Neurons communicate by generating brief electrical impulses known as action potentials, or spikes. These signals travel along the cells and trigger communication with neighboring neurons, allowing networks of interconnected cells to process information. Conventional calcium imaging has enabled scientists to observe the activity of large numbers of neurons, but it does so indirectly. When a neuron fires, calcium ions flow into the cell, producing a chemical signal that can be detected through fluorescent indicators. Because calcium concentrations rise and fall relatively slowly, however, calcium imaging usually records activity over timescales of seconds or longer and may miss the individual spikes that carry information through neural circuits.

Voltage imaging offers a more direct alternative. Researchers can introduce genetically encoded voltage indicators into neurons, causing the cells to produce fluorescent proteins whose brightness changes when the electrical potential across the cell membrane changes. When a neuron fires, the indicator responds to the rapid shift in voltage, allowing the electrical event to be observed optically. In principle, this makes it possible to follow the timing and sequence of individual neural impulses. In practice, voltage signals are extremely brief and often faint, making it difficult to image them across a large three-dimensional brain at the speed and resolution required to distinguish individual cells.

To overcome this limitation, the MIT team modified a light-sheet microscope, an instrument designed to image large biological samples rapidly while reducing light exposure. A light sheet illuminates only a thin plane of tissue, and the microscope records the fluorescence emitted from that plane before moving through successive layers. Combining those images produces a three-dimensional representation of the sample. The researchers accelerated both parts of the process: they increased the acquisition speed of the camera and used a technique called remote refocusing to shift the imaging plane rapidly without mechanically moving the specimen or the main optical components.

The resulting instrument was able to scan the entire brain of a larval zebrafish 200 times per second, completing one full volume every five milliseconds. This rate is fast enough to capture many of the electrical events that conventional whole-brain imaging would blur or miss. The zebrafish is particularly useful for this kind of experiment because its larval brain is small and relatively transparent, allowing researchers to image neural activity throughout the organism without the need to physically remove tissue. Its nervous system also contains many of the major functional structures found in vertebrates, making it a valuable model for studying how brain-wide circuits operate.

For their demonstration, the researchers engineered larval zebrafish to express a genetically encoded voltage indicator called Positron2-Kv. The indicator did not produce usable signals in every neuron, but approximately one-quarter of the neurons showed fluorescence changes strong enough for analysis. Even this partial coverage allowed the team to observe activity across many brain regions at once. In fish that were resting, the microscope detected individual voltage spikes as well as rapid bursts of activity. These recordings provided a direct view of the timing of electrical signals and offered information that would be difficult to obtain from slower calcium measurements.

The system also revealed how activity spread through the brain after the fish received ultraviolet light. Soon after the stimulus, neurons in the optic tectum became active. This brain region receives visual information from the retina and performs early stages of visual processing. The activity then propagated across the tectum, moving from one side of the structure to the other. Because the microscope recorded the activity throughout the brain rather than in a single visual-processing area, the researchers could also observe stimulus-independent sequences in groups of neurons located in the cerebellum and hindbrain. These patterns suggest that spontaneous brain activity is organized across distributed networks, even when the animal is not responding to an obvious external signal.

The ability to observe voltage signals across a complete brain could change the way researchers formulate questions about neural computation. Brain functions rarely depend on isolated groups of neurons; instead, they emerge from interactions among circuits that may be separated by considerable anatomical distances. A neuron in a sensory region may influence cells involved in movement, attention, or internal state within milliseconds. If experiments examine only one region at a time, important participants in these network-wide processes can be overlooked. Whole-brain voltage imaging could help scientists identify coordinated activity patterns first and then investigate how specific neurons and connections contribute to them.

The new microscope is not yet a complete solution to the challenges of brain-wide recording. The researchers aim to increase the proportion of neurons that produce strong voltage-indicator signals while improving spatial resolution, imaging speed, and data analysis. Fluorescence must be collected quickly enough to distinguish a faint electrical response from background noise, and the enormous data volumes generated by high-speed three-dimensional imaging require sophisticated computational methods. The team is also exploring whether the approach can be adapted for other experimental organisms, including mice, whose larger and more complex brains present additional optical and technical obstacles.

Despite those limitations, the demonstration represents a significant step toward observing the brain as an integrated electrical network. By measuring voltage directly from neurons distributed throughout an entire organism, the method could help researchers connect rapid neural events with sensory responses, spontaneous activity, behavior, and internal mental states. Future experiments may use it to investigate how brain-wide circuits support learning, decision-making, movement, or states such as daydreaming. The broader goal is to understand not only what individual neurons do, but how thousands of cells coordinate their electrical activity to produce the unified functions of a living brain.

Subject of Research: Brain-wide voltage imaging of neuronal activity in larval zebrafish

Article Title: Voltage imaging of neurons distributed across entire brains of larval zebrafish

News Publication Date: 14-Aug-2026

Web References: https://doi.org/10.1038/s41592-026-03179-7

References: Nature Methods, DOI: 10.1038/s41592-026-03179-7

Keywords

Voltage imaging, genetically encoded voltage indicators, neurons, zebrafish, whole-brain imaging, light-sheet microscopy, neuroscience, neural networks, brain activity, Positron2-Kv

REM sleep, characterized by vivid dreaming and heightened brain activity, has long fascinated neuroscientists for its unique signatures and critical role in memory consolidation, emotional regulation, and brain plasticity. Despite decades of intensive research, the precise neural underpinnings gating the transition into REM sleep remained largely speculative due to the complexity of brainstem circuitry and limitations in capturing its dynamic activity. This new study leverages cutting-edge population recording techniques combined with innovative computational analyses to strip down this complexity into core dynamic modes that define REM gating.

The investigators used multi-site electrophysiological recordings from the brainstem of freely behaving rodents, capturing hundreds of neurons simultaneously as animals cycled through awake, non-REM, and REM states. Employing dimensionality reduction algorithms, they distilled these high-dimensional signals into a succinct set of population dynamics—essentially revealing the ‘neural language’ that brainstem neurons use to collectively initiate REM sleep. This approach departs from traditional single-cell analyses, recognizing that sleep state transitions arise from coherent network-level dynamics rather than isolated neuronal activity.

At the core of the findings is a low-dimensional manifold—a mathematical representation of neural activity trajectories—that succinctly encodes the progression into REM sleep. This manifold was observed to act as a ‘gateway’, whereby brainstem population activity follows distinct trajectories that gate the timely and selective entry into REM epochs. Crucially, the researchers demonstrated that perturbing this neural trajectory via optogenetic or pharmacological manipulations disrupted the normal onset and maintenance of REM sleep, underscoring the causal role of these dynamics.

The study’s detailed mapping of brainstem population states revealed an intricate interplay between cholinergic, glutamatergic, and GABAergic neurons, each contributing to the shape and flow of the activity manifold underlying REM gating. By combining cell-type-specific recordings with network-level models, the research sheds light on how these diverse neurotransmitter systems coordinate to sculpt brainstem activity sequences critical for sleep regulation. This integrated view advances our mechanistic insight beyond traditional neurotransmitter-centric frameworks.

Notably, the research addresses a longstanding question concerning the variability and robustness of REM sleep transitions. The low-dimensional manifold framework explains how neural population dynamics maintain consistency despite inherent neural noise and variability, ensuring reliable REM gating. This suggests that the brainstem employs a stable attractor landscape—an energy-efficient strategy—to buffer against perturbations and maintain sleep architecture integrity.

Furthermore, the temporal precision of the identified dynamics aligns elegantly with behavioral markers of REM sleep, such as rapid eye movements and cortical desynchronization. This temporal coordination across brain regions implies that brainstem population dynamics serve as a master regulator, orchestrating downstream cortical and subcortical circuits essential for the full expression of REM sleep phenomenology.

The translational implications of these results are profound. REM sleep disturbances feature prominently in a range of neuropsychiatric conditions, including depression, schizophrenia, and neurodegenerative diseases. By unraveling the fundamental gating mechanisms in the brainstem, this study lays the groundwork for targeted therapeutic strategies aimed at restoring normal REM sleep transitions. Precision modulation of brainstem network dynamics could become a viable avenue to ameliorate symptoms linked to REM sleep dysregulation.

Another remarkable aspect is the study’s demonstration of how machine learning and computational neuroscience can synergize to uncover hidden organizational principles in complex biological networks. The use of dimensionality reduction and dynamical systems theory enabled the researchers to transcend the limitations imposed by the sheer scale and heterogeneity of brainstem neurons. This methodological advance could inspire a new wave of research dissecting other elusive brain states and transitions.

Additionally, the study highlights the evolutionary conservation of REM sleep regulating circuits. Comparative analyses suggest similar low-dimensional dynamics govern REM gating across mammalian species, emphasizing the fundamental and conserved nature of these neural strategies. This evolutionary perspective enhances the relevance of animal models in deciphering human sleep disorders and paves the way for cross-species translational research.

Beyond sleep, the principles elucidated here may extend to other brain state transitions such as anesthesia induction, arousal regulation, and even pathological states like epilepsy. Understanding how low-dimensional neural manifold trajectories control state gating can inform broader neurophysiological models and interventions across diverse conditions where brain state regulation is disrupted.

In sum, this pioneering work provides a comprehensive and mechanistic framework that connects neural population dynamics in the brainstem to the gating of REM sleep. By revealing a low-dimensional landscape that governs REM entry, the research creates a conceptual bridge between microcircuit activity and whole-brain behavioral states. This paradigm shift radically advances sleep neuroscience and opens rich avenues for innovative clinical applications.

As neuroscientists continue to explore the brain’s complex state transitions, the integration of population-level recordings with computational tools—exemplified by this study—will be indispensable. Unlocking the brain’s low-dimensional signatures offers unprecedented capacity to decode its inner workings and design precise interventions for disorders tethered to disrupted brain states.

With REM sleep’s role expanding from mere ‘dream time’ to a critical function in brain health and disease, understanding its gated entry through low-dimensional dynamics marks a pivotal moment in neuroscience. This knowledge not only enriches our grasp of sleep biology but also catalyzes novel translational possibilities, embodying the future of integrative brain research.

Subject of Research: Neural population dynamics in the brainstem regulating REM sleep gating

Article Title: Low-dimensional population dynamics in the brainstem gate REM sleep

Article References:
Lozano, D.E., Hong, J., Jin, X. et al. Low-dimensional population dynamics in the brainstem gate REM sleep. Nat Neurosci (2026). https://doi.org/10.1038/s41593-026-02314-z

Image Credits: AI Generated

DOI: https://doi.org/10.1038/s41593-026-02314-z

The article by Major et al. serves as both a rebuttal and a clarification aimed at refining interpretations of prior experimental findings. Their reply underscores the intrinsic complexity of neural networks, advocating for a multidimensional analysis approach that integrates electrophysiological data, advanced imaging techniques, and computational modeling. This comprehensive methodology seeks to transcend the reductionist paradigms that have historically limited the field, pushing toward a more holistic understanding of how neural ensembles coordinate behaviorally relevant patterns.

Technical rigor is a hallmark of their reply, as the authors delve into nuanced discussions of synaptic integration, temporal coding strategies, and spatial heterogeneity within microcircuits. By dissecting the interplay between excitatory and inhibitory inputs at the cellular level, they emphasize the necessity of considering variegated synaptic weights and dendritic processing capabilities. This consideration, they argue, is crucial for accurately modeling neural computations and predicting network responses under physiological and pathological conditions.

Major and colleagues also confront the challenges posed by recent high-throughput approaches, notably single-cell transcriptomics and optogenetic manipulation. They caution against simplistic interpretations of data obtained from such technologies, urging for careful calibration of experimental parameters and rigorous validation of findings through complementary methods. Their perspective highlights the risks of overgeneralization, particularly when extrapolating molecular signatures to functional phenotypes without accounting for dynamic state-dependent variables.

The response further illuminates the role of interneuron diversity in shaping circuit output, a theme that has garnered significant attention in recent years. Through a synthesis of anatomical, electrophysiological, and genetic evidence, the authors present a compelling case for subclass-specific contributions to network oscillations and synchronization phenomena. This insight not only enriches the conceptual landscape but also offers potential targets for intervention in neurological disorders characterized by dysregulated inhibitory control.

Integral to their argument is an emphasis on the temporal dimension of neural activity. By examining how transient synaptic events cascade into sustained network-level effects, Major et al. challenge conventional static models that fail to capture the fluidity of information processing in the brain. Their analysis leverages cutting-edge computational frameworks to simulate dynamic interactions over multiple time scales, providing new avenues for understanding phenomena such as plasticity, memory consolidation, and attentional modulation.

The importance of spatial context within neural tissue also receives substantial attention. The authors discuss the limitations of oversimplified localization assumptions and advocate for employing sophisticated imaging modalities capable of resolving fine-grained structural-functional relationships. In particular, they underscore the promise of integrative microscopy techniques that allow simultaneous assessment of morphological, molecular, and functional attributes within intact circuits.

Major et al.’s reply ventures beyond methodological critique to propose an ambitious conceptual synthesis. They argue for a paradigm shift towards viewing neural circuits as adaptive, self-organizing systems whose emergent properties cannot be fully understood through linear cause-and-effect models. This perspective aligns with contemporary theories in systems neuroscience and complexity science, which emphasize feedback loops, nonlinearity, and probabilistic computation as foundational elements of brain function.

In tackling the issue of reproducibility and data transparency, the authors commend recent efforts to standardize protocols and share datasets openly. Nonetheless, they highlight persistent obstacles related to biological variability, experimental design heterogeneity, and analysis pipeline discrepancies. Their call to action advocates for community-wide initiatives to foster collaborative frameworks that harmonize methodologies without stifling innovation.

Beyond the technical and theoretical discourse, the article reflects a wider philosophical contemplation regarding the trajectory of neuroscience research. Major and colleagues acknowledge the tension between technological advancements and conceptual clarity, cautioning researchers to maintain a critical eye towards data interpretation. They stress the value of iterative hypothesis testing and integrative modeling as means to avoid premature conclusions driven by methodological trends rather than substantive insights.

The reply also provides a nuanced discussion of translational implications. By elucidating fundamental mechanisms of neural circuit operation, the authors open new possibilities for developing targeted therapeutic interventions. They particularly emphasize the need to bridge basic neuroscience with clinical applications, highlighting how refined circuit-level understanding can inform pharmacological strategies as well as neuromodulation techniques for treating neuropsychiatric conditions.

In synthesizing these themes, the article exemplifies the dynamic, self-correcting nature of scientific progress. Major et al.’s contribution embodies the spirit of rigorous peer discourse, demonstrating how constructive criticism and thoughtful engagement propel the field forward. Their response is not merely reactive but proactive, setting a bright agenda for future investigations that promises to unravel the brain’s deepest mysteries.

In summary, the reply by A.J. Major and collaborators represents a pivotal moment in contemporary neuroscience dialogue. It not only addresses critical methodological points but also advances a visionary framework that champions integration, dynamic complexity, and translational relevance. This multifaceted approach is poised to inspire new generations of neuroscientists dedicated to decoding the intricate language of neural circuits and ultimately translating these insights into tangible benefits for human health.

Subject of Research: Neural circuit dynamics, synaptic integration, and network function within the brain

Article Title: A. J. Major et al. reply

Article References: Major, A.J., Abdaltawab, A., Phillips, J.M. et al. A. J. Major et al. reply. Nat Neurosci (2025). https://doi.org/10.1038/s41593-025-02168-x

Image Credits: AI Generated

DOI: https://doi.org/10.1038/s41593-025-02168-x