Memory is Dependent on Synapses Between Engram Cells

What's the science?

Memory is thought to be encoded in the brain by a set or pattern of dispersed neurons in the brain called an ‘engram’. It has also been suggested that changes in the strength of the connections between neurons (i.e. synapses) in the brain is how memories are formed. This week in Science, Choi and colleagues used a new green fluorescent protein technique in mice to better understand how engram cells are connected in the brain and whether synaptic connections between engram cells underlie memory formation.

How did they do it?

Mice underwent a fear conditioning task, in which some mice received foot shocks (either weaker or stronger) during an experiment and some mice did not. Mice typically freeze due to fear when placed in a particular environment in which they remember experiencing something negative (such as a shock to their foot). Using doxycycline, they labeled the brain cells which were activated during the conditioning - these are referred to as the engram cells, because they are involved in memory during the fear conditioning task. Next, they designed a novel modified green fluorescent protein labelling technique called dual-eGRASP, which allows for visualization of the synapse between neurons, including identifying synapses formed by two separate groups of pre-synaptic (input) neurons as well as post-synaptic neurons. They then used this technique to visualize the connectivity (strength) of synapses in engram cells (labeled as active during fear conditioning) versus non-engram cells.

What did they find?

The dual-eGRASP technique successfully visualized synapses from separate populations of pre-synaptic neurons to hippocampal CA1 neurons from two different inputs, even when the synapses from the two populations were interspersed. When comparing engram-engram synapses versus synapses that involved non-engram cells (which were not labeled as active during fear conditioning), they found that dendritic spines were larger and more numerous on engram cells receiving input from other engram cells only. This indicates that synapses activated between engram cells during fear conditioning could cause synaptic connectivity between these cells to strengthen. When assessing the relationship between synaptic strength and memory, the authors found that synapses between engram cells were stronger and denser after receiving a stronger shock versus after receiving a weaker shock. This indicates that strength of a memory (due to strength of a foot shock) is related to synaptic connectivity and the strength of connections between engram cells in the hippocampus.

What's the impact?

This is the first study to show that engram neurons that are activated during a learning task (fear conditioning) are more likely to have stronger, denser synaptic connections with each other. We now know that the, strength of connections between engram cells in the hippocampus underlies memory formation and strength.

J. Choi et al., Interregional synaptic maps among engram cells underlie memory formation. Science (2018). Access the original scientific publication here.

The Association Between Cannabis Use and Cognitive Function

What's the science?

Recent policies in North America have shifted towards increased acceptance of cannabis use, and perceived harm has decreased. Chronic cannabis use could have effects on cognitive function in developing adolescents, however, this risk is not well understood. This week in JAMA Psychiatry, Scott and colleagues perform a meta-analysis of studies on cannabis use to determine whether there is an association between cannabis use and cognitive function in youth and young adults. 

How did they do it?

They performed a systematic review of the literature. They included all cross-sectional studies where participants were adolescents or young adults, frequent or heavy cannabis use was assessed as a variable of interest (not just acute cannabis use) and cognitive performance on at least one neuropsychological test was measured. They also ensured that each of these studies had sufficient power to calculate effect sizes. 69 studies met their criteria (2152 cannabis users and 6575 control participants with minimal cannabis use). They performed the statistical analysis using a mixed effects multivariate model, which accounts for correlated within-study effects and multiple different effect sizes between studies. Cognitive performance on neuropsychological tests was the measure of interest.

What did they find?

The overall effect size in a model without explanatory variables (i.e. variables that could potentially affect the association between cannabis and cognitive function such as age) was small but significant, showing a lower cognitive performance in cannabis users compared to the comparison group. Cognitive performance in heavy cannabis users was lower in the domains of learning, executive function, information processing speed, delayed memory, inhibition, working memory and attention. When explanatory variables were included, the effects were not explained by age category, age at first cannabis use or alcohol use. Hours of abstinence from cannabis (self-reported) was associated with a lower effect size, suggesting that a longer period of abstinence from cannabis reduced the negative effects on cognitive performance. When looking at studies where a period of greater than 72 hours of abstinence prior to study participation was required, effect size was no longer significant.

What's the impact?

This is the first meta-analysis of the effects of heavy cannabis use specifically in youth and young adults. This study confirms findings from studies in adults, suggesting that in youth, the effects of cannabis use on cognition are small and abstaining from cannabis for at least 3 days, may reduce negative effects on cognitive performance.

A Word of Caution: Studies which take into account the type and amount of cannabis used will be needed to better understand the effects on cognitive performance. Additionally, studies looking at functional outcomes like employment or other health measures (not just cognitive performance) with cannabis use over time may provide more meaningful information. Furthermore, even a small effect on cognitive performance may impact youth differently depending on their genetic makeup or life circumstances.

C. Scott et al., Association of Cannabis with Cognitive Functioning in Adolescents and Young Adults: A Systematic Review and Meta-Analysis. JAMA Psychiatry (2018). Access the original scientific publication here.

Brain Activity Predicts How Much Individuals Like Each Other

What's the science?

Within a group, people tend to like each other over time as they get to know one another. When considering interpersonal factors, if Person A likes Person B, this can lead Person B to like Person A over time (reciprocal liking). Interpersonal reward might affect this process; if a person experiences something rewarding during a social encounter, this can further influence reciprocal liking (a positive feedback loop). Two brain regions, the ventromedial prefrontal cortex and the ventral striatum are known to be involved in processing reward. In some cases, activity in these brain regions can predict an individual’s preferences, and therefore may also be able to predict how much an individual will like someone else. This week in PNAS, Zerubavel and colleagues tested whether interpersonal liking and brain activity in reward-related regions could predict how much people liked each other after a period of time.

How did they do it?

Sixteen healthy young adults, most of whom did not previously know each other, were recruited for the study while volunteering for a summer program that took place over nine weeks. At two timepoints, once at the beginning of the summer program and again 9 weeks later, participants rated how much they liked each other ‘not very’ to ‘very’, 0-100 scale, and also viewed pictures of each other's faces while undergoing functional magnetic resonance imaging (fMRI) scans. The authors employed structural equation modelling (which can account for multiple interrelated variables) to assess liking cross all possible pairs of participants. Each participant was in turn referred to as the ‘actor’ and other participants were referred to as ‘partners.’ The authors assessed: a) the effects of both the actor and the partner’s liking of each other at timepoint 1 on the actor’s liking of the partner at timepoint 2 and b) the effects of both the actor’s and the partner’s neural responses in the ventromedial prefrontal cortex and ventral striatum at timepoint 1 on how much the actor liked the partner at timepoint 2 (i.e. whether either person’s brain activity at timepoint 1 predicted how much the actor liked the partner, controlling for how much they reported liking each other).

What did they find?

The authors found that how much the actor liked the partner, and how much the partner liked the actor at time point 1 predicted how much the actor liked the partner at timepoint 2. Next, after accounting for how much the partner and actor liked each other at timepoint 1 (and other known predictors of future liking) the authors found that neural responses in the ventromedial prefrontal cortex and ventral striatum in both the actor and the partner (brain activity in response to viewing each other's face) at timepoint 1 predicted how much the actor liked the partner at timepoint 2. This indicates that neural processes leading to future liking could be occurring subconsciously.

What's the impact?

This is the first study to demonstrate that fMRI responses (measuring brain activity) in reward-related regions of the brain can predict how much one individual will like another individual months in the future. Specifically, neural responses in both people can affect how much one person likes the other person at a later timepoint. The study furthers our understanding of the neural basis for interpersonal relationships.

N. Zerubavel et al., Neural precursors of future liking and affective reciprocity. PNAS (2018). Access the original scientific publication here.