How memories are processed and where they are stored have long been the subject of scholarly inquiry. In ancient times, the Greek philosopher Aristotle wrote a treatise called ‘De memoria et reminiscentia’ that explored the properties of memory, positing that experiences are stored in the soul and later retrieved for further use (Jorge, 2013). However, it was not until the early 20th century that the concept of memories being stored in specific areas of the brain, termed “engram”, was proposed by Richard Semon (Semon and Simon 1921). Nevertheless, Semon's theory faced limited acceptance, as the neuroscientific knowledge of that era rendered it challenging to empirically verify the physical traces of memory (Smith 2003). However, recent advancements in neuroscience, driven by the ability to record the activity of individual nerve cells and their populations, have facilitated a deeper and more precise understanding of the mechanisms underlying memory storage.

Memory is defined as the ability to retain and recall information about past experiences, and consists of three basic processes: encoding, storage, and retrieval (Earhard 1969). Therefore, in order to know how memories are processed in the brain, it is first necessary to understand how the neurons that constitute the neural circuits respond to external inputs. A population of neurons involved in a particular computation is called a neural ensemble (Guzowski et al., 1999, Leutgeb et al., 2004, Sorensen et al., 2016, DeNardo et al., 2019, Hansel and Yuste, 2024). If memory traces remain in neurons that respond to external inputs, those neurons will undergo a series of physical/chemical changes. Neurons that have undergone such changes can be reactivated during the memory recall. In addition, memory recall could be artificially induced through optogenetic activation of the corresponding neurons (Liu et al. 2012). Those neurons that retain traces of memory are called engram neurons (Josselyn et al., 2015, Tonegawa et al., 2015, Poo et al., 2016, Kitamura et al., 2017, Josselyn and Tonegawa, 2020, Han et al., 2022, Ortega-de San Luis and Ryan, 2022, Williamson et al., 2024). Whether these traces persist throughout neurons or within subcellular structures such as synapses remains an actively investigated topic. Considering the current engram studies, evidence is being discovered that the engram of a single memory is distributed across multiple regions of the brain rather than confined to a specific area (Dixsaut and Gräff, 2022, Roy et al., 2022, Franceschini et al., 2023). Recent studies on engrams provide evidence that these neural engrams are not immutable but can undergo modifications over time (DeNardo et al., 2019, Cho et al., 2021, Ryan et al., 2021, Lee et al., 2023, Refaeli et al., 2023, Williamson et al., 2024). A recent study revealed that glial cells, which are non-neuronal cells that support and protect neurons in the nervous system, can also contribute to the formation of engrams (Williamson et al. 2024).

In this article, we will summarize research on how to observe neural ensembles, then discuss important studies on engrams and essential elements for the study of engrams. We will also discuss what areas will be pioneered to engram research in the future and how engram research will be conducted to address our unsolved questions about learning and memory.