Scientists demonstrate technique that could reveal the mechanics of memory formation

A team of scientists led by Professor Hye Yoon Park have developed a new RNA imaging technique that has the potential to reveal how engram cells operate, which could lead to better understanding of memory and learning processes. The method is documented in a paper titled, “Real-time visualization of mRNA synthesis during memory formation in live mice” published in the Proceedings of the National Academy of Sciences (PNAS). The work involved researchers at the University of Minnesota and two institutions in South Korea.

Commenting on the technique, lead scientist Hye Yoon Park says, "The breakthrough that this work represents is a novel technology that enables real-time imaging of mRNA in the live mammalian brain. Because long-term memory formation requires new RNA synthesis, this technique will offer a powerful tool for studying the dynamics of memory in the live brain."

Richard Semon coined the term “engram” to describe the physical manifestation of a memory in the brain a little more than a century ago. Twenty years later, Donald Hebb proposed that memories are encoded in cell assemblies. However, it took several decades after these theoretical propositions for scientists to be able to experimentally validate long-term changes in neurons exposed to stimuli to result in what is known as memory trace or engram cells. Yet little is known about the dynamics of these cells as access to them for real-time monitoring has remained difficult, which in turn has made it challenging to understand the processes of learning and memory retrieval. 

Activity-regulated cytoskeletal gene (Arc) creates a protein critical to memory formation and consolidation. While it has been known for a while that the gene is a gateway to the understanding of cellular processes surrounding memory creation and consolidation, the exact mechanism by which it controls those functions has proven to be elusive. Arc is an immediate-early gene (IEG) and studies have shown that IEGs are transcribed within minutes of stimulation, which makes them widely used for the identification of neurons activated during learning and memory tasks. More recently, studies have also indicated that IEG expressing neurons are engaged in the formation of engram cells. However, the transient and rapid activation of Arc in response to stimuli poses a challenge in tracking its transcription, translation, and the intricacies of its role in learning and memory.

Although existing methods for imaging IEG expression have been able to capture snapshots of neuronal populations activated during memory encoding and retrieval, these methods are insufficient for continuous and real time monitoring in live animals. Considering the transient nature of the IEGs, the delayed expression of reporter proteins makes it difficult to accurately point to the types of neuronal activities that trigger IEG transcription at the level of the single cell. Methods that involve short-lived fluorescent proteins have also been unsuccessful. These proteins decay over hours to days, which complicates the identification of distinct neuronal populations triggered by specific stimuli with a time interval of less than a day. Other methods exist, but as they are tied down to a window of time that extends from a few hours to days, they run the risk of presenting results that overestimate the neuronal population connected to a single event. 

Tracking Arc expression

Park and her colleagues use a novel imaging technique to address the challenge of tracking IEG expression in real time. Using a genetically-encoded RNA indicator (GERI) mouse, in which every endogenous Arc mRNA was labeled with 48 green fluorescent proteins (GFPs) (the protein fluoresces to a bright green when exposed to light in the blue to ultraviolet range) and two-photon excitation microscopy, the team could visualize individual Arc transcription sites in vivo. The GERI mice were subject to contextual fear conditioning and tests were run to examine their fear memory retrieval over several days in the same context and a new context. The method yielded that Arc-positive neuronal populations had distinct dynamic properties in different regions of the brain. Delving deeper, the team went on to demonstrate that GERI can be used alongside a genetically encoded calcium indicator (GECI) to investigate neuronal activity in Arc-positive neurons in mice navigating a virtual environment. 

Dynamics of neurons in hippocampal and cortical regions

The GERI technique developed by the team allows scientists to directly study IEG transcription in real time which results in greater accuracy unlike prevailing methods involving reporter proteins. The technique has revealed results that contribute to a new understanding of the physiological aspects of memory encoding and retrieval. It has shown that Arc-positive neuronal populations in the CA1 region of the hippocampus and the retrosplenial cortex (RSC) operated in distinctly different ways. In the CA1 region, approximately the same fraction of neurons expressed Arc during recent and remote memory retrieval, but there was a significant change in the specific neurons involved within a span of two days. Contrast that with neurons in the RSC, where a small but significant fraction of neurons displayed Arc expression persistently over a significantly longer period of time of one month.

Results also showed a correlation between the persistently Arc-positive proportion of neurons in the RSC with the freezing rate in the mice. This indicates that these particular neurons play a role in maintaining contextual fear memory. Given the prevailing memory indexing theory (which suggests that ensembles of hippocampal neurons perform the indexing function, while those in the cortical region store information about the content) the results of the present study indicate that each cluster of the hippocampal Arc-positive neuronal population represented an updated index of each retrieval event, while the stable RSC Arc-positive population is reflective of the context which remained the same. The drop in the overlapping population of Arc-positive neurons also indicates the time limited role of the hippocampus in memory consolidation.  

The combination of GERI and genetically-encoded calcium indicator (GECI) imaging in virtual reality (VR) environment tests (where mice navigated a linear track in a virtual environment) indicated higher calcium activity in the Arc-positive population as compared to the Arc-negative population. However, there was a wide difference in the levels of activity within individual neurons in each group. The overlapping neuronal population that showed Arc positivity on days one and two of the VR test where the mice were exposed to the same environment showed higher calcium activity. The results indicate that neurons that showed Arc positivity on both days are involved in maintaining contextual memory. 

Bearing in mind that the Arc gene modifies the strength of synaptic transmission, the transcription and translation of the gene when the mice are first exposed to the fear conditioning stimulus might be sufficient for the creation of memory engrams. The results of the VR experiments indicating calcium activity levels suggest that the overlapping neuronal population that showed Arc positivity on the first two days of exposure to the VR environment were involved in encoding and retrieval of contextual memory (unlike the population that showed Arc positivity initially and then were Arc-negative on the second day). The team conclude that consistent Arc expression over multiple exposures plays a role in stabilizing the memory engrams.

Impact of the study

Reacting to the unique nature of the technique, RNA imaging expert Dr. Robert H. Singer of Albert Einstein College of Medicine says, “[It is] quite incredible! To be able to image mRNA longitudinally in the awake behaving mouse is such an advance.” 

The current study by Park and her team of scientists is significant on many counts. The GERI technique provides researchers a window into a live animal brain, which allows them to track immediate-early genes as they are expressed in real time. The team have also successfully demonstrated that the GERI technique can be used in conjunction with the GECI technique to study neuronal activity of Arc-positive neurons in mice that are awake. Overall, the success of their study establishes GERI as a reliable tool for tracking transcriptional activity of IEGs during learning and memory processes. Besides the technical significance of the work, the results of the study are also particularly salient for the new insights they offer regarding the distinct dynamics of Arc-expressing neuronal populations in different regions of the brain.

In addition to Park, the research team included Byung Hun Lee, Jae Youn Shim, Hyungseok C. Moon, Dong Wook Kim, Seoul National UniversityJiwon Kim, Jang Soo Yook, Jinhyun Kim, Korea Institute of Science and Technology. The research was supported by the Samsung Science and Technology Foundation and the Wellcome Trust.

Read the paper "Real-time visualization of mRNA synthesis during memory formation in live mice" at the PNAS site.

Learn more about Professor Hye Yoon Park's research interests.