This research identifies the protein UBQLN2 as a vital bridge between protein quality control and lipid metabolism in neurodegenerative diseases like ALS and FTD. By studying mutated neurons and brain organoids, scientists discovered that UBQLN2 normally regulates the breakdown of two enzymes, ILVBL and ALDH3A2, which are essential for processing fats during energy stress. When UBQLN2 is mutated or obstructed by TDP-43 protein aggregates, these enzymes accumulate, leading to lipid droplet depletion and excessive fatty acid oxidation that starves and kills neurons. Experimental treatments that reduced these enzyme levels or provided cholesterol supplements successfully restored metabolic balance and improved survival in animal models. Ultimately, the study establishes the UBQLN2–ILVBL/ALDH3A2 axis as a potential therapeutic target to prevent the metabolic collapse associated with dementia and motor neuron loss.

References:

• Liu Y, Huang Z, Hsu Y W, et al. UBQLN2 links proteotoxicity with lipid metabolism in neurodegeneration[J]. Nature Neuroscience, 2026: 1-14.

The paper details a technological breakthrough in restoring rapid communication for people with severe paralysis using an intracortical brain-computer interface. Researchers developed a typing neuroprosthesis that allows users to operate a bimanual QWERTY keyboard by simply attempting finger movements. This system achieved a record-breaking speed of 110 characters per minute, significantly surpassing previous hand-motor interfaces while maintaining high accuracy through a recurrent neural network and language models. The study included participants with ALS and spinal cord injuries, demonstrating that neural signals for bimanual typing are robust even when recorded from a single hemisphere. Because it utilizes a familiar keyboard layout, the interface is intuitive, easy to learn, and requires minimal daily calibration. Ultimately, this innovation offers a faster and more private alternative to existing assistive technologies like eye-gaze trackers or speech-to-text systems.

References:

• Jude J J, Levi-Aharoni H, Acosta A J, et al. Restoring rapid natural bimanual typing with a neuroprosthesis after paralysis[J]. Nature Neuroscience, 2026: 1-10.

This research profiles the epigenomic landscape of the adult human central nervous system using single-nucleus sequencing to map chromatin accessibility and histone modifications. By analyzing various brain regions and the spinal cord, the authors identified a persistent epigenetic memory of developmental programs within adult glial cells, specifically oligodendrocytes. While these signatures do not always correlate with active gene expression, they appear to prime cells for regenerative responses or, conversely, increase vulnerability to certain types of brain tumors. The study provides a comprehensive multimodal resource that details how different cell populations, such as neurons and astrocytes, maintain distinct regulatory environments. Ultimately, these findings reveal how chromatin architecture and histone marks influence both the functional plasticity and the disease susceptibility of the mature human nervous system.

References:

• Kabbe M, Agirre E, Carlström K E, et al. Single-nucleus epigenomic profiling of the adult human central nervous system unveils epigenetic memory of developmental programs[J]. Nature Neuroscience, 2026: 1-15.

This research from Nature Neuroscience utilizes computational whole-brain modeling to demonstrate that competitive interactions are essential for understanding how mammalian brain networks function. By analyzing data from humans, macaques, and mice, the authors found that models incorporating both cooperative and competitive ties more accurately replicate real-world brain activity than traditional models. These negative-valued interactions typically link regions with opposing biological traits, such as different gene expressions or cellular structures, and are characterized by long-range, diffuse connections. The study suggests that such competition promotes synergistic dynamics and hierarchical organization, which are critical for complex information processing. Ultimately, the work establishes a generative link between anatomical architecture and the sophisticated computational performance of the living brain.

References:

• Luppi A I, Sanz Perl Y, Vohryzek J, et al. Competitive interactions shape mammalian brain network dynamics and computation[J]. Nature Neuroscience, 2026: 1-19.

Unraveling the cellular and molecular characteristics of human prefrontal cortex (PFC) development is crucial for understanding human cognitive abilities and vulnerability to neurological and neuropsychiatric disorders. Here, in this study, we created a comparative repository for gene expression, chromatin accessibility and spatial transcriptomics of human and macaque postnatal PFC development at single-cell resolution. Integrative analyses outlined species-specific dynamic trajectories of different cell types, highlighting key windows and gene regulatory networks for processes such as synaptogenesis, synaptic pruning and gliogenesis. We identified regulatory correlates of the prolonged development of human PFC relative to macaques. Glial progenitors showed higher proliferation capability in humans compared to macaques, associated with distinct gene expression profiles. Furthermore, we uncovered cell types and lineages most susceptible t­ o n­ euro­ developmental a­ nd n­ eu­ ro­ ps­ yc­ hi­ atric disorders, focusing on transcription factors with human-specific expression features. In summary, our discoveries shed light on human-specific regulatory programs extending postnatal cortical maturation through coordinated neuronal and glial development, with implications for cognition and neurodevelopmental disorders.

References:

• Zhang J, Li M, Wang M, et al. Single-cell spatiotemporal transcriptomic and chromatin accessibility profiling in developing postnatal human and macaque prefrontal cortex[J]. Nature Neuroscience, 2026, 29(3): 746-758.

This research from Nature Neuroscience explores how the mammalian central nervous system regulates gene expression following injury. By utilizing single-nucleus multiomics, the authors identified thousands of injury-responsive enhancers (IRENs) that control cell-type-specific repair and stress programs in the mouse spinal cord. They developed deep learning models to decode the complex DNA sequences of these enhancers, revealing how generic stimulus signals combine with specific cell identity programs. The study demonstrates that these decoded sequences can be used to selectively target reactive astrocytes with therapeutic gene delivery vectors. Ultimately, this work provides a blueprint for engineering precision genomic tools to treat disease-associated cell states in the brain and spinal cord.

References:

• Zamboni M, Martínez-Martín A, Rydholm G, et al. The regulatory code of injury-responsive enhancers enables precision cell-state targeting in the CNS[J]. Nature Neuroscience, 2026, 29(2): 337-349.

This research explores how the Alzheimer’s drug Lecanemab utilizes microglia to eliminate harmful amyloid-beta (Aβ) plaques from the brain. By utilizing human microglia xenograft models and spatial transcriptomics, scientists discovered that the therapy’s effectiveness depends on its Fc fragment to trigger a specific genetic program in immune cells. This program boosts phagocytosis, improves lysosomal degradation, and increases the expression of SPP1/osteopontin, a protein that directly aids in clearing deposits. Unlike other treatments, Lecanemab appears to reprogram these cells for a protective response without causing significant synaptic loss or extreme inflammation. Ultimately, the study concludes that effective Aβ reduction requires active microglial engagement, providing a blueprint for refining future immunotherapies.

References:

• Albertini G, Zielonka M, Cuypers M L, et al. The Alzheimer’s therapeutic Lecanemab attenuates Aβ pathology by inducing an amyloid-clearing program in microglia[J]. Nature Neuroscience, 2026, 29(1): 100-110.

This research introduces CCF-ME, a high-resolution mouse brain atlas developed by analyzing the dendritic microenvironments of over 100,000 neurons. Unlike traditional maps based on cell density, this framework groups neurons by integrating the morphological features of their local neighborhoods to reveal a more detailed anatomical organization. This innovative approach nearly doubles the number of identifiable brain regions, providing a finer-grained view than the standard Allen Common Coordinate Framework. The study demonstrates that these local structural patterns are highly predictive of long-range axonal connections, particularly within the hippocampus. By aligning structural granularity with functional and projection specificity, the atlas offers a robust tool for understanding the relationship between neuronal shape and global brain connectivity. Undergoing rigorous validation, the microenvironment representation proves more effective than single-neuron analysis in identifying distinct subregions and spatial patterns.

References:

• Liu Y, Zhao S, Yun Z, et al. A mouse brain atlas based on dendritic microenvironments[J]. Nature Neuroscience, 2026, 29(1): 111-122.

This research investigates how oxidized phosphatidylcholines (OxPCs) act as primary drivers of chronic neurodegeneration in progressive multiple sclerosis (P-MS). By using a mouse model, the authors demonstrate that depositing these toxic byproducts into the central nervous system creates persistent lesions that mirror the pathological and transcriptomic signatures of human disease. The study highlights that aging significantly worsens these injuries by promoting microglial dysfunction and increasing the recruitment of inflammatory macrophages. Furthermore, the findings reveal a destructive feedback loop involving IL-1β signaling, which fuels the continuous accumulation of OxPCs and prevents tissue repair. Ultimately, the results suggest that neutralizing OxPCs or blocking IL-1β could offer a promising therapeutic strategy for treating progressive forms of MS.

References:

• Yu R, Lozinski B M, Seifert A, et al. Oxidized phosphatidylcholines deposition drives chronic neurodegeneration in a mouse model of progressive multiple sclerosis via IL-1β signaling[J]. Nature Neuroscience, 2026, 29(1): 67-80.

This research outlines a complex, brain-wide network called the allostatic–interoceptive system, which is responsible for regulating the body's metabolic resources and interpreting internal signals. By utilizing 7 Tesla fMRI technology, the study provides unprecedented detail regarding the connections between the cerebral cortex and small, deep structures within the brainstem and midbrain. The findings demonstrate that this system involves an integration of the default mode and salience networks, mirroring anatomical pathways previously identified in animal research. Notably, the study highlights that many of these regions act as "rich club" hubs, serving as a high-capacity backbone for global communication across the brain. Ultimately, the results suggest that allostasis and interoception are fundamental processes that shape a wide range of psychological experiences, including emotion, cognition, and perception.

References:

• Zhang J, Chen D, Deming P, et al. Cortical and subcortical mapping of the human allostatic–interoceptive system using 7 Tesla fMRI[J]. Nature Neuroscience, 2025: 1-12.