Individual Variation in Amygdala Connectivity

Post by Deborah Joye

What's the science?

The amygdala is a part of the brain most known for its role in aggression and fear. Amygdala function is affected in many psychiatric illnesses including post-traumatic stress disorder, anxiety, depression, and phobias. Most in-depth studies of the amygdala in humans measure brain signals averaged across a group of individuals to determine the location and connections of the amygdala. While this approach has helped us to better understand the amygdala in general, it has limited our ability to tailor treatment of amygdala dysfunction to individual patients. This week in PNAS, Sylvester and colleagues use extensive functional magnetic resonance imaging (fMRI) of individuals to characterize three functional subdivisions of the amygdala and their specific patterns of connectivity with other networks in the brain.

How did they do it?

The authors analyzed over 5 hours of fMRI data per individual from 10 individuals to determine different amygdala subdivisions based on activity patterns within the amygdala and associated activity in other cortical regions. The authors then used both group-averaged and individualized data to demonstrate that group-averaged analyses can obscure the specific locations of amygdala regions and mask their functional patterns. The authors compared their amygdala subregions from the individualized dataset against a much larger independent dataset to understand whether amygdala subdivisions and their connectivity patterns were roughly consistent across people. Finally, the authors investigated possible differences in the timing of activity across the amygdala and other cortical networks to investigate whether individual differences exist in the timing of amygdala-cortex connectivity.

What did they find?

The authors characterized three subdivisions of the amygdala that are roughly consistent across individuals with some differences in spatial location. The authors also found that each subdivision of the amygdala had its own unique connections to other brain networks and that the magnitude of these connections varied from person to person. One subdivision was anatomically superior and preferentially connected to the default mode network, a widespread neural network important for reflecting on the self and others, as well as thinking of the past and future. This amygdala connection’s role could be to integrate important environmental information with an individual’s past history regarding the emotional significance of that stimuli. Another subdivision was anatomically medial in most people and preferentially connected to the dorsal attention network which is active during attention-demanding tasks. This amygdala connection’s role might be the top-down modulation of attention networks.

The last subdivision was anatomically ventral and did not show a preferential connection to a specific neural network but had connectivity properties that were shared across the rest of the amygdala. When the authors compared these findings with a larger, publicly available dataset they found similar amygdala subdivisions, but the selectivity of each subdivision for particular neural networks was much weaker compared to individual analyses. Lastly, the authors found that the timing of activity between amygdala subdivisions and other neural networks was consistent across both datasets, suggesting that though location and magnitude of amygdala connections may vary from person to person, the networks themselves are consistent.

What's the impact?

This study is the first to use extensive fMRI from individuals to demonstrate that three distinct subdivisions of the amygdala are roughly consistent across people, but with important individual variation in location and magnitude of connectivity. The study also revealed that subdivisions of the amygdala can have preferential connectivity with specific neural networks, providing a framework for a more detailed understanding of how the amygdala interacts with other brain regions in individual patients. These findings could lead to improvements in personalized psychiatry and potential therapeutics for amygdala dysfunction.

Sylvester et al., Individual-specific functional connectivity of the amygdala: A substrate for precision psychiatry, PNAS (2020). Access the original scientific publication here.

Threshold for Odor Detection Adapts Based on Past Experience

Post by Shireen Parimoo

What's the science?

Animals react to sensory input from the environment, but sometimes the input isn’t strong enough to elicit a behavioral response. How much sensory input is needed for organisms to detect it? Several models attempt to explain how external sensory information, like sound, is detected in the brain. For example, the absolute threshold model proposes that a sound will be detected once it reaches a certain intensity (i.e. the threshold). According to the derivative model, the rate at which a sound’s intensity changes will determine when it is detected, whereas the fold change model posits that detection depends on how much the sound changes in proportion to its original intensity. Although these models have been applied to explain sensory detection in various organisms and across different modalities, no study has directly compared them with each other. This week in Neuron, Levy and Bargmann used computational modeling and calcium imaging to develop a unified model for odor detection in Caenorhabdtis elegans (roundworms).

How did they do it?

Roundworms have a simple nervous system that makes it possible to record the activity of specific neurons. The authors measured the sensory activity of an olfactory neuron called AWCON in response to changes in levels of the odorant butanone. Specifically, using a microfluidic setup, AWCON calcium activity was recorded in immobilized animals across a wide range of odor concentrations and timescales. Neuronal activity and navigation decisions were also examined in animals freely moving in odor gradients controlled by a specialized microfluidic device.

The authors rigorously tested many models that predict neuronal activity features (such as neuronal response and latency of response) and navigation behavior, including the absolute threshold, derivative, and fold change models. They also created an adaptive concentration threshold (ACT) model in which sensory activity is initiated when the odor concentration reaches a threshold, however, this threshold is continuously adapting to the odor. The ACT model includes (i) a threshold constant, which changes the neuron sensitivity, and (ii) adaptation time, which determines how long is the neuron memory of the external information. To determine whether the ACT model is generalizable, it was also tested on a separate dataset of neuronal activity in zebrafish in response to visual input. To identify the molecular basis of sensory detection, they examined the role of EGL-4, a protein kinase in the AWCON neuron that is involved in olfactory learning. They compared the effect of butanone concentration in loss-of-function mutants without functional EGL-4, gain-of-function mutants with enhanced EGL-4 activity, and wild-type animals. Finally, they performed theoretical studies to determine which model can allow both accurate and fast sensory responses, two key features for sensory neurons performance. 

What did they find?

Previous models did not adequately predict the observed neuronal responses and latencies, and could only match a subset of the experimental observations. For instance, they found that calcium responses depend on butanone concentration and the rate of concentration change, inconsistent with the absolute threshold and the derivative change models. The ACT model, on the other hand, predicted neuronal responses for both slow and fast changes in butanone concentration. The ACT model also predicted neuronal activity and aversive navigation decisions, like reversals and pauses, in more natural conditions, while animals freely navigated in odor gradients. This indicates that odor sensation and navigation are driven by an adaptive threshold mechanism that allows a comparison of past and current sensory inputs.

Interestingly, loss of EGL-4 function elongated the threshold adaptation time relative to wild-type animals and enhancement of EGL-4 function shortened it, suggesting that the protein kinase EGL-4 tunes the adaptation time of the sensory detection threshold. The ACT model also predicted activity in the optic tectum of zebrafish in response to visual input, demonstrating generalizability. Finally, the authors show that in contrast to alternative models, an adaptive-threshold mechanism allows sensory neurons to respond both fast and accurately to external stimuli, highlighting its benefit in reliable environment sensation.  

What's the impact?

Combining computational modeling with quantitative assays, this study is the first to systematically compare previous sensation models and to demonstrate how sensory detection is driven by a combination of current and past sensory inputs from the environment. The ACT model is powerful because it encompasses elements of previous models under different conditions and further generalizes to visual stimuli. These findings pave the way for future research to uncover the neurobiological basis of sensory detection and test the generalizability of the model across organisms and sensory modalities. 

Levy & Bargmann. An adaptive-threshold mechanism for odor sensation and animal navigation. Neuron (2020). Access the original scientific publication here.

Learning from Your Mistakes: The Role of Dopamine Activity in Prediction Errors

Post by Lincoln Tracy 

What's the science?

Understanding how associative learning occurs in the brain is one of the most important questions in neuroscience. One of the key concepts in associative learning relates to the idea of a prediction error — a mismatch between what we expect to happen and what actually happens. Both humans and animals use prediction errors to learn; the greater the error, the greater the learning. Prediction errors can be calculated using the method of temporal difference. The ability to map millisecond by millisecond changes in neuronal dopamine firing activity has been a major step forward in understanding prediction errors. However, there are still aspects of prediction errors that are yet to be fully explored. Previous research has demonstrated that optogenetics can be used to shunt—or attenuate—neuronal dopamine activity to prevent learning about a reward when it is delivered. This week in Nature Neuroscience, Maes and colleagues used second-order conditioning to determine whether blocking or shunting neuronal dopamine activity with laser light when a visual cue that predicts a reward is presented prevents learning from occurring in a similar fashion.

How did they do it?

The authors took rats whose genome had been altered to express Cre recombinase, an enzyme derived from bacteria, that could be controlled by a tyrosine hydroxylase promoter. The rats underwent surgery, where a Cre-dependent viral vector carrying halorhodopsin was injected into the ventral tegmental area (VTA) of the brain. Optic fibers were also implanted into the VTA; these would be targeted by the lasers during optogenetic stimulation. The rats were then placed on a food-restricted diet for four weeks before they were conditioned to associate a specific visual cue (stimulus A; a flashing light) with a reward (a chocolate-tasting sucrose pellet). After the training period, the rats completed two experiments; a second-order conditioning experiment and a blocking experiment. In both experiments, the percentage of time the rats spent approaching the food port where the pellet was delivered was taken as a measure of how conditioned they had become. The second-order training experiment had two types of trials. In both types of trials, the previously conditioned cue (the flashing light) was used to reinforce learning about two novel cues. That is, a second novel stimulus, either a chime (stimulus C) or a siren (stimulus D), was presented after the flashing light. In the C trials, continuous laser light was beamed onto the VTA half a second before the presentation of the flashing light so to disrupt the dopamine transmission that would normally occur when the reward predicting cue was presented. In the D trials, the light was beamed onto the VTA at a random time point after the flashing light was presented. Following the training, the rats also completed probe testing, where the chime and siren were presented without a reward. The authors then compared the behavioral response between the two trial types to determine if disrupting dopaminergic transmission impacted learning.

In the blocking experiment, the conditioning cue (the flashing light) was presented in separate compounds with each of two novel audio stimuli, a tone (stimulus X) or a click (stimulus Y). Each of these compounds was paired with reward. Normally, under these conditions, the conditioned light blocks learning about the relationships between X (or Y) and the reward. The question was, if the conditioned cue carries information about the prediction of up going reward, then disrupting this prediction would prevent the light from blocking learning about X. To test this, the laser light was beamed onto the VTA in the X and Y trials at the flashing light or at a random time point between trials, respectively. Learning these compounds was compared to a compound that consisted of a non-conditioned steady light and another audio cue, a white noise (stimulus D), which was also paired with a reward. The rats underwent probe testing following the blocking (compound) training, where the X, Y, and Z stimuli were presented alone and without a reward.

What did they find?

Optogenetic manipulation did not alter responding during second-order training. However, during the probe test, the rats responded to the D stimulus more frequently than the C stimulus. These results indicate that attenuating dopaminergic activity at the start of the reward-predictive cue prevented second-order conditioning from occurring to stimulus C. Similar to the second-order experiment, optogenetic manipulation did not alter responding during blocking training. During the probe test of the blocking experiment, the rats responded more to the control stimulus, Z, compared to the blocked stimuli, X or Y. These results confirmed that the conditioned flashing light was able to block learning about the novel cues, X and Y, but showed that attenuating the dopamine signal to the flashing light did not disrupt the ability of this stimulus to block learning about the novel stimulus X, suggesting that the dopamine signal to good predictors of reward represents a prediction error and not a prediction about reward.   

What's the impact?

This study provides clear evidence that the increases in firing activity of dopaminergic neurons following the presentation of a reward-predicting cue serve as prediction errors to support associative learning in a similar fashion to the previously shown reward-evoked changes in dopaminergic firing. Importantly, these findings suggest a broader role for dopaminergic signaling in driving associative learning than what is thought in current theories. 

Maes et al. Causal evidence supporting the proposal that dopamine transients function as temporal difference prediction errors. Nature Neuroscience (2020). Access the original scientific publication here.