Astrocytic Ca2+ Prevents Synaptic Depotentiation by Limiting Repetitive Activity in Dendrites During Motor Learning

Background and Motivation

In the field of neuroscience, learning and memory processes rely on the intricate regulation of cellular activities in the brain. Previous research has predominantly focused on synaptic plasticity between neurons—such as long-term potentiation (LTP) and long-term depression (LTD)—as the material basis for the remodeling of neural circuits, thereby promoting the advancement of neuroscience. However, in recent years, a burgeoning area of research—the impact of astrocytes on brain function—has drawn increasing attention. Astrocytes are not merely “supporting actors” to neurons; they actively influence neuronal activity and synaptic transmission by modulating neuronal metabolism, buffering extracellular ion environments, uptaking neurotransmitters, and secreting regulatory molecules. Nevertheless, the specific mechanisms by which these functions operate at the behavioral level, such as in learning and memory, remain largely unresolved.

In particular, it is still largely unknown whether, and how, astrocytic calcium (Ca2+) activity directly participates and determines synaptic plasticity associated with learning in vivo. Although substantial experiments on brain slices or cultured cells have suggested that activation of astrocytic Ca2+ can modulate synaptic function by releasing signaling molecules such as ATP, glutamate, and D-serine, there is ongoing debate regarding the true regulatory network in the intact brain. Limitations of experimental technologies and research approaches have also led to contradictory results among different studies.

Furthermore, genetic or pharmacological interventions targeting astrocyte functions—such as blocking lactate shuttling, glutamate transport, IP3R2 (inositol 1,4,5-trisphosphate receptor 2)-mediated Ca2+ release, or G-protein coupled receptor (GPCR) activation—show highly variable impacts on learning and memory. Some result in behavioral deficits, while others show no detectable effect, reflecting the complex regulatory mechanisms at play.

This study accurately focuses on a core scientific question: In the true process of learning in vivo, particularly during motor learning in live mice, does an astrocytic Ca2+ increase determine the dynamics of synaptic strengthening/weakening and thereby provide a safeguard for the formation of motor-related memories? The authors aim to elucidate the regulatory mechanisms of astrocyte activity, its direct impacts on synaptic plasticity during learning, and the involved molecular signaling pathways—filling a major gap in the study of intercellular interactions in neuroscience.

Source of the Paper and Author Information

This original research paper, titled “astrocytic ca2+ prevents synaptic depotentiation by limiting repetitive activity in dendrites during motor learning,” was published in Nature Neuroscience (volume 28, 2296–2309) in November 2025, with DOI: https://doi.org/10.1038/s41593-025-02072-4. The main authors include Baoling Lai, Deliang Yuan, Zhiwei Xu, Feilong Zhang, Ming Li, Alejandro Martín-Ávila, Xufeng Chen, Kai Chen, Kunfu Ouyang, Guang Yang, Moses V Chao, and Wen-Biao Gan. The author team is affiliated with top international research institutions, including New York University Grossman School of Medicine, Shenzhen Bay Laboratory, Peking University Shenzhen Graduate School, Beijing Normal University, Columbia University Irving Medical Center, and Lingang Laboratory, reflecting the world-class academic background and interdisciplinary characteristics of this study.

Research Workflow and Detailed Experimental Design

1. Detection of Astrocytic Calcium Dynamics Induced by Motor Learning

At the beginning of the study, the authors employed in vivo two-photon microscopy to achieve high spatial and temporal resolution monitoring of astrocytic Ca2+ dynamics in the motor cortex of live mice. Using adeno-associated virus (AAV5) and the astrocyte-specific GFAP-C1D promoter to drive the expression of the genetically encoded calcium indicator GCaMP6f, and in combination with a thin-skull window strategy, they successfully recorded Ca2+ transients in cortical layer 1 astrocyte somas and processes during mouse treadmill training.

The experiment included at least five groups of mice, with each group’s Ca2+ signal dynamics, peak latency, half-width, and other parameters recorded and quantified over 15 training sessions. Through a cross-breeding strategy using genetically engineered mice (GLAST-creER;PC::G5-tdt) expressing another Ca2+ indicator gcamp5/tdTomato, the authors ruled out the possibility that the observed Ca2+ rise was induced by viral infection—demonstrating that motor training alone was sufficient to evoke astrocytic Ca2+ dynamics, thus enhancing the credibility of their findings.

At the behavioral level, the experiment simultaneously recorded and quantified changes in mouse forelimb stride width, accurately evaluating the contribution of motor training to behavioral improvement and effectively linking cellular signals to enhancements in motor capability.

2. Analysis of the Signaling Mechanisms of Astrocytic Ca2+ Elevation

To further analyze the mechanisms underlying the rise in astrocytic Ca2+ induced by motor training, the research team employed multiple pharmacological and genetic tools:

• Blockade of Neuronal Activity: Local application of the sodium channel blocker tetrodotoxin (TTX) significantly reduced astrocyte Ca2+ elevation in the cortex, indicating that neuronal input activity is essential for Ca2+ signal generation.
• GABA_A Receptor Activation: Administration of muscimol also notably suppressed astrocytic Ca2+ elevation, further confirming that local cortical neuronal activity is closely related to astrocytic Ca2+ dynamics.
• Adrenergic Signaling Pathway: Through GCaMP6s expression in the locus coeruleus (LC) and its descending axons, dynamic imaging revealed that motor training significantly increases norepinephrine content, which in turn activates the GPCR pathway via α1-adrenergic receptors on astrocytes to trigger Ca2+ signaling.
• Pharmacological and Genetic Interventions: Local application of the α1-receptor antagonist prazosin or chemical depletion of norepinephrine (with DSP4) both effectively inhibited the training-induced astrocytic Ca2+ elevation, whereas administration of the receptor agonist phenylephrine alone could induce Ca2+ transients.
• IP3R2-Mediated ER Ca2+ Release: Using genetic tools such as IP3R2 knockout (KO) mice, it was found that in the absence of IP3R2, training-induced Ca2+ elevation in astrocytes was significantly impaired, further confirming the importance of the GPCR–IP3R2 pathway.

In addition, the study employed an astrocyte-specific Gq-DREADD (Designer Receptor Exclusively Activated by Designer Drug) system and CNO (clozapine-N-oxide) to allow precise control of astrocytic Ca2+ signals. In vivo ER calcium measurement with G-CEPIA1er further confirmed that sustained GPCR activation could lead to exhaustion of ER Ca2+, explaining the molecular basis for the disappearance of astrocytic Ca2+ responsiveness over longer periods.

3. Probing the Relationship Between Astrocytic Ca2+ and Synaptic Plasticity

As one of the study’s core components, the authors achieved high-resolution in vivo imaging of the dendrites and dendritic spines of layer 5 pyramidal neurons in the mouse brain, using YFP/tdTomato labeling to track changes in their structure (size, formation, and elimination) and Ca2+ dynamics. In the control group under motor training, dendritic spine volumes increased on average (synaptic potentiation), whereas following prazosin intervention, IP3R2 knockout, or DREADD activation via CNO (all of which suppressed astrocytic Ca2+), dendritic spine volumes significantly decreased (synaptic weakening) and the formation rate of new spines diminished while elimination rates increased, with the behavioral improvement from training also diminished.

Further analysis revealed that the impact of astrocytic Ca2+ was prominent during training, with little effect on dendritic spines observed in the absence of training. Moreover, larger initial spine size predicted a greater likelihood of shrinkage after training, suggesting that highly active spines are particularly sensitive to astrocytic Ca2+ regulation.

4. Astrocytic Calcium Signaling Limits Repetitive Dendritic Ca2+ Activation and Prevents Synaptic Weakening

By coupling calcium imaging and dendritic spine structure tracking, the authors found that inhibition of astrocytic Ca2+ (by either pharmacological or genetic means) led to a subset of dendritic branches exhibiting high-frequency repetitive Ca2+ spikes during motor training, whereas the control group had only a very small fraction of branches with more than 8 Ca2+ spikes. Dendritic spines on these branches with high-frequency repetitive Ca2+ spikes generally shrank, indicating weakened synaptic function. This phenomenon suggests that astrocytic Ca2+ helps suppress excessive dendritic activity, thereby protecting synapses from training-induced weakening.

The experiment also investigated the relationship between the timing of dendritic Ca2+ spikes and dendritic spine Ca2+ activity (spike-timing-dependent mechanism), finding that if a spine Ca2+ transient occurred before a dendritic Ca2+ spike, it was more likely to shrink in size. Further experiments using the CaMKII (calcium/calmodulin-dependent protein kinase II) inhibitor KN62 demonstrated that activation of this kinase is a key molecular mechanism underlying dendritic spine weakening.

5. Astrocyte-Secreted ATP/Adenosine Signaling Regulates Dendritic and Synaptic Activity

To further explore the molecular pathways of the aforementioned high-frequency dendritic Ca2+ spikes, the authors investigated the involvement of signaling molecules. They found that motor training rapidly elevated extracellular adenosine levels in astrocytes (as detected with the genetically encoded GRAB-ADO1.0MED adenosine sensor). Local application of the adenosine A1 receptor agonist CPA significantly reduced dendritic and spine Ca2+ activity, peak amplitude, and the number of pre-activated spines, while the antagonist DPCPX increased these indicators, confirming that adenosine signaling dynamically suppresses extensive synchronous activity of dendritic synapses and preserves synaptic function.

Crucially, exogenous adenosine was able to reverse the synaptic weakening caused by IP3R2 KO or DREADD-CNO, demonstrating that astrocytic ATP release and adenosine signaling constitute a novel in vivo synaptic homeostasis protection mechanism.

Main Findings, Significance, and Scientific Value

In summary, through multi-level imaging, genetic and pharmacological intervention, and behavioral analysis, the authors systematically and, for the first time, demonstrated that the rapid increase of astrocytic Ca2+ during in vivo motor learning is essential for the maintenance of synaptic function—by suppressing repetitive high-frequency dendritic Ca2+ activation, they prevent reductions in dendritic spine size and number, thus protecting synaptic homeostasis. This intercellular interaction depends on neuronal norepinephrine input, the GPCR–IP3R2 signaling pathway, and downstream ATP/adenosine signaling.

The scientific value of this research includes:

1. First revealing, in the living environment, that astrocytic Ca2+ directly regulates the homeostasis of learning-related synapses, re-writing the previous neuron-centric understanding of synaptic plasticity and providing a new perspective on intercellular modulation.
2. Establishing a causal chain linking motor learning, astrocytic Ca2+ signaling, norepinephrine signaling, repetitive dendritic activation, and synaptic weakening.
3. Innovatively proposing ATP/adenosine signaling as a synaptic homeostasis protection factor, offering new molecular targets for intervention in brain diseases and cognitive disorders.
4. Validating the high repeatability and reliability of technologies such as in vivo imaging, genetic engineering, and chemogenetics (DREADD) under living conditions, driving advancements in neuroscience experimental methods.

Research Highlights and Application Prospects

• Innovative Experimental Design: In vivo two-photon multicolor imaging, genetically encoded Ca2+/adenosine sensors, and high spatial-temporal resolution dynamic detection in living animals.
• Multiple Intervention Approaches: Manipulation at the neuronal, astrocytic, and signaling pathway levels, ensuring thorough mechanistic analysis.
• Integration of Behavioral and Cellular Analysis: Precisely links cellular-level signal changes to animal motor performance, pushing neuroscience towards an integrated “molecule-to-behavior” paradigm.
• Clinical Translational Value: Opens new avenues for interventions targeting synaptic plasticity mechanisms in diseases such as cognitive disorders, motor disorders, and psychiatric conditions including Alzheimer’s and Parkinson’s disease.
• Academic Impact: Research team and institutions with top international backgrounds, published in Nature Neuroscience and rigorously peer-reviewed, representing significant academic influence.

Conclusion and Outlook

This study systematically establishes a complete multicellular interactive chain: “motor learning—neuronal activity—astrocytic Ca2+ response—ATP/adenosine signaling—dendritic synaptic homeostasis,” laying the cellular and molecular foundation for understanding the glia-neuron cooperative mechanisms in brain learning and memory. Its results not only deepen our biological understanding of neural circuit plasticity regulation but also provide solid scientific support for future mechanistic studies and the development of novel therapeutic strategies for related diseases.

In today’s neuroscience, where research into complex cellular interactions is deepening, this paper undoubtedly sets a new milestone for the field and opens a new chapter in exploring the role of astrocytes in cognitive function regulation.