Mitochondria are not uniform organelles. Across the brain, they exhibit profound molecular, biochemical, and functional diversity shaped by cell type, anatomical region, subcellular compartment, and lived experience. Recent advances in cell-type- and subcellular domain-targeted proteomics, transcriptomics, advanced live imaging, and functional biochemistry have begun to map this landscape with unprecedented resolution. Together, these findings challenge the conventional view of mitochondria as generic metabolic engines and position mitochondrial molecular diversity as a fundamental feature of brain organization, with direct relevance to behavior, aging, and neurological disease. This mini review synthesizes key recent studies in this field, highlighting their findings, methodological novelty, and significance, and formulates theories and hypotheses for future investigations.
Much attention has rightly been paid in recent years to the importance of rigorous experimental design and the use of appropriate statistics for data analysis. This is of tantamount importance in the evaluation of clinical trials and because new methods that generate large data sets may not be transparent without advanced analytical and statistical methods. Most findings in the biological literature are confirmed, either by the authors in the process of doing the work, or subsequently by the authors and others, who build on the findings. That said, it is also crucially important to recognize that failures to replicate are to be expected every so often, even when investigators are careful in studies of complex, degenerate, biological systems. Failures to replicate may not reflect issues with experimental design or interpretation, but may instead reveal new principles and provide important routes to new discoveries.
Across every scale at which we study the brain, from folded proteins and single neurons to cortical populations and the moving body, artificial intelligence (AI) has shifted from a bespoke tool into a driver of measurement and, increasingly, a generative engine for hypotheses. Here, I review recent progress (and open challenges) in applying AI for neuroscience along this scale axis: structure prediction for proteins, simulation-based inference for biophysical neurons, latent and dynamical models for neural populations, task-trained networks as minimal models of circuit computation, computer vision for animal behavior, neuromusculoskeletal modeling for biomechanics, and the multimodal, agentic systems now promising to automate discovery itself.
Communication is a natural social behavior that influences reproduction and survival; consequently, central questions in neuroethology involve discovering how activity in distinct neural populations regulates the performance, perception, and development of communication signals. Advances in the monitoring and manipulation of neural activity (including tools to record, stimulate, or inactivate specific neuron types) and the acquisition and analysis of communication signals (e.g., machine learning) provide unprecedented power to address fundamental questions about animal communication. Here we highlight recent insights into the neural mechanisms of communication, focusing on vocal development and plasticity in songbirds. We review recent discoveries about the sensory processing of birdsong, the contributions of inhibitory, excitatory, and neuromodulatory populations to song learning and performance, and the interplay between motor and sensory circuitry including the role of vocal practice. These studies not only deepen our understanding of behavioral plasticity but also challenge and expand models of song development, plasticity, and performance.
Classic experiments by Albert Bandura more than 50 years ago first illustrated that aggression is shaped not just by innate predisposition but also by social learning. In this work, human observers “mirrored” aggressive behaviors after watching aggressive demonstrators, showing that observation can lead to action imitation. Beyond imitation of aggression itself, aggression observation may be critical for updating internal models of the social world and promoting adaptive behavioral states. As the tendency to watch aggression is both ubiquitous and conserved, this highly ethological behavior is a useful framework for understanding the neural mechanisms underlying social learning and behavioral refinement across species. Recent work has shown that the observation of aggression recruits widespread neural activity in subcortical brain regions that are known to be active during self-directed attack and learning during naturalistic behavior. This suggests that social learning of aggression might recruit systems for aggressive motivation, arousal, and reward in addition to traditional “action mirroring” systems. Unlike other forms of social learning with a single demonstrator, aggression observation represents a more complex form of social learning because the observed behavior consists of multiple demonstrators (a winner and loser) whose unique goals must be simultaneously inferred by the observer. Several recent computational approaches, including multiagent reinforcement learning and inverse reinforcement learning, are promising strategies for formalizing how individuals might update behavior following aggression observation.
Statistical learning (SL) - the ability to detect patterns in sensory input without explicit instruction - is crucial for building internal models of the environment. In humans, it notably supports language acquisition, including word segmentation and grammar learning. Evidence across primates, songbirds, rodents, and insects indicate that SL is a widely shared evolutionarily conserved ability. The computational complexity of the mechanisms involved; however, varies between species, likely reflecting specific cognitive limitations. Alternatively, specific competences may have evolved to support the emergence of demanding, ecologically relevant, functions such as complex communication systems. These findings challenge the idea of a human-specific SL module while raising key questions about its evolutionary drivers and underlying mechanisms. Addressing these questions requires the expansion of cross-species, cross-modal studies with ecologically valid, unsupervised paradigms. Comparative approaches are indeed essential to uncover shared properties and species-specific SL adaptations. Framing SL as a foundational component of cognition, informed by animal research, offers new insights into brain function, and implicit learning in light of the evolution of sophisticated, emergent cognitive abilities such as complex communication systems.
Despite the phenomenological experience of vision being stable and uniform in both space and time, our visual system constructs this representation from markedly inhomogeneous spatial and temporal structures. While considerable progress has been made in understanding the inhomogeneities in spatial vision, much less attention has been given to the inhomogeneities in the temporal dynamics of vision. Specifically, little is known about how speed of processing varies across the visual field. In this non-systematic review, we focus primarily on psychophysical and neurophysiological studies of low-level photopic vision to elucidate how the speed of visual processing changes with eccentricity. Specifically, we examine key findings from major investigations on this topic, highlighting areas of convergence, points of divergence, and aspects that remain unresolved.
What is a natural behavior? I argue that the study of natural behaviors is often the study of the spontaneous behaviors of animals placed in quantifiably different environments. For behavioral generalists such as rodents, humans, and many other species, there may be no such definable construct as a native habitat or natural behavior, due to their successful abilities and needs to rapidly adapt to a wide range of different ecosystems. Instead of prioritizing naturalness, it may be more essential to determine objective outcome measures related to specific behaviors; i.e., which sequences of behaviors and adaptive mechanisms allow animals to survive and reproduce, across a range of dynamic or hazardous physical and social environments.
Autism spectrum disorder is a heterogeneous condition marked by social communication difficulties and restricted/repetitive behaviors. Although major progress has been made over the past two decades in understanding its genetics and molecular mechanisms, effective treatments remain limited. Report from the 2021 Lancet Commission on autism recommends that research should focus on improving quality of life through personalized assessment and intervention. Due to its heterogeneity, multiple treatment strategies will likely be needed. For some individuals, especially those with severe syndromic autism, gene therapy may offer future therapeutic options. To match patient subgroups to treatments, both "mutation clustering to treatment" forward approach and "treatment to disease subgroup" reverse approach can be used. Building a broad treatment portfolio will take time, but even incremental advances would be meaningful. Principles of neural plasticity, such as early intervention and repeated practice, may also enhance outcomes alone or alongside other therapies.
Computational models of visual system function are largely based on the mammalian visual hierarchy, as the exemplar of a large complex visual system. However, the cephalopod visual system provides an intriguing alternative model for vision; it is relatively similar in size, complexity, and acuity to that of mammals, but has a fundamentally different neural architecture. While this system has been largely overlooked relative to more standard model species, renewed interest in the diversity of neural computations across species, along with technical developments allowing for further investigation, have led to insights into this evolutionarily distinct visual system. Here we provide a brief overview of the unique architecture of the cephalopod visual system, and review recent advances in understanding its neural circuitry, visual coding, and potential computational mechanisms. Throughout, we highlight how investigating cephalopod visual processing provides the opportunity to identify both shared and novel principles for visual computations and their neural implementation.
Animals exhibit remarkable flexibility in their behavior. Such adaptability is tuned by integrating evolving physiological internal states, external contexts, and previous experiences. In this review, we aim to summarize recent findings on the circuit mechanisms of the reciprocal modulation of sensory processing and behavior through internal states, external contexts, and behavioral actions. We discuss how metabolic states reconfigure odor valuation, how social contexts generate motivational states that redirect action selection, and how behavioral execution recursively modulates sensory perception to drive adaptive decision making, toward a multiscale mechanistic framework of the interplay between states, contexts and actions.
Context-dependent decision-making enables flexible behavior by allowing identical sensory inputs to guide different actions depending on memory, rules, or goals. Recent advances in large-scale neural recordings have shifted the focus from single-neuron tuning to population-level representations, revealing principles by which neural populations support such flexibility. Here, we review evidence of how context-dependent decisions are implemented by population coding mechanisms, including nonlinear mixed selectivity, task-dependent population geometry, shared representational subspaces, and structured across-neuron correlations. Nonlinear mixed selectivity expands representational dimensionality, allowing downstream readout of arbitrary combinations of task variables. Learning reshapes population geometry, and it may balance flexibility and generalization by promoting the reuse of shared representations when task components overlap. Structured correlations between neurons that share a projection target enhance transmission of context-dependent information to downstream circuits. These population-level coding mechanisms provide a conceptual framework for understanding how neural circuits integrate sensory and contextual information to guide behavior.
Social dysfunction is a prominent feature of many neuropsychiatric conditions, yet its biological basis remains incompletely understood. Progress has been limited in part by the difficulty of experimentally linking genes, circuits, and hormones to flexible social behavior. African cichlid fish offer a powerful and underutilized vertebrate system for addressing this gap. We highlight the African cichlid Astatotilapia burtoni, in which social status is dynamic and rapidly reversible, enabling direct analysis of socially driven changes in behavior, physiology, and brain state. These transitions engage conserved components of the vertebrate social decision-making network and are amenable to genetic and endocrine manipulation. Recent advances in CRISPR/Cas9 gene editing and transcriptomic profiling establish cichlids as tractable models for dissecting the control of social behavior and probing mechanisms underlying social dysfunction. By combining ecological validity with experimental precision, cichlids provide a unique entry point for identifying conserved mechanisms relevant to human social dysfunction.
Traditional laboratory assays are limited in capturing the full range of evolved brain function, particularly in the domain of social behavior. While laboratory approaches offer control and causal precision, they constrain how animals interact through artificial groupings and the elimination of sociospatial structure. Here, we outline an emerging complementary approach-field neuroethology-which investigates neural mechanisms of behavior and their socioecological consequences in organisms living within semi-natural or natural contexts. We define its aims, highlight promising domains for its application, and note the technical innovations enabling its practice. Rather than framing field neuroethology in opposition to laboratory studies, we emphasize its potential to broaden the questions we can ask about neurobehavioral relationships-particularly those related to ecological validity and real-world fitness outcomes. Field neuroethology is not a replacement for traditional approaches, but rather an expansion of the experimental toolkit for investigating neurobehavioral functions expressed only in dynamic socioecological contexts.
Seasonal migration requires precise coordination between environmental sensing, neural processing, and physiological state. The monarch butterfly (Danaus plexippus) is a powerful model for understanding how neural systems generate adaptive responses across time and space. Migratory monarchs exhibit distinct seasonal reproductive traits and orientations, all regulated by environmental cues such as photoperiod, temperature, sunlight, and the Earth's magnetic field. Recent work has identified key neurobiological mechanisms underlying these processes, including circadian clock-dependent photoperiodic signaling, seasonal remodeling of the blood-brain barrier, and navigational compass systems guiding orientation. Here, we synthesize current knowledge and propose how anthropogenic stressors-artificial light at night, climate warming, and electromagnetic noise-could potentially disrupt the already threatened monarch migration by targeting key neurobiological interfaces that link environmental cues to migratory physiology and behavior. Understanding how these interfaces integrate environmental information will be critical for predicting the vulnerability of migratory monarchs, and possibly other insects, to rapid environmental change.
Perception depends on the brain's ability to transform high-dimensional sensory inputs into low-dimensional internal models that support adaptive behavior. Evidence supports two frameworks for sensory perception-representational processing, in which stimulus features are progressively integrated into complex perceptual objects across a cortical hierarchy, and predictive processing, in which internally generated predictions are continuously reconciled with incoming sensory signals. Yet how these frameworks are mechanistically implemented in neural circuits, and whether they can be unified, remains an open question. Here, we review recent studies in mouse primary sensory and higher-order association cortex demonstrating that cell-type-specific transcriptional programs may provide a critical mechanistic link between these frameworks and circuit functions. In primary sensory cortices, neurons that function as stable feature detectors or respond to sensory prediction errors correspond to distinct molecularly defined cell types. In higher-order association cortices, distinct inhibitory cell-type compositions and plasticity-related gene expression support both associative learning for representational processing and error learning for predictive processing. We discuss how cell-type-specific transcriptional programs may endow cell types and circuits with the capacity to support both representational and predictive processing modes in a behavioral state-dependent manner. This could potentially enable active sensation during behavioral engagement as well as memory consolidation and model updating during behavioral quiescence. Together, these studies suggest that examining how gene expression programs equip specific cell types with relevant computational properties is a promising approach that can integrate these frameworks and provide a new understanding of how sensory perception is implemented in the brain.
Parental care enhances offspring survival but requires profound alteration of parental physiological and behavioural states. Among vertebrates, teleost fishes exhibit great diversity in parental strategies, providing opportunities to investigate how the brains of different species integrate internal and external cues to produce adaptive care behaviours. Key regulators of parenting include prolactin, vasopressin, oxytocin and gonadal steroids, showing that endocrine signals coordinate parental motivation, protection, and feeding behaviour. At the neural level, recent studies highlight the hypothalamus as central for the integration of reproductive and energetic states. Key regulators of the balance between offspring-directed behaviour and self-maintenance include neurons that express galanin, mirroring conserved motifs described in mammals. As emerging tools advance our understanding of the mechanisms that underlie fish parental care, we will gain deeper insights into the evolution of neural circuits for social behaviour.
Sensory systems form the interface between organisms and their environment, enabling detection of external stimuli and guiding behavior. Across animals, sensory systems have repeatedly diversified, giving rise to specialized organs tuned to particular ecological roles. Sea robins provide a striking example of this process. These benthic fishes possess leg-like appendages derived from modified anterior pectoral fin rays that are used for walking and probing the seafloor. The legs function as multimodal sensory organs integrating chemosensory, mechanosensory, and proprioceptive inputs to detect and excavate buried prey. Correspondingly, the neural circuits associated with these appendages are dramatically expanded. Despite their unusual morphology and behavior, sea robins have received relatively little modern study, with much of the anatomical literature dating back more than a century. Recent molecular, cellular, and neurobiological studies are beginning to illuminate how such sensory appendages emerge and integrate into existing circuits.
Alzheimer's disease (AD) is increasingly conceptualized as a system-level disorder shaped by bidirectional communication between the gut and the brain. The gut-brain axis (GBA) integrates neural, immune, and metabolic signaling pathways that influence central neuroinflammation and proteopathy. Recent mechanistic studies demonstrate that gut dysbiosis alters microbial metabolite profiles, promotes microglial immunometabolic reprogramming, and facilitates amyloid and tau pathology. The vagus nerve functions as a bidirectional conduit enabling neural transmission of inflammatory signals and tau propagation directly. Emerging evidence implicates microbiota-derived extracellular vesicles as mediators of peripheral-to-central immune modulation. Human gut-brain organoid platforms now allow causal interrogation of these interactions in physiologically relevant systems. Together, these advances reframe AD as a disorder of dysregulated neural-immune communication and identify the GBA as a tractable therapeutic target.
