Neural Newsletter | Dec 30 - Jan 5
Network energy landscapes, social media negativity, and dendrite-glia interactions
Article 1: The brain selectively allocates energy to functional brain networks under cognitive control
Keywords: Network energy, Cognitive control, Executive functions, Functional connectivity, Canonical functional networks, Structural balance theory, Brain biomarker, Predictive modeling
“The assessment of brain network energy through the lens of structural balance theory provides valuable insights into the organization of functional connections… These energy dynamics can illustrate the brain’s capacity to shift between stability and instability, essential for managing the competing demands of efficiency and flexibility.”
- Saberi, M., Rieck, J.R., Golafshan, S. et al., Scientific Reports, December 30, 2024
Background and Methods
An exciting new framework is transforming how researchers conceptualize functional neuroimaging data. Structural Balance Theory, with its roots in statistical physics, imagines the set of all possible brain states as a jagged energy landscape. The precarious peaks of this landscape represent unstable, high-energy patterns of connectivity between brain regions, which need only a slight nudge to shift into a new functional configuration. The troughs in this landscape represent stable, low-energy patterns that are easier to get stuck in, requiring stronger stimuli to induce a change in brain state. In this context, “energy” is an abstract representation of the instability of connections between brain regions, and does not directly represent the brain’s metabolic expenditure in any particular functional configuration. However, low-energy states generally represent efficiency in neural processing and low conflict between connected brain regions, while high-energy states represent more tension between connected regions and allow for more immediate adaptability into different functional patterns.
A paper from Saberi et al., published last week in Scientific Reports, applied this energy modeling framework to investigate the functional organization of the brain under different cognitive demands. In particular, the researchers measured the network energy of established brain networks at rest and during tests of working memory (n-back), inhibitory control (go/no-go), and cognitive flexibility (shifting task). Additionally, they tested whether incorporating network energy as a feature in existing models could reliably improve their performance. The study utilized fMRI data from a publicly available dataset comprising 144 participants aged 20 to 86.
Results and Conclusions
The study revealed significant differences in whole-brain energy levels across cognitive tasks. As anticipated, the resting state exhibited the lowest energy configuration, reflecting the efficient operation of the whole-brain network. In contrast, task states corresponded to increased total energy, indicating a brain-wide increase in cognitive flexibility. Modifying task difficulty did not significantly affect total energy levels, meaning that the energy change was specific to the type of task and not the complexity of the task.
After examining the whole-brain network, the researchers refined their investigation to 10 canonical networks: auditory, visual, dorsal attention, ventral attention, frontoparietal, cingulo-opercular, somatomotor, salience, default mode, and subcortical. All of these networks were at their lowest energy configuration during the resting state, though their energy levels varied relative to one another. For example, networks involved in low-level sensory processing (e.g., visual, auditory, and somatosensory) had lower energy levels at rest compared to networks that are engaged by more complex cognitive functions (e.g., frontoparietal, default mode, salience, and ventral attention). The authors propose that higher-order networks may intrinsically operate at elevated energy levels to enable greater flexibility and responsiveness to situational demands.
Task-specific differences in network energy were also observed. For example, the energy configurations of the dorsal attention and subcortical networks were especially low during the cognitive flexibility task, highlighting their role in attention-shifting. The frontoparietal network exhibited only a modest energy increase during the working memory task compared to the resting state, suggesting its efficient at processing working-memory demands compared to other networks. The visual and auditory networks displayed the greatest energy increases during the task states compared to the resting state, emphasizing that these systems must become highly flexible to process task-relevant sensory stimuli.
Finally, the team sought to investigate whether network energy could be a valuable input variable for predictive classification and regression models. Their findings show that including network energy in their machine-learning models significantly improves accuracy in classifying subjects by task-state and predicting subjects’ age from functional connectivity data. Ultimately, Saberi et al. successfully demonstrate how the brain allocates energy across networks in response to specific cognitive demands, and emphasize network energy as an informative feature for enhancing predictive insights from functional neuroimaging data and for conceptualizing brain connectivity on a broader scale.
Article 2: Opening the Pandora Box: Neural processing of self-relevant negative social information
Keywords: Information seeking, Social feedback, Social incentive delay, EEG
“Our findings challenge prior evidence suggesting that humans instinctively avoid aversive stimuli, and they shed light on the neurophysiological mechanisms that may underlie this counterintuitive behavior. Receiving self-relevant negative feedback appears to be both rewarding and salient, demonstrating that negative information about oneself can engage attention and curiosity under certain conditions.”
- Nicolaou S., Vega Moreno D., Marco Pallarés J., Biological Psychology, December 30, 2024
Background and Methods
The urge to gossip is a central part of human social behavior, but whether people will actively seek out negative information about themselves remains less understood. In an time when immediate social feedback is more accessible than ever, researchers from the University of Barcelona conducted a timely study that used EEG recording and a Social Incentive Delay paradigm to reveal counterintuitive findings regarding the brain’s response to negative comments on personal social media posts.
The study involved 30 healthy participants (21 women) with an average age of 23. These participants were told that photos from their personal Instagram profiles would be evaluated by a group of external volunteers, who would leave a positive comment if they liked the photo and a negative comment if they did not. They were told that each of their photos would be presented to four volunteers, and that performing well on a reaction time task would earn them the chance to view positive comments or avoid viewing negative comments.In reality, the comments were pre-generated by the researchers, consisting of 186 generic remarks rated on a 1-5 positivity scale by independent evaluators. To enhance believability, a smaller subset of personalized comments (e.g., “This sunset is beautiful!” or “Nothing new, everyone posts sunsets”) was crafted for each participant. At the end of the study, all participants reported believing the cover story.
To assess the brain’s reward response to positive and negative social feedback, participants performed a variation of the Social Incentive Delay task while undergoing EEG recording. This paradigm incorporates social rewards and punishments into a reaction-time task, enabling researchers to evaluate participants’ implicit motivations to receive different types of feedback. In the case of this study, participants were instructed to press the spacebar as quickly as possible when a white square appeared on screen. Before each trial, a visual cue indicated the type of reward they could receive: viewing a positive comment on one of their own or another person’s Instagram photo, or avoiding viewing a negative comment. The comments were displayed immediately after each trial. This paradigm aimed to uncover the implicit value participants assigned to each type of feedback by examining their reaction times and corresponding EEG patterns.
Results and Conclusions
The study revealed that participants were more motivated to view positive comments than to avoid negative ones on their Instagram posts. This was evidenced by slower reaction times during the negative information trials compared to the positive information trials, and is corroborated by the EEG results. The researchers examined the reward-positivity (RewP) and feedback-P3 (FB-P3) event-related potentials, which are electrical responses to reward and motivational salience, respectively. Intriguingly, the RewP response, which distinguishes aversive from reinforcing feedback, encoded negative comments about a subjects own photos as more rewarding than negative comments about a stranger’s photos. The FB-P3 response, which encodes the salience of a stimulus, was most pronounced when participants viewed negative feedback on their own photos.
These findings suggest that the brain interprets negative social information, such as unfavorable Instagram comments, as both rewarding and highly salient. This may represent a drive to seek out opportunities for self-improvement in the eyes of the community. It might also reflect a preference for attention over indifference, or it could be that the relative rarity of negative comments on one’s personal social media posts leads the brain to perceive them as more salient. Whatever the explanation, this study warns us that our brains may drive us towards information that could damage our self-esteem. In a digital environment where negative opinions are more accessible than ever, we must be intentional about relinquishing control to our biological drives.
Article 3: Glia detect and transiently protect against dendrite substructure disruption in C. elegans
Keywords: Glia, Cilia disruption, Dendrite substructure, Extracellular matrix, Neuron-glia interaction
“We demonstrate that disruption of C. elegans sensory neuron dendrite cilia elicits acute glial responses, including increased accumulation of glia-derived extracellular matrix around cilia, changes in gene expression, and alteration of secreted protein repertoire. Our studies reveal a homeostatic, protective, dendrite-glia signaling interaction regulating dendrite substructure integrity.”
- Varandas, K.C., Hodges, B.M., Lubeck, L. et al., Nature Communications, January 2, 2025
Background and Methods
With its 302 neurons, C. elegans is the simplest animal model in neuroscience research. Still, the nervous system of this microscopic worm continues to harbor biological mysteries that can inform future studies in higher organisms. In a recent paper published in Nature Communications, researchers from The Rockefeller University explored a novel function of C. elegans glial cells using mutant strains, RNA and genome sequencing, electron microscopy, and other advanced techniques.
Dendrites, the branching structures through which neurons receive chemical signals, are known to interact closely with glial cells across species. In humans, astrocyte endfeet often enwrap these structures to facilitate synaptic communication, though the full extent of this interaction is not understood. Some mammalian sensory neurons have cilia extending from their dendrites to encode mechanical signals, and disorders that negatively impact cilia structure can produce devastating sensory deficits.
The anatomical, cellular, and molecular features of C. elegans dendrites are largely conserved in mammals, making this organism an attractive model for studying glia-cilia interactions. The primary sensory organs in C. elegans are the amphids, which each contain sensory neurons with cilia at their dendritic tips. These cilia are enwrapped by the endfeet of amphid sheath glial cells (AMsh), which secrete proteins that form a thin extracellular matrix (ECM) around them. The researchers sought to determine how AMsh glia respond to mutations that disrupt cilia structure.
To investigate this glial response, the team utilized a che-2 mutant strain of C. elegans, which exhibits truncated sensory neuron cilia. They also employed CRISPR/Cas9 to attach a fluorescent protein to VAP-1, one of the proteins secreted by AMsh glia around sensory cilia, to visualize how these protein dynamics are altered in response to cilia disruption. They also performed RNA sequencing on AMsh glia in cilia mutants and wild-type worms to identify transcriptional changes.
The team also developed a molecular tool to induce acute cilia damage in wild-type worms, allowing them to compare glial responses to acute versus hereditary cilia disruption. Finally, genetic tools were employed to identify specific proteins involved in mediating this glial response and to propose a potential mechanism underlying glia-cilia signaling
Results and Conclusions
Using a comprehensive experimental approach, Varandas et al. identified a robust protective glial response to the disruption of dendritic cilia in C. elegans sensory neurons. They observed a significant accumulation of the AMsh glia-secreted fluorescent VAP-1 around cilia in mutant strains compared to wild-type animals, an effect detectable in larvae and persisting into adulthood. This was accompanied by a twofold increase in vap-1 transcriptional reporter expression, suggesting that AMsh glia alter their protein expression to deposit a large ECM around truncated cilia. RNA sequencing of AMsh glia revealed that cilia defects alter the expression of over 1,000 genes, many of which are involved in protein secretion.
An inducible system to acutely degrade cilia structure demonstrated that damage to multiple dendrite cilia triggers a robust glial response within hours. A forward genetic screen identified two key proteins mediating this response: DGS-1, a dendritic cilia transmembrane protein, and FIG-1, a glial transmembrane protein. Localization studies supported a model in which DGS-1 signals cilia integrity to FIG-1 via direct physical interaction, with disruptions in either protein triggering increased ECM secretion by glia.
To assess whether this glial response is protective, the researchers induced cilia defects in wild type C. elegans and in DGS-1 and FIG-1 mutants, which exhibit elevated glial ECM secretion around cilia. Using a dye-filling assay, they demonstrated that the glial response delays cilia damage, consistent with previous findings that certain secreted matrix proteins may enhance ciliogenesis.
Given the basic conservation of sensory organs and cilia structure, it is possible that astrocytic endfeet may similarly protect dendritic substructures in mammals. While future research is needed to identify functional analogs in more complex species, this study highlights a novel and elegant interaction between glia and dendritic cilia in one of the most established neuroscience model organisms.
Interesting…