Figure 1G). Orientation: top: dorsal, right: medial. Scale bars: (A) LFP raw and filtered: 0.4 mV, unit: 1 mV, time: 400 ms; (B) unit: 1 mV, ECoG: 0.25 mV, time: 1 s; (C) 10 μm; (D) top 2 μm, bottom 250 nm; (E) 20 μm; (F) 5 μm; (G) 100 μm, inset 500 μm. See also Figures S1–S7 and Tables S1–S3.

Figure 3

CB⁺ Dendrite-Targeting Cells: Firing In Vivo and Anatomical Characterization (A) tjx22c fired preferentially before the peak of dCA1 θ. (B) tjx59b did not change its firing rate during a noxious stimulus. (C) Axon varicosities of tjx21i avoid CaMKII⁺ somata and occasionally make visible appositions with CaMKII⁺ dendrites (arrows; projection of a confocal z stack of 2.10 μm thickness). (D) Electron micrograph of a synapse formed by an axon bouton of tjx22c with a CaMKIIα⁺ dendrite (de). Inset: higher magnification. (E) tjx22c is immunopositive for calbindin (structured illumination, single optical section). (F) tjx22c is GABAergic: confocal images (projection of z stack of 0.75 μm thickness) showing Neurobiotin (NB)-filled axon varicosities containing VGAT (arrows). (G) Reconstruction of tjx22c. Soma and entire dendritic tree are drawn from 8 sections of 60 μm thickness. Axon is drawn from 2 sections, for clarity. Inset: position of tjx22c dendrites in the BLA and axonal field extent estimated from the two drawn and surrounding sections (color code as in Figure 1G). Orientation: top: dorsal, right: medial. Scale bars: (A) LFP raw and filtered: 0.4 mV, unit: 1 mV, time: 400 ms; (B) unit: 1 mV, ECoG: 0.5 mV; (C) 10 μm; (D) 500 nm, inset 100 nm; (E) 20 μm; (F) 2 μm; (G) 100 μm, inset 500 μm. See also Figures S1–S6 and Tables S1–S3.

Figure 4

AStria-Projecting Cells: Firing In Vivo and Anatomical Characterization All panels show data from one neuron (tjx52a). (A) Preferential firing during the ascending phase of hippocampal θ oscillations. (B) Reduced firing during hindpaw pinches. (C) Reconstruction of tjx52a. Soma and entire dendritic tree are drawn from 5 sections of 60 μm thickness. Axon is drawn from 2 sections, for clarity. Inset: position of tjx52a dendrites in the BLA and axonal field extent in the two drawn and surrounding sections. Color code is as in Figure 1G, and AStria is purple. Orientation: top: dorsal, right: medial. (D) Electron micrograph of a labeled bouton making a synapse (arrow) with a CaMKIIα⁺ soma in the AStria. (E) Axon varicosities making close appositions with a principal neuron (CaMKIIα⁺/DARPP-32⁺) soma in the AStria. Projection of a confocal z stack of 3.06 μm thickness. (F) Electron micrograph of a labeled bouton making a synapse (arrow) with a CaMKIIα⁺ principal cell dendrite in the lateral amygdala. (G) Axon varicosities making close appositions with principal neuron somata (CaMKIIα⁺) or unidentified structures (arrows) in the lateral amygdala (projection of confocal z stacks of 8.05 μm thickness). (H) Immunopositivity for parvalbumin (structured illumination z stack; NB: Neurobiotin). (I) tjx52a is GABAergic: confocal images (projections of z stack of 0.94 μm thickness) showing axon varicosities (arrows) positive for VGAT and GAD. NB: Neurobiotin. Scale bars: (A) units and LFPs: 0.5 mV, time: 400 ms; (B) units 1 mV, time: 4 s; (C) 100 μm, inset 500 μm; (D) 500 nm; (E) 10 μm; (F) 250 nm; (G) 10 μm; (H) 10 μm; (I) Scale bar: 2.5 μm. See also Figures S1–S7 and Tables S1–S3.

Figure 5

Firing Modulation with Hippocampal Theta Oscillations and Responses to Noxious Stimuli Are Cell Type Specific (A) Cell types mean θ phase histograms. CB⁺ dendrite-targeting interneurons are homogeneously and strongly modulated in phase with dCA1 θ, in contrast to the other cell types. Two θ cycles are represented for clarity. 0°, 360°, and 720°: θ troughs. (B) Polar distribution of individual neurons' preferred θ phases and modulation depths. Each symbol represents a significantly modulated cell. (C and D) Cell types mean peristimulus histograms for noxious stimulation. Axo-axonic cells are dramatically excited and AStria-projecting cells are inhibited by electrical shocks and pinches delivered to the controlateral hindpaw. Error bars: SEM; dashed lines: mean ± 2 standard deviations. n values represent the number of neurons tested in each analysis. See also Figures S2, S4, and S5.

Figure 6

Anatomical Features and Theta Modulation of BLA Principal Neurons (A) The neuron tjx89c had large dendrites covered with spines (projection of a confocal z stack of 7.99 μm thickness). (B) tjx89c expressed VGluT1, demonstrating its identity as a glutamatergic neuron (single optical section confocal micrographs; recordings from this neuron are shown in Figure 7A). (C) Moderate levels of CaMKIIα, (^∗) as well as weak calbindin (^∗∗) expression were detected in the soma of tjx89c (confocal images, single optical sections). (D) Phase histograms. Two θ cycles are shown for clarity. Error bars represent SEM. (E) Polar distribution of preferred θ phase and modulation depth of the θ-modulated principal neurons (n = 9/23). Each symbol represents a modulated neuron. Note that all 9 cells were strongly modulated. Scale bars (A) 10 μm; (B) 2 μm (C) 10 μm. See also Figures S1 and S8, and Table S4.

Figure 7

Dorsal CA1 Reference Recapitulates BLA Unit Firing Modulation by Ventral Hippocampal θ (A and B) A principal neuron (tjx89c) was recorded along with LFPs from dorsal (CA1 str. pyramidale) and vHippocampus (subiculum str. radiatum). (A) Raw data showing tjx89c firing during an exemplary period of ventral hippocampal θ. Traces filtered for θ frequencies are overlaid. (B) Theta oscillations in the ventral subiculum are phase-locked to those in dCA1. Phase histogram representing the distribution of vSubiculum θ troughs across the dCA1 θ cycle. (C and D) Single units modulated in phase with dCA1 θ are similarly modulated with vHippocampus θ. (C) Left: phase histogram of tjx89c spikes computed with the dCA1 reference. vSubiculum θ trough distribution was superimposed to illustrate the predicted firing phase relationship of tjx89c with vSubiculum θ rhythm (predicted phase: 135.7°). Right: distribution of tjx89c spike angles relative to vSubiculum θ, consistent with the prediction. (D) Results obtained with dCA1 θ phase as a reference accurately reflect firing modulation of neurons in phase with ventral hippocampal θ. Top: linear correlation between the predicted (LFPs) and actual phase differences of neuron modulation by dorsal and ventral hippocampal θ. Bottom: linear correlation between modulation depths calculated with the dorsal and ventral hippocampal references. Diagonal dashed lines: unity line. (E) Anatomically confirmed recording sites in the ventral subiculum from the same experiment. The white box highlights the Neurobiotin (green) deposit made at the str. pyramidale recording site. NeuN immunoreactivity (red) was used as a panneuronal nuclear and cytoplasmic marker to delineate str. pyramidale. Calbindin immunoreactivity (light blue) highlights hippocampal layers. Colocalization with NeuN is indicated by white and pink. Inset: higher magnification of the area delineated in the main panel. Arrow: recording site in str. radiatum. SO: str. oriens; SP: str. pyramidale: SR, str. radiatum; SLM: str. lacunosum moleculare. Scale bars: (A) time: 1 s, LFPs: 0.5 mV, units: 2 mV; (E) 1 mm, inset: 100 μm. See also Figure S9 and Table S6.