Search

The medial temporal lobe drives semantic congruence dependent memory formation. However, the exact roles of hippocampal subfields and surrounding brain regions remain unclear. Here, we used an established paradigm and high-resolution functional magnetic resonance imaging of the medial temporal lobe together with cytoarchitectonic probability estimates in healthy humans. Behaviorally, robust congruence effects emerged in young and older adults, indicating that schema dependent learning is unimpaired during healthy aging. Within the medial temporal lobe, semantic congruence was associated with hemodynamic activity in the subiculum, CA1, CA3 and dentate gyrus, as well as the entorhinal cortex and laterobasal amygdala. Importantly, a subsequent memory analysis showed increased activity for later remembered vs. later forgotten congruent items specifically within CA3, and this subfield showed enhanced functional connectivity to the laterobasal amygdala. As such, our findings extend current models on schema dependent learning by pinpointing the functional properties of subregions within the medial temporal lobe.

Long-term memory can improve when incoming information is congruent with known semantic information. This so-called congruence effect has widely been shown in younger adults, but age-related changes and neural mechanisms remain unclear. Here, congruence improved recognition memory in younger and older adults (i.e. congruence effect), with only weak evidence for age-related decline in one behavioral study. In an EEG study, however, no significant behavioral differences in the congruence effect could be observed between age-groups. In line with this observation, electroencephalography data show that, in both groups, congruence led to widespread differences in Event-Related Potentials (ERPs), starting at around 400 ms after stimulus onset, and theta, alpha and beta oscillations (4–20 Hz). Importantly, these congruence-related ERPs were associated to increases in memory performance for congruent items, in both age groups. Finally, the described ERPs and neural oscillations in the theta-alpha range (5–13 Hz) were less pronounced in the elderly despite a preserved congruence effect. Together, semantic congruence increases long-term memory across the lifespan, and, at the neural level, this could be linked to neural oscillations in the theta, alpha and beta range, as well as ERPs that were previously associated with semantic processing.

As the stream of experience unfolds, our memory system rapidly transforms current inputs into long-lasting meaningful memories. A putative neural mechanism that strongly influences how input elements are transformed into meaningful memory codes relies on the ability to integrate them with existing structures of knowledge or schemas. However, it is not yet clear whether schema-related integration neural mechanisms occur during online encoding. In the current investigation, we examined the encoding-dependent nature of this phenomenon in humans. We showed that actively integrating words with congruent semantic information provided by a category cue enhances memory for words and increases false recall. The memory effect of such active integration with congruent information was robust, even with an interference task occurring right after each encoding word list. In addition, via electroencephalography, we show in 2 separate studies that the onset of the neural signals of successful encoding appeared early (∼400 ms) during the encoding of congruent words. That the neural signals of successful encoding of congruent and incongruent information followed similarly ∼200 ms later suggests that this earlier neural response contributed to memory formation. We propose that the encoding of events that are congruent with readily available contextual semantics can trigger an accelerated onset of the neural mechanisms, supporting the integration of semantic information with the event input. This faster onset would result in a long-lasting and meaningful memory trace for the event but, at the same time, make it difficult to distinguish it from plausible but never encoded events (i.e., related false memories).SIGNIFICANCE STATEMENTConceptual or schema congruence has a strong influence on long-term memory. However, the question of whether schema-related integration neural mechanisms occur during online encoding has yet to be clarified. We investigated the neural mechanisms reflecting how the active integration of words with congruent semantic categories enhances memory for words and increases false recall of semantically related words. We analyzed event-related potentials during encoding and showed that the onset of the neural signals of successful encoding appeared early (∼400 ms) during the encoding of congruent words. Our findings indicate that congruent events can trigger an accelerated onset of neural encoding mechanisms supporting the integration of semantic information with the event input.

Neurobiological models of long-term memory explain how memory for inconsequential events fades, unless these happen before or after other relevant (i.e., rewarding or aversive) or novel events. Recently, it has been shown in humans that retrospective and prospective memories are selectively enhanced if semantically related events are paired with aversive stimuli. However, it remains unclear whether motivating stimuli, as opposed to aversive, have the same effect in humans. Here, participants performed a three phase incidental encoding task where one semantic category was rewarded during the second phase. A memory test 24h after, but not immediately after encoding, revealed that memory for inconsequential items was selectively enhanced only if items from the same category had been previously, but not subsequently, paired with rewards. This result suggests that prospective memory enhancement of reward-related information requires, like previously reported for aversive memories, of a period of memory consolidation. The current findings provide the first empirical evidence in humans that the effects of motivated encoding are selectively and prospectively prolonged over time.

Recent accounts of Posttraumatic Stress Disorder (PTSD) suggest that the encoding of an episode within a fearful context generates different implicit and explicit memory representations. Whilst implicit memory traces include the associated emotional states, explicit traces include a recoding into an abstract or gist-based structural context of the episode. Theoretically, the long-term preservation of implicit memory traces may facilitate the often untreatable memory intrusions in PTSD. Here, we tracked in two experiments how implicit and explicit memory traces for fearful episodes dissociate and evolve over time. Subjects (N=86) were presented with semantically-related word-lists in a contextual fear paradigm and tested for explicit memories either immediately (i.e., 30 min) or after a delay (i.e., 1 or 2 weeks) with a verbal recognition task. Skin Conductance Response (SCR) was used to assess implicit memory responses. Subjects showed high memory accuracy for words when tested immediately after encoding. At test, SCR was higher during the presentation of verbatim but not gist-based words encoded in a fearful context, and remained unchanged after 2 weeks, despite subjects being unaware of words' encoding context. We found no clear evidence of accurate explicit memory traces for the fearful or neutral contexts of words presented during encoding, either 30 min or 2 weeks afterwards. These findings indicate that the implicit, but not the explicit, memory trace of a fearful context of an episode can be detected at long-term through SCR and is dissociated from the gist-based memory. They may have implicationstowards the understanding of how the processing of fearful memoriescould lead to PTSD.

Summary Animal models of human anxiety often invoke a conflict between approach and avoidance [1, 2]. In these, a key behavioral assay comprises passive avoidance of potential threat and inhibition, both thought to be controlled by ventral hippocampus [2–6]. Efforts to translate these approaches to clinical contexts [7, 8] are hampered by the fact that it is not known whether humans manifest analogous approach-avoidance dispositions and, if so, whether they share a homologous neurobiological substrate [9]. Here, we developed a paradigm to investigate the role of human hippocampus in arbitrating an approach-avoidance conflict under varying levels of potential threat. Across four experiments, subjects showed analogous behavior by adapting both passive avoidance behavior and behavioral inhibition to threat level. Using functional magnetic resonance imaging (fMRI), we observe that threat level engages the anterior hippocampus, the human homolog of rodent ventral hippocampus [10]. Testing patients with selective hippocampal lesions, we demonstrate a causal role for the hippocampus with patients showing reduced passive avoidance behavior and inhibition across all threat levels. Our data provide the first human assay for approach-avoidance conflict akin to that of animal anxiety models. The findings bridge rodent and human research on passive avoidance and behavioral inhibition and furnish a framework for addressing the neuronal underpinnings of human anxiety disorders, where our data indicate a major role for the hippocampus., Highlights • Human behavior in a spatial approach-avoidance conflict resembles animal behavior • Threat level in this context specifically engages anterior hippocampus • Hippocampal lesions reduce passive avoidance and inhibition • The study provides human evidence for an anterior hippocampus role in anxiety, Using a combined functional magnetic resonance imaging and hippocampus lesion approach, Bach et al. demonstrate that the human anterior hippocampus is engaged in monitoring threat level during approach-avoidance conflict, replicating and extending a rodent literature on models of anxiety.