Hippocampus-dependent memory enhancement by the amygdala

stimulation of rat amygdala

Emotional events are often remembered better than neutral events, and the basolateral complex of the amygdala (BLA) is thought to mediate this memory enhancement in part via its direct projections to the hippocampus. In previous studies with experimental animals, pharmacological activation of the BLA was shown to enhance memory without necessarily inducing emotional arousal, but a question left unanswered by the slow time course of pharmacological activation was whether or not direct activation could enhance memory for some events and not others, a prerequisite for effective emotional prioritization of event memories. In an initial study (Bass, Partain, and Manns, 2012), we demonstrated that brief (1 s) electrical stimulation of the BLA following offset of novel object exploration in rats resulted in those objects being remembered one day later whereas interleaved encounters with control objects were forgotten by that point. In a follow-up study (Bass, Nizam, Partain, Wang, and Manns, 2014), we replicated the results and extended the findings by showing that inactivation of the hippocampus (via infusion of muscimol) eliminated any benefit to memory of BLA stimulation. In a third study (Bass and Manns, 2015), we recorded neural activity in the hippocampus of rats performing the same task. We found that electrical stimulation of the BLA elicited the same type of low gamma synchrony in the hippocampus (CA3-CA1 low gamma coherence) as we previously observed to be associated with good recognition memory performance without BLA stimulation (Trimper, Stefanescu, Manns, 2014). The results of this program of research indicate that activation of the BLA results in improved memory in part by targeted modulation of activity in the hippocampus. The long-term goal of this research program is to understand how modulation of the hippocampus by the amygdala can at times lead to memory enhancement and at times lead to memory dysfunction, such as that observed in post-traumatic stress disorder (PTSD).

human amygdala

In collaboration with colleagues in the department of Neurosurgery, we recently published results from similar experiments in human patients who volunteered to participate in amygdala stimulation studies after having electrodes implanted to monitor for epileptic seizures (Inman et al., 2017). Stimulation of the amygdala in humans resulted in improved one-day recognition memory despite all the participants indicating that they could not feel the stimulation.


The role of spatiotemporal context in hippocampus-dependent memory


Spiking activity in hippocampus

Another line of research has focused on electrophysiological recordings from the hippocampus of rats performing memory tasks and has asked how spiking activity in the hippocampus relates to memory for individual items encountered in a spatial or temporal context. One set of results (Manns et al., 2007) suggested that the hippocampus contributed to memory for temporal order neither by directly encoding the ordinal position of odors nor by supporting simple judgments of recency but instead by allowing items to be associated with the context in which they were encountered. By looking at multi-neuron ensemble activity, we found that temporally adjacent items were represented more similarly than temporally distant items and that this similarity predicted successful performance. Our findings were consistent with previous results from both computational modeling and cognitive psychology, which had identified a type of running average of experience often referred to as temporal context. Another recently completed study showed clear evidence of temporal context effects on performance resulting from simple temporal contiguity of items (Manns, Galloway, and Sederberg, 2015). Specifically, repeating the first two objects from a once-encountered sequence of three objects incidentally cued memory for the third object, even in its absence. One interpretation is that repeating the first two objects reactivated the original temporal context, an interpretation consistent with human behavioral and modeling data. More broadly, these results revealed that temporal context influences item memory in rats similar to the manner in which it influences item memory in humans. These results related to temporal context complement other results in which the pattern of hippocampal activity suggested that nonspatial items were also embedded in a spatial context (Manns and Eichenbaum, 2009).

Neuronal oscillations and memory

Hippocampus cover 

In addition to our interest in studying patterns of spiking activity, we also pursue investigations of oscilaltions (e.g., theta, gamma) in the local field potentials of the hippocampus and related structures. For example, one recent study (Trimper, Stefanescu, and Manns, 2014) focused on how neural synchrony in the hippocampus—particularly theta and gamma oscillations—related to successful memory performance by rats on a novel object recognition memory task described previously.  During exploration, gamma synchrony in the hippocampus (i.e., coherence between hippocampal subregions CA3 and CA1 at ~40 Hz) increased markedly and related to subsequent memory. Specifically, CA3-CA1 coherence was significantly higher during exploration of objects for which the rats subsequently showed good memory performance as compared to objects for which the rats subsequently showed poor memory performance. This subsequent memory paradigm controls for overt behavior at the time of the increase in gamma coherence and makes a strong case for the role of event-related CA3-CA1 synchrony specifically in the low gamma range for memory encoding for individual, trial-unique items. Further, we observed a novel relationship between memory and interactions between gamma synchrony and the hippocampal theta rhythm. In particular, we found that the portion of the theta cycle that contained the greatest levels of gamma synchrony was elongated during exploration of objects for which the rats subsequently showed good memory performance. Taken together, the results suggested that the shape of the theta rhythm mediated hippocampal gamma synchrony to the benefit of memory encoding. More broadly, the results provide insight into how hippocampal subregions dynamically couple during memory formation.

We recently extended this work by pursuing recordings from four major subregions of the hippocampus (DG, CA3, CA1, and subiculum) as rats performed an object-location recognition memory task (Trimper, Galloway, Jones, Mandi, and Manns, 2017). We found that spiking and local field potentials across the subregions of the hippocampus related to both behavior (e.g., object exploration versus locomotion) and memory performance. In particular, slow gamma activity--both in terms of power within regions and coherence between regions--correlated with associative object-location memory encoding. The goal of this overal line of research is to better understand the complicated dynamics of the extended hippocampal network as it relates to memory encoding and retrieval.


Muscarinic Acetylcholine Receptors and Memory

The muscarinic M1 acetylcholine receptor is a key target for potential drug therapies for many disorders and diseases of the brain, including Alzheimer’s disease. In one project (see Lebois et al., 2016), we have focused on two selective activators of the M1 receptor, the bitopic M1 agonist VU0364572 and the M1 positive allosteric modulator BQCA (we also included for comparison the current standard of care for Alzheimer’s disease, donepezil). We recorded spatially-specific “place cells” from the hippocampus of freely-foraging rats and asked how systemic administration of VU0364572, BQCA, and donepezil altered the spatial correlates of these hippocampal pyramidal neurons. The goal was to use place cells, one of the best studied correlates of behavior and memory in neuroscience, to assess how these drugs would impact spatial memory-related neural function at the network level in awake and freely-moving animals. We found that VU0364572 but not BQCA or donepezil increased the spatial correlations of hippocampal place cells when the walls of the recording enclosure were reshaped in 15-minute intervals (from square, octagon, hexagon, circle, then square again). The results suggested that VU0364572 favored a pattern-matching memory state often termed pattern completion, whereas for BQCA it favored an orthogonalizing memory state often termed pattern separation. These results are important because the findings show how two promising drugs impact hippocampal function in an awake and freely-foraging animal and indicate that two drugs that both target the M1 receptor can have different network-level effects. These results also have important implications for future development of drugs targeting memory dysfunction in aging and Alzheimer’s disease.






Department of Psychology

Emory Homepage