Reviews & Analysis

The solutions found by neural networks to solve a task are often inscrutable. We have little insight into why a particular structure emerges in a network. By reverse engineering neural networks from dynamical principles, Dubreuil, Valente et al. show how neural population structure enables computational flexibility.

How and where negative affect is represented in the brain is a central neuroscientific question. A new study identifies neural correlates of both general negative affect and those specific to stimulus type by conducting multimodal functional MRI experiments.

Sleep disruption is a common but poorly understood feature of neurodevelopmental disorders including autism spectrum disorder. A study by Bian et al. reveals that sleep disruption in adolescent mice leads to long-lasting changes in social novelty preferences. Importantly, these perturbations can be restored through balanced actions in midbrain dopamine systems.

The extent to which neurogenesis occurs in adult primates is still controversial. Single-cell RNA sequencing, immunofluorescence staining, and ex vivo neurosphere culture experiments were performed using the adult macaque hippocampus. The results reveal robust adult neurogenesis in the dentate gyrus of the primate hippocampus.

Myelin, iron, and calcium are major constituents of brain tissue with magnetic properties that can be detected non-invasively using MRI. Using quantitative susceptibility mapping, we estimated the magnetic susceptibility of brain structures in 35,273 participants, creating a new resource to identify novel, non-invasive markers of brain health.

A new ‘meta-matching’ algorithm developed by He et al., published in this issue of Nature Neuroscience, enables small MRI datasets to piggyback on larger datasets to boost prediction accuracy. This innovation may aid in efforts toward personalized psychiatry.

Eating disorders are prevalent and, in far too many cases, fatal. This review covers advances in genetics, neuroimaging, and animal models, and encourages a more unified science of eating disorders.

Dopamine (DA) neurons in the ventral tegmental area bidirectionally regulate the activity of serotonin neurons in the dorsal raphe nucleus. Low-strength activity causes inhibition via dopamine receptor D2, whereas high- strength activity causes activation via dopamine receptor D1, and this circuit contributes to anorexia nervosa-like behaviors in mice.

Degeneration of dopaminergic neurons in the substantia nigra is a pathological hallmark of Parkinson’s disease (PD). However, not all dopamine-producing neurons degenerate. Kamath, Abdulraouf et al. find that there are ten transcriptionally defined dopaminergic subpopulations in the human substantia nigra, but only one carries significant PD genetic risk and is vulnerable to neurodegeneration in PD.

Activity-regulated myelination adaptively tunes neural circuit function in health. In rodent models of generalized epilepsy, recurrent seizures aberrantly increase myelination specifically within the seizure circuit. Blocking this seizure-induced myelination abrogates the progressive increase in seizure burden and ictal hypersynchrony that occurs in mice with intact activity-regulated myelination, indicating that maladaptive myelination can contribute to disease progression in epilepsy.

Two recent papers reveal that the brain can regulate its own immune responses by sending molecular cues to immune cells in the skull bone marrow via the cerebrospinal fluid. Furthermore, experimental spinal cord injury or bacterial meningitis specifically activate local vertebral and skull-resident hematopoietic cell injury responses.

Pettit, Yuan and Harvey find that hippocampal spatial maps degrade when mice voluntarily disengage from a navigation task, even without changes in sensory or self-motion cues. This finding suggests that internal state could have an active role in supporting navigational coding and, perhaps, spatial memory.

Human neuroscience methods (for example, electroencephalography, functional near-infrared spectroscopy and electrodermal response) are biased to exclude data from dark skin and coarse hair—traits common in Black people—and possibly people with racial trauma. We outline strategies to prevent a biased ‘unusable data crisis’.

An efficient tool for neurite tracing has been developed that reconstructed the complete axons of 6,357 individual projection neurons in the mouse prefrontal cortex (PFC). The resulting single-neuron projectome analysis revealed comprehensive PFC neuron subtypes, topographic organization of PFC axon projections, modular structure within the PFC and correspondence with single-cell transcriptomes.

Wang et al. used transcriptomic profiles of olfactory sensory neurons to determine the identity of their odorant receptors and map the location of their corresponding glomeruli on the olfactory bulb surface. The method enables high-throughput molecular mapping of the glomerular layout and opens up new venues to understand olfactory processing.

A study by Cook et al. shows that mice can use auditory feedback from their own actions to precisely time an interval. This feedback is processed in the secondary auditory cortex and regulates performance via a corticostriatal circuit.

As researchers uncover new roles for the microbiome in health and disease, recent studies have emerged linking the microbiome to aspects of aging. Mossad and colleagues demonstrate that a microbiome-mediated disruption of the intestinal barrier, associated with aging, can contribute to dysfunction of microglia in mice.

Palser et al. show that at the top 100 journals in psychology and neuroscience, male editors and those from the USA outnumber female editors and those affiliated with other countries, at rates significantly beyond their participation in the fields.

Complex and intelligent behavior depends not just on sensory evidence but also on internal cognitive state. Ashwood et al. use a powerful statistical method to identify hidden internal states in choice data.

Roy, Zhang et al. discuss how modern neuroscience is revealing underappreciated heterogeneity in thalamic cell types, which leads to the idea that ‘thalamic subnetworks’ provide a more appropriate level of functional description than thalamic nuclei.