Super-Resolution Imaging of Fast Morphological Dynamics of Neurons in Behaving Animals
Super-Resolution Imaging Reveals New Advances in Neural Morphodynamics in Mouse Brains: Dynamic Observations in Head-Fixed Awake Animals
Background
In the field of neuroscience, understanding the morphological changes and functional dynamics of neurons is key to deciphering how the brain processes information and sustains network plasticity. Despite their critical roles in learning and behavioral adaptation, structures such as dendritic spines, axonal boutons, and synapses remain challenging to study dynamically in living organisms. Limited by the resolution and acquisition speed of traditional microscopy, many studies of neuron microstructures rely on fixed tissues or cultured cells, restricting insights into how plasticity correlates with natural behaviors and physiological states.
Recent breakthroughs in super-resolution microscopy (SRM) have surpassed the diffraction limit of traditional optical imaging, providing a pathway to bridge the gap between studying ultrafine neural structures and observing dynamic behavior in vivo. However, significant hurdles remain for SRM applications in awake animals due to motion artifacts and imaging speed constraints. Some studies have attempted to suppress motion artifacts by anesthetizing animals and using stimulated emission depletion (STED) microscopy to observe mouse brains. Yet, since anesthesia alters the normal physiological functions of neural systems, such observations cannot fully reflect neural dynamics in natural behavioral states.
Against this backdrop, Yujie Zhang and colleagues developed a novel technique named multiplexed, line-scanning structured illumination microscopy (MLS-SIM) to address these challenges. This breakthrough technology opens new avenues for super-resolution applications in head-fixed awake animals.
Paper Source
The research paper, titled “Super-resolution imaging of fast morphological dynamics of neurons in behaving animals,” was authored by Yujie Zhang, Lu Bai, Kai Wang, and others from institutions such as the Center for Excellence in Brain Science and Intelligence Technology at the Chinese Academy of Sciences and the School of Life Science and Technology at ShanghaiTech University. The paper was published in Nature Methods (doi: https://doi.org/10.1038/s41592-024-02535-9) in the January 2025 issue of Nature Methods, Volume 22.
Research Workflow and Methods
Research Subjects and Experimental Design
The primary goal of the research team was to develop and validate a method capable of long-term longitudinal super-resolution imaging in head-fixed awake animals while investigating dendritic spine and axonal bouton dynamics. The experiments were divided into three parts: technology development, resolution characterization, and multidimensional applications.
Development of MLS-SIM Technology
MLS-SIM integrates the advantages of traditional structured illumination microscopy (SIM) with innovative enhancements. The technique utilizes two distinct line-scanning excitation modes, each tailored to enhance resolution in specific directions using two-dimensional structured light patterns, while ensuring background rejection and rapid sampling. To complement these scanning modes, the authors developed a novel reconstruction algorithm that effectively deconvolves the raw low-resolution data into super-resolved images.Resolution and Motion Tolerance Characterization
Purkinje cells in model organisms (zebrafish and mice) were labeled with membrane-targeted enhanced green fluorescent protein (EGFP) to characterize the lateral (~150 nm) and axial (~450 nm) resolutions of MLS-SIM. By testing under various levels of motion (up to 100 μm/s) and signal-to-noise ratio (SNR) conditions, the method demonstrated strong resistance to motion artifacts.Applications in Head-Fixed Awake Mice
To verify its translational value, the authors employed MLS-SIM to study morphological dynamics of dendritic spines and axonal boutons during sleep-wake cycles in mouse brains. Additionally, through dual-color imaging, they observed PSD-95 substructures (postsynaptic density protein 95, a marker of synapse maturation) and their correlations with neuronal morphologies.
Image Reconstruction and Algorithmic Innovations
MLS-SIM image reconstruction is based on a multilayer joint deconvolution algorithm specifically optimized for varying point spread functions (PSFs) under different scanning modes. The research team segmented each two-dimensional image into multiple groups corresponding to distinct PSFs, reconstructed imaging layers accordingly, and ultimately fused them into a comprehensive high-resolution image through 3D stitching. Compared to existing methods, MLS-SIM dramatically reduces the number of sample frames required for super-resolution reconstruction, enhancing tolerance for motion artifacts.
Research Results and Data Analysis
Resolution and Motion Tolerance
- In static sample tests, MLS-SIM achieved lateral resolutions between 120–150 nm, comparable to existing SIM techniques.
- Under dynamic motion testing, the system demonstrated exceptional tolerance: resolutions of approximately 100 nm were maintained even with motions at speeds of 50 μm/s; at 100 μm/s, lateral resolutions perpendicular to the motion direction remained comparable to static conditions.
Dynamic Observations of Neural Morphology
- Dendritic Spines: Time-lapse imaging, lasting up to 20 minutes, revealed that 62% of dendritic spines exhibited dynamic spinule activity. These spinules were small, with exploration ranges typically extending ~1 μm, and most exhibited transient behavior positively correlated with their lifetimes.
- Axonal Boutons: MLS-SIM imaging of excitatory axonal boutons showed that 47% of the population exhibited small protrusions resembling spinules. These structures demonstrated rapid dynamics, unfolding on timescales of seconds.
Dual-Color Imaging of PSD-95
- PSD-95 proteins displayed morphological diversity, ranging from simple puncta to complex multi-domains, and were primarily located at dendritic trunk bases or within dendritic spine heads. Over 30 minutes of imaging, over 75% of PSD-95 clusters were spatially correlated with small protrusions, suggesting a potential link between spine emergence and PSD-95 recruitment.
Sleep-Wake Cycle Dynamics
- By monitoring EEG and EMG signals, the researchers captured morphological changes in dendrites and axons during rapid eye movement (REM) and non-REM (NREM) sleep states. While no significant area changes were observed in dendritic spines during sleep, spinule numbers fluctuated during specific periods, suggesting a potential modulatory role of sleep in synaptic plasticity.
Implications and Insights
Scientific Significance:
MLS-SIM allows scientists to achieve high-resolution imaging of neuronal morphology in awake animals, enabling deeper analyses of how plasticity changes under natural behavioral conditions.Technological Value:
MLS-SIM addresses the limitations of traditional super-resolution imaging in awake animals, overcoming motion artifacts while balancing imaging resolution and stability.Novel Discoveries:
The study demonstrated that small spinules exhibit rapid formation and retraction dynamics, linked to PSD-95 substructures. These dynamics may be critical for synapse maturation, disassembly, and neural network plasticity.Future Applications:
Beyond neuroscience, MLS-SIM can be widely applied to study dynamic biological systems, such as beating hearts and breathing lungs, offering a versatile new tool for medical diagnostics and tissue engineering research.
Conclusion
This study not only achieved successful super-resolution imaging in head-fixed awake mice but also provided profound insights into the correlation between neural structures and molecular dynamics. As an innovative tool, MLS-SIM lays a theoretical foundation for future research, particularly in understanding how neural networks regulate behaviors and memory mechanisms.