The Hippocampus and Amygdala: Functions and Interaction

The hippocampus and amygdala sit just centimeters apart in the brain’s temporal lobes, and they do fundamentally different jobs: the hippocampus builds and retrieves memories, especially memories tied to places and events, while the amygdala evaluates emotional significance and triggers responses like fear. What makes them fascinating as a pair is that they are wired together by direct neural projections and constantly influence each other’s output. The result is a system where emotion stamps certain experiences as worth remembering, spatial maps get redrawn around threats, and stress reshapes the two structures in strikingly opposite ways.

What the Hippocampus Does

The hippocampus is most closely associated with episodic memory, your ability to recall specific events in the context of when and where they happened. When you remember what you had for dinner last Tuesday or the layout of a friend’s apartment, the hippocampus was involved in encoding and retrieving those details. A key mechanism is pattern completion: given a partial cue, the hippocampus can reconstruct the full experience. Brain-imaging work has shown that this hippocampal pattern completion contributes to what researchers call “holistic reinstatement,” the phenomenon of an entire event springing back to mind from a single reminder. This process operates both immediately and after a delay of at least 24 hours, though after consolidation the neocortex begins contributing as well, working alongside the hippocampus rather than replacing it.¹

The hippocampus also builds spatial maps. Neurons called place cells fire when an animal occupies a particular location in its environment, collectively forming an internal representation of the surrounding space. This spatial-mapping function is ancient. Comparative studies across goldfish, turtles, lizards, rodents, monkeys, and humans indicate that the hippocampal homologue originally evolved to house map-like representations of odors and sights encountered at various locations, including the order and timing in which they would be encountered during a journey.² These spatial and episodic memory functions are deeply intertwined: remembering where something happened is often inseparable from remembering what happened.

What the Amygdala Does

The amygdala is often described as the brain’s alarm system, but that undersells its complexity. It is not a single structure but a cluster of nuclei, each with distinct cell types and roles. The basolateral complex (BLA) receives sensory information from the cortex and thalamus and is the primary site where emotional associations are formed. The central nucleus (CeA) is the main output station, sending signals to brain-stem areas that trigger the physical expressions of emotion: a racing heart, a freeze response, a spike in stress hormones. Within the CeA, different neuron populations handle different aspects of fear processing. For instance, neurons expressing corticotropin-releasing factor (CRF) in the central amygdala help the brain distinguish between a genuine threat and a neutral stimulus, facilitating what researchers call discriminative fear.³

Beyond fear, the amygdala processes a range of emotionally charged information, including reward, social signals, and ambiguity. Its job, broadly, is to flag what matters. That flagging function is what connects it so tightly to the hippocampus: once something is tagged as emotionally important, the amygdala helps ensure it gets stored.

The Physical Wiring Between Them

The hippocampus and amygdala communicate through direct axonal projections, not just through indirect cortical routes. Tracing studies in rodents have shown that excitatory neurons in the basolateral amygdala send projections specifically to the ventral (lower) portion of the hippocampal CA1 region, with little to no input reaching the dorsal (upper) hippocampus.⁴ The ventral hippocampus, in turn, sends projections back to the basal amygdala, creating a bidirectional loop.

This ventral hippocampus–basal amygdala pathway is not just an anatomical curiosity. When researchers silenced activity in this pathway during contextual fear learning using chemogenetic tools, mice showed significantly reduced fear responses when returned to the same context a day later. The pathway was not needed for learning about an isolated cue like a tone, but it was essential for encoding fear tied to the surrounding environment.⁵ In other words, the hippocampus tells the amygdala where you are, and the amygdala uses that information to decide whether you should be afraid.

How Emotion Shapes Memory Consolidation

The most studied interaction between the hippocampus and amygdala involves memory consolidation, the hours-long process by which a fresh experience becomes a stable long-term memory. Emotional events are remembered better than neutral ones, and the amygdala is a central reason why. The two structures activate following an emotional event and cross-talk with each other during consolidation, working together to form durable memories of events that carried emotional weight.⁶

The BLA is the key player on the amygdala side. Decades of animal research using drug infusions, selective lesions, and more recently optogenetic activation have converged on a clear picture: the BLA modulates memory consolidation by influencing neural plasticity in downstream brain regions. This modulation involves stress hormones (particularly norepinephrine) and several neurotransmitter systems converging within the BLA. The BLA then sends its modulatory signals outward to the hippocampus, cortex, caudate nucleus, and nucleus accumbens.⁷ The effect is not limited to fear memories; the BLA modulates consolidation of many different kinds of information, from spatial learning to reward-based tasks.⁸

The interaction is not one-directional. Hippocampal input to the amygdala is essential for contextual modulation of fear. When rats learned to associate a specific context with danger, the functional connection between the hippocampal subiculum and the BLA was required for that contextual fear memory to consolidate. Blocking synaptic plasticity at the junction between these regions after training impaired contextual fear memory while leaving cue-based fear memory intact.⁹ The hippocampus provides the spatial and contextual scaffold; the amygdala provides the emotional charge. Neither alone produces a full contextual fear memory.

Opposing Roles in the Stress Response

The hippocampus and amygdala exert opposite influences on the body’s main stress-hormone system, the hypothalamic-pituitary-adrenal (HPA) axis. The hippocampus acts as a brake. When hippocampal neurons fire, they predominantly inhibit the stress-hormone-releasing neurons in the hypothalamus through an indirect (disynaptic) pathway, dampening cortisol output.¹⁰ The amygdala, by contrast, acts as an accelerator, enhancing glucocorticoid secretion when it detects a threat.¹¹

This push-pull arrangement means that a healthy brain can mount a fast stress response when danger appears (amygdala pushes the gas) and shut it down efficiently once the danger passes (hippocampus applies the brake). Problems emerge when chronic stress tips the balance.

How Chronic Stress Reshapes Both Structures

Prolonged stress does not wear down the hippocampus and amygdala equally. It drives them in opposite structural directions. In rodents exposed to chronic immobilization stress, hippocampal CA3 neurons showed dendritic atrophy and lost branches, which is consistent with earlier reports linking chronic stress to hippocampal shrinkage. The amygdala, meanwhile, responded to the same stress protocol with enhanced dendritic branching in its basolateral neurons.¹²

The practical implication is troubling. A shrinking hippocampus means a weaker brake on the stress response. A growing amygdala means a stronger accelerator. The system drifts toward a state where stress becomes self-reinforcing: more stress leads to more amygdala reactivity and less hippocampal regulation, which leads to more stress-hormone exposure, which drives further remodeling. This feedback loop is thought to underlie some of the persistent anxiety and emotional dysregulation seen in chronic-stress-related conditions. The hippocampus also continues to produce new neurons in the dentate gyrus throughout adulthood, and stress hormones (glucocorticoids) suppress this neurogenesis, further reducing the hippocampus’s capacity to support learning and regulate mood.¹³

When Emotion Sharpens or Blurs Similar Memories

One of the hippocampus’s subtler tasks is pattern separation, the ability to store two similar experiences as distinct memories rather than blending them together. If you parked in a slightly different spot at the same parking lot on two consecutive days, pattern separation is what lets you remember Tuesday’s spot without confusing it with Monday’s. The dentate gyrus, a subregion of the hippocampus, is considered critical for this process.

Emotion complicates pattern separation in an interesting way. In imaging studies, participants showed stronger hippocampal signals during correct discrimination of similar emotional items compared with neutral ones, and the amygdala was also more active during emotional trials regardless of discrimination accuracy.¹⁴ Emotional arousal, in other words, can boost the hippocampus’s ability to keep similar threatening experiences separate in memory.

But the mechanism has a darker side. Intracranial recordings in epilepsy patients revealed that successful pattern separation of emotional stimuli was associated with theta-band (roughly 3–7 Hz) bidirectional communication between the amygdala and hippocampus. When pattern separation failed and patients incorrectly lumped similar experiences together, the communication shifted to an alpha-band (roughly 7–13 Hz) pattern dominated by a one-way signal from the amygdala to the hippocampus.¹⁵ This matters because overgeneralization of threatening memories, seeing danger everywhere because your brain cannot separate one bad experience from superficially similar safe ones, is a hallmark of PTSD. The alpha-band coupling that drives overgeneralization represents a potential target for future therapies.

A complementary line of evidence suggests that norepinephrine release during emotional arousal can actually enhance pattern separation. Higher levels of a norepinephrine marker in saliva correlated with better performance on a pattern-separation task administered shortly after viewing emotionally arousing images.¹⁶ So the amygdala’s influence can cut both ways: the right kind of arousal sharpens distinctions between similar memories, while the wrong kind of coupling collapses those distinctions.

How Threats Rewrite Spatial Maps

The hippocampus’s spatial maps are not static. They respond to emotional context, and the amygdala is the structure that drives those changes. When rats foraged in an environment that included a “predatory” threat (a robot that lunged at them), hippocampal place cells near the threat zone underwent remapping: their location-specific firing patterns changed significantly compared to pre-threat sessions, while place cells farther from the threat stayed stable.¹⁷ The hippocampus essentially rewrote its internal map of the dangerous zone while leaving the safe areas alone.

Beyond threat-related remapping, the dorsal hippocampus can incorporate valence information more broadly. Place fields cluster around sites associated with positive outcomes (rewards) and negative stimuli (shocks, predator odors), effectively marking the locations of both rewards and threats while retaining spatial precision.¹⁸ Some hippocampal neurons fire selectively at reward sites regardless of spatial position, while others fire during freezing behavior. The spatial map is never purely geometric; it is always colored by what happened at each location.

Coordinated Replay During Sleep

Memory consolidation does not stop when you fall asleep. During non-REM sleep, the hippocampus produces sharp wave-ripple complexes, brief bursts of high-frequency activity thought to replay recent experiences and push them toward long-term cortical storage. The amygdala generates its own ripples and sharp waves during the same sleep stages, and these events are coordinated between the two structures. Combined with evidence of interregional spindle synchronization, this coordination offers a potential physiological basis for how emotional experiences get consolidated and regulated during sleep.¹⁹

There is also evidence that the ventral hippocampus–BLA pathway contributes to sleep disruption after traumatic experiences. Research using calcium imaging and optogenetics has begun dissecting how reactivation of fear-memory-linked neurons in this circuit during sleep can fragment normal sleep architecture, a finding with obvious relevance to the sleep disturbances that plague people with PTSD and other anxiety-related conditions.²⁰

Clinical Relevance in PTSD and Alzheimer’s Disease

The interplay between the hippocampus and amygdala is not just an academic curiosity; it is central to understanding several neurological and psychiatric conditions. In PTSD, a large multisite neuroimaging study found that people with the disorder had significantly smaller hippocampi compared with trauma-exposed controls, along with smaller amygdalae, though the amygdala finding was less robust statistically.²¹ A scoping review of subregion-level imaging found that the hippocampal shrinkage in PTSD concentrates in the CA1, CA3, dentate gyrus, and subiculum, while amygdala changes affect the lateral, basal, and central nuclei.²² Functionally, hyperarousal symptoms in PTSD have been linked to increased activation in the amygdala and dorsal anterior cingulate cortex.²³

In Alzheimer’s disease, the amygdala is attracting renewed attention. While the hippocampus has long been recognized as an early target of Alzheimer’s pathology, tau-PET imaging studies in cognitively healthy but amyloid-positive individuals have found among the highest tau signals in the amygdala at the earliest detectable stages, suggesting that the amygdala may be affected even before cognitive symptoms emerge.²⁴ MRI-based studies in clinical Alzheimer’s patients have confirmed that the strongest atrophy concentrates in the hippocampal subiculum and CA1 regions alongside the basolateral and centromedial subregions of the amygdala.²⁵ The fact that both structures are hit early helps explain why Alzheimer’s patients lose not just factual memories but also the emotional coloring of those memories and the ability to learn new emotional associations.

The Hippocampus-Amygdala Circuit in Epilepsy

Mesial temporal lobe epilepsy (MTLE) is the most common form of drug-resistant epilepsy, and the hippocampus and amygdala are both deeply implicated. Recordings of high-frequency oscillations in patients suggest that the amygdala contributes to seizure generation, while the hippocampus plays a critical role in seizure propagation.²⁶ The anatomical pattern of how seizures spread through the temporal lobe is associated with the structural reinforcement of the white-matter tracts connecting these regions.²⁷

An intriguing finding is that roughly 14% of MTLE patients with hippocampal sclerosis show significant amygdala enlargement, almost always on the opposite side of the brain from the damaged hippocampus. Patients with this pattern had an earlier age of epilepsy onset.²⁸ Whether this contralateral amygdala growth represents a compensatory response or a pathological one is still unclear, but it underscores how tightly these two structures are linked: damage to one consistently produces measurable changes in the other.

How the Circuit Develops in Early Life

The amygdala’s connectivity with the rest of the brain undergoes dramatic changes during the first years of life. A cross-sectional imaging study of children from 3 months to 5 years of age found that amygdala-thalamus connectivity shifted from positive to negative over that period, while the lateral-basal subregion of the amygdala developed increasing connectivity with sensory and motor cortical regions. The superficial subregion, in contrast, showed decreasing connectivity with subcortical and cortical sensory areas over the same period.²⁹ These shifts suggest that the amygdala’s different subregions mature on different timelines and establish their adult-like communication patterns gradually. The period from infancy through early childhood is thought to be a sensitive window during which adverse experiences can shape amygdala reactivity for years to come.

Gut Bacteria and the Amygdala-Hippocampus System

One of the more surprising lines of recent research connects gut microbes to the function and even the structure of both the hippocampus and amygdala. Mice raised in completely germ-free conditions, meaning they have no gut bacteria at all, show changes in the morphology of both structures and alterations in fear-related behavior.³⁰ In the amygdala specifically, germ-free mice show upregulation of genes involved in synaptic transmission, nervous system development, and neurogenesis, along with signs of hyperactivity, as if the amygdala is running hotter without microbial input to calibrate it.³¹ These are animal studies, and translating them to human health requires caution. But they open the possibility that gut-brain communication influences emotional memory processing at a more fundamental level than anyone suspected a couple of decades ago.

Fear Extinction and the Prefrontal Cortex

The hippocampus-amygdala story is incomplete without a third player: the ventromedial prefrontal cortex (vmPFC). Fear extinction, the gradual reduction in fear response when a threat no longer materializes, depends on all three structures working together. The hippocampus provides contextual information, letting the brain know where and when it is. The vmPFC exerts top-down inhibition on the amygdala, suppressing the fear response in safe contexts. Hippocampal input to both the vmPFC and the BLA is essential for context to appropriately gate when fear is acquired and when it is extinguished.³²

This three-way circuit is directly relevant to exposure-based therapies for anxiety and PTSD. The goal of such therapies is, in neuroscience terms, to strengthen the prefrontal cortex’s ability to suppress amygdala reactivity and to help the hippocampus encode a new, safe-context memory that competes with the old fear memory. When any part of this circuit is compromised, whether by chronic stress atrophying the hippocampus, by an overactive amygdala, or by weakened prefrontal connectivity, extinction learning suffers and fear responses persist.