Comprehensive Notes: Encoding of Conditioned Fear in Central Amygdala Inhibitory Circuits

Functional Architecture and Role of Central Amygdala (CEA) Local Inhibitory Circuits

• The Central Amygdala (CEA) is a nucleus primarily composed of inhibitory GABAergic neurons and is considered essential for the acquisition and expression of conditioned fear.
• The classic model of fear circuitry positions the Lateral Nucleus of the Amygdala (LA) as the primary site for associations between the Conditioned Stimulus (CS) and Unconditioned Stimulus (US).
• In this classical view, the CEA is primarily designated for the behavioral expression of fear responses, with output neurons in the Medial Subdivision (CEm) projecting to the brainstem and hypothalamus to orchestrate autonomic and motor responses.
• This study identifies a functional dissociation within the CEA:     * The Lateral Subdivision (CEl) is required for fear acquisition.     * The Medial Subdivision (CEm) drives conditioned fear responses.
• The CEA circuitry consists of highly organized microcircuits where cell-type-specific plasticity in phasic and tonic activity regulates fear expression and generalization.

Differential Roles of CEl and CEm Subnuclei in Fear Behavior

• Methodology for CEm Activation: Bilateral targeted injection of an adeno-associated virus expressing channelrhodopsin-2 (rAAV-Syn-ChR22A-tdimer) was used to activate neurons optogenetically.
• Observation of CEm Activation: In freely moving animals, bilateral activation of CEm induced strong and reversible freezing responses.
• Methodology for Inactivation: Microiontophoresis of a fluorescently labeled $GABAA$ receptor agonist, muscimol-bodipy (muscimol-BPY), was used to inhibit neuronal activity reversibly.
• Impact of CEm/CEA Inactivation on Baseline Behavior: Bilateral inactivation of the CEm or the entire CEA (CEm and CEl) did not elicit freezing behavior in unconditioned mice.
• Impact of CEl Inactivation on Baseline Behavior: Targeted inactivation of the CEl alone induced unconditioned freezing, indicating that the CEm is under tonic inhibitory control from the CEl.
• Specific Roles in Acquisition:     * Bilateral inactivation of the entire CEA during fear conditioning resulted in a deep memory deficit when animals were tested 24 hours later in a drug-free state.     * A comparable memory deficit occurred with inactivation of the CEl.     * Inactivation of the CEm during conditioning did not impair memory acquisition.     * Inactivation of the CEA/CEl did not impair the detection of the US (foot-shock).
• Specific Roles in Retrieval/Expression:     * Local application of muscimol-BPY 24 hours after conditioning (during the retrieval phase) showed that inactivation of the entire CEA or the CEm resulted in an expression deficit.     * In contrast, inactivation of the CEl did not reduce conditioned freezing levels, although the results might be occluded by unconditioned freezing.

Identification and Classification of CEl Neuronal Subtypes

• Recordings of 167 units in the CEl were used to investigate CS-evoked firing changes after fear conditioning.
• $CEl_{on}$ Neurons: Representing $30\%$ of units, these neurons acquired an excitatory response to the CS after fear conditioning.
• $CEl_{off}$ Neurons: Representing $25\%$ of units, these neurons displayed a strong inhibitory response to the CS after conditioning.
• Non-responsive Neurons: Approximately $45\%$ of units did not exhibit tone-evoked responses.
• The onset latency of CS-evoked excitation in $CEl_{on}$ neurons was very short ($< 15\,ms$), suggesting they receive direct input from the sensory thalamus.
• Fear conditioning induces rapid, specific, and persistent changes in these activity patterns.

Functional Organization and Connectivity of CEA Microcircuits

• Intra-CEl Interactions: Cross-correlation analysis of spontaneous spikes revealed asymmetrical short-latency inhibitory interactions.     * $CEl_{on}$ to $CEl_{off}$: Inhibitory interaction in $26\%$ (9 of 35) of pairs.     * $CEl_{off}$ to $CEl_{on}$: Inhibitory interaction in $9\%$ (3 of 35) of pairs.     * Inhibitory interactions among only $CEl_{on}$ units or only $CEl_{off}$ units were rare.
• Projections to CEm: Retrograde tracing using Herpes Simplex Virus 1 (HSV-1) injected into the CEm resulted in intense labeling in the CEl, while CEm remained largely devoid of GFP when HSV-1 was injected into the CEl, suggesting a unidirectional CEl-to-CEm projection.
• Morphological Reconstructions: Intracellular recordings and reconstructions showed that axons of both $CEl_{on}$ and $CEl_{off}$ subtypes arborize locally within the CEl and send collaterals to the CEm.
• CEl to CEm Inhibition: Multi-site single-unit recordings confirmed that both $CEl_{on}$ and $CEl_{off}$ neurons exhibit inhibitory interactions with CEm neurons in vivo, with no reverse interactions found.

Encoding of Disinhibition in CEm Output Neurons

• Identification of Output Neurons: CEm output neurons were identified by placing a stimulation electrode in the mesencephalic axon bundle (containing projections to brainstem targets).
• Antidromic Responses: Antidromic spikes followed high-frequency stimulation ($200\,Hz$) with low temporal jitter ($< 0.1\,ms$).
• Firing Patterns: $83\%$ of CEm neurons (including all identified brainstem-projecting cells) showed a marked increase in CS-evoked firing 24 hours after fear conditioning.
• Biphasic CS Response in CEm:     * First Component: Brief, short-latency ($10\text{--}15\,ms$), likely driven by direct thalamic input.     * Second Component: Sustained, paralleling the timeframe of $CEl_{off}$ inhibition, suggesting disinhibition plays a significant role in triggering fear responses.

Plasticity of Tonic Activity and Fear Generalization

• Tonic Activity Changes: Fear conditioning caused a decrease in the absolute level of spontaneous (tonic) activity in CEm output neurons and $CEl_{on}$ neurons, but an increase in $CEl_{off}$ neurons.
• Correlation with Freezing: Phasic CS-evoked activity (z-scored) was highly correlated with freezing behavior ($r^2 = 0.243, P < 0.05$ for CEm; $r^2 = 0.119, P < 0.01$ for $CEl_{off}$; $r^2 = 0.098, P < 0.01$ for $CEl_{on}$).
• Generalization vs. Discrimination:     * Discrimination was defined as a freezing ratio of $CS1/CS2 > 1.4$.     * Generalization was defined as a freezing ratio of $CS1/CS2 < 1.4$.     * A decrease in the tonic activity of CEm output neurons was associated with generalization.     * In contrast, $CEl_{on}$ and $CEl_{off}$ neurons showed the inverse correlation regarding generalization.
• ROC Analysis: A Receiver Operating Characteristic (ROC) analysis of CEl neurons yielded an Area Under Curve (AUC) of $0.67$, indicating changes in tonic activity are significant discriminators ($P < 0.01$) for identifying generalizing mice versus discriminating ones.
• Network State: Changes in tonic activity appeared before CS onset, indicating generalization is associated with a distinct functional network state.

Discussion of Learning Mechanisms and Afferent Pathways

• The CEl is essential for fear acquisition, suggesting synaptic plasticity at glutamatergic inputs onto $CEl_{on}$ neurons.
• Potential glutamatergic inputs include the Basolateral Amygdala (BLA), insular cortex, and the parabrachial nucleus.
• The parabrachial nucleus provides strong synapses that may serve as a "teaching signal" for inducing plasticity at other inputs.
• The biphasic response of CEm indicates an integration of thalamic excitation and disinhibition through the $CEl_{off}$ pathway.
• The CEA circuit is structurally similar to striatal circuits, implying that winner-share-all mechanisms or coordinated changes in phasic/tonic inhibition may be a widespread mechanism for output selection in the CNS.

Experimental Methods and Parameters

• Subjects: Male C57BL6/J mice (2-3 months old).
• Fear Conditioning Paradigm: Discriminative auditory conditioning using a CS1 (paired with US) and a CS- (unpaired white noise or 7.5 kHz pips, 80 dB SPL).
• Foot Shock Parameters: US was a $1\,s$ foot shock at $0.6\,mA$.
• Freezing Quantification: Automatic infrared beam detection; freezing defined as no movement for $> 2\,s$.
• Iontophoresis Currents: Cationic current of $112\text{--}115\,nA$ for $15\,min$ per side for muscimol-BPY application.
• Electrophysiology: 8-16 individually insulated nichrome wires ($13\,\mu m$ diameter); impedance $50\text{--}300\,k\Omega$; digitizing at $40\,kHz$; filtering from $250\,Hz$ to $8\,kHz$.
• Statistical Thresholds: Z-score normalization used a $500\,ms$ pre-stimulus baseline. Significant onset was defined as the first bin exceeding $2.5$ standard deviations of baseline activity.