In a new study Indiana University researchers observed episodic memory in rats to a degree never documented before, suggesting that rats can serve as a model for complex cognitive processes often considered exclusively human. Unlike semantic memory, which involves isolated facts, episodic memory involves replaying events in the order and context in which they occurred.
“The ability to replay a stream of episodic memories in context suggests that rats can serve as a model for complex cognitive processes,” said principal investigator Jonathon Crystal, the Eliot S. Hearst and Provost Professor of Psychological and Brain Sciences in the IU College of Arts and Sciences. “What we're systematically trying to do with this work is to push the envelope on what types of cognitive processes can be modeled in animals, particularly in rats, that are as complex and sophisticated as the things people seem to do effortlessly.”
Their findings, he believes, can ultimately enable researchers to further explore the biological mechanisms of memory, disorders of memory such as Alzheimer’s and dementia, as well as therapeutic interventions for these conditions. The findings also imply that these sophisticated forms of memory are evolutionarily quite ancient.
The study, “Rats replay episodic memories in context,” was published on January 23 in Current Biology.
Steppingstones to episodic memory
Asking rats what they remember, Crystal admits, is a somewhat painstaking process. Yet, on account of the rats’ extraordinary sense of smell, the researchers’ large collection of scents, four distinct settings/contexts, and an increasingly complex series of five experiments, the rats could tell the researchers precisely what they remember, as well as where and how they acquired the information.
In each of the five experiments the rats are presented with two separate sequences of scented lids, each sequence in a uniquely designed setting, the memory-encoding context. Through rigorous prior training, the rats were taught to search their memory for the third-to-last scented lid in a sequence encountered at one of the encoding contexts. The rats are then moved to one of two other uniquely designed locations, the memory-assessment context, each consistently associated with the same encoding context, where they are “asked” to find the third-to-last scent from the list. Because each sequence includes a different number of scented lids (from 4-11, randomly selected), the rats could only identify the correct one (which provides a pellet of food) in retrospect by mentally replaying their stream of memories.
Each experiment in the study builds on the one before, upping the ante on what the rat must remember to “win” the coveted morsel of food, and culminating in a final astounding feat. Yet, even at this final stage, what Crystal calls “the holy grail” of memory testing, the rats defied expectations.
IU graduate student and first author on the study Siyan Xiong condensed this fifth and final experiment into a single graphical abstract. Yet, to understand that experiment and the abstract, it is helpful to have some sense of the four that lead up to it.
Road to “the holy grail”
In the first experiment, for example, the rats learned to associate the specific memory-encoding contexts with the corresponding memory-assessment contexts, even when one sequence of scents is presented immediately after another, before either of the memory tests are given.
Experiment 2 further ensures the rats are attending to context by presenting two lists of odors followed by a slightly more challenging memory test in which the rats must distinguish between the third-to-last item in one context and the third-to-last item in the other.
A new challenge arises in Experiment 3 when researchers introduced two identical scents to both lists, as a third-to-last item and a non-third-to-last item in each context. Despite the same two odors appearing in both sequences, the rats successfully identified which scent was third-to-last in the specific context they were being tested on.
The challenge escalated in Experiment 4 when the lists were interleaved. Rats were moved back and forth between contexts, receiving pieces of one list and then the other. Remarkably, they still accurately picked out the target scent from the requested context.
Next up, the holy grail. In this final feat, the interleaved lists were interrupted by a 30-minute delay before the sequences continued. Despite the time gap and constant switching of locations, the rats remembered each sequence and its context apparently searching memories that bridged a long gap in time.
Each experiment gave the researchers greater confirmation than the one before that the rats were using episodic memory. “We previously knew that rats could remember lots of events, and the order of events,” said Crystal, “but did not know whether they knew how they came to that information, making it less relevant for human episodic memory.”
This study, however, confirms the degree to which rats do in fact replay episodic memories, remembering not just the flow of events, but how and where they acquire information. And by expanding what we know about non-human animal cognition, it pushes the limits on the ability of future research to model complex cognitive processes.
The project was completed with support from the National Science Foundation and the National Institute of Mental Health.
The College of Arts