Volume 26 Issue 1, January 2023
Deep reinforcement learning to study combinatorial expansion of a behavior repertoire
Despite rich behavioral evidence, it is unclear how the brain expands its behavior repertoire. By building theoretical models with a deep reinforcement learning algorithm, I show that the brain composes a behavior to solve a novel task by combining previously acquired skills and augmenting their variability.
Confronting racially exclusionary practices in the acquisition and analyses of neuroimaging data
The use of field-standard approaches in neuroscience and psychology can exclude participants from research, biasing our understanding of brain–behavior relations. Here the authors discuss how we might address inequity in our scientific methodology.
Microglia ferroptosis is regulated by SEC24B and contributes to neurodegeneration
Iron-laden microglia assume a disease-relevant, ferroptosis-associated signature and cause neurotoxicity. CRISPR screen uncovered regulators of ferroptosis in microglia. This ferroptosis–microglia–neurodegeneration axis could be targeted therapeutically.
An RNA-targeting CRISPR–Cas13d system alleviates disease-related phenotypes in Huntington’s disease models
Leveraging RNA-targeting CRISPR–Cas13d technology, Morelli et al. engineered a novel therapeutic strategy that safely and effectively eliminates toxic expanded huntingtin RNA in multiple models of Huntington’s disease.
Ketamine triggers a switch in excitatory neuronal activity across neocortex
Cichon et al. show that ketamine induces a switch in the active population of excitatory neurons across cortical layers and regions that contributes to impairments in sensory processing characteristic of a dissociative-like state.
Serotonin regulation of behavior via large-scale neuromodulation of serotonin receptor networks
Combining brain imaging, genetics and behavioral data, this study shows that distinct brainwide distributions of serotonin receptors explain the known division of serotonin effects on human impulsivity and aversive processing.
Intracellular chloride regulation mediates local sleep pressure in the cortex
Alfonsa et al. show that wakefulness causes shifts in cortical EGABAA, weakening synaptic inhibition and resulting in markers of local sleep pressure, and identify Cl− regulation as a link between sleep–wake history, cortical activity and behavior.
Early adversity promotes binge-like eating habits by remodeling a leptin-responsive lateral hypothalamus–brainstem pathway
Early-life trauma is a risk factor for binge eating and obesity later in life. Shin et al. identify a hypothalamus–brainstem circuit in mice that underlies the early-life trauma-induced binge-like consumption of high-fat diet and obesity-prone characteristics.
Disinhibition of the orbitofrontal cortex biases decision-making in obesity
This article shows that reduced inhibitory tone in the lateral orbitofrontal cortex of obese mice impairs their ability to use the value of an outcome to guide behavior. This could explain why access to obesogenic food biases behavior toward eating beyond satiety.
The role of the lateral orbitofrontal cortex in creating cognitive maps
Animals form cognitive maps of the world to guide behavior. This study shows that the lateral orbitofrontal cortex is essential for creating precise, outcome-specific cognitive maps during initial learning, but not for general map creation in itself.
Region-selective control of the thalamic reticular nucleus via cortical layer 5 pyramidal cells
Hádinger et al. found selective innervation of the thalamic reticular nucleus (TRN) by frontal cortex L5 cells. The spike output of TRN provides a precise readout of the cortical population activity in thalamic regions connected to frontal cortex.
Hippocampal spatial representations exhibit a hyperbolic geometry that expands with experience
Most spatial navigation studies assume a linear map of space. This study reports that neural representations of space follow non-linear hyperbolic coordinates. These representations also expand with experience along a maximally informative limit.
Arithmetic value representation for hierarchical behavior composition
Using mice and artificial deep reinforcement learning agents trained in the same task, it is discovered that composition of a novel behavior entails a simple arithmetic operation on action values of constituent subtasks and their stochastic policies.
Integrative transcriptomic analysis of the amyotrophic lateral sclerosis spinal cord implicates glial activation and suggests new risk genes
Humphrey et al. analyzed genetic and gene expression data from the postmortem spinal cords of patients with amyotrophic lateral sclerosis (ALS), observing changes in cell type composition and highlighting new risk genes.
Increasing diversity in connectomics with the Chinese Human Connectome Project
Brain images from the Chinese Human Connectome Project (CHCP) are now publicly available to facilitate transcultural and cross-ethnic brain–mind studies. Comparisons found reproducible brain parcellations but most differences were in language processing.