Abstract
The spatial interactions of synaptic vesicles in synapses were assessed after a detailed characterization of size, shape, and orientation of the synaptic vesicles. We hypothesized that shape and orientation of the synaptic vesicles are influenced by their movement toward the active zone causing deviations from spherical shape and systematic trends in their orientation. We studied three-dimensional representations of synapses obtained by manual annotation of focused ion beam scanning electron microscopy (FIB-SEM) images of male mouse brain. The configurations of synaptic vesicles were regarded as marked point patterns, where the points are the centers of the vesicles, and the mark of a vesicle is given by its size, shape, and orientation characteristics. Statistics for marked point processes were employed to study spatial interactions between vesicles. We found that the synaptic vesicles in excitatory synapses appeared to be of oblate ellipsoid shape and in inhibitory synapses appeared to be of cigar ellipsoid shape, and followed a systematic pattern regarding their orientation toward the active zone. Moreover, there was strong evidence of spatial alignment in the orientations of pairs of synaptic vesicles, and of repulsion between them only in excitatory synapses, beyond that caused by their physical extent.
Introduction
In classical neurobiology, synaptic vesicles (SVs) are frequently modeled as uniform spheres distributed randomly within the pre-synaptic terminal. However, this simplified view overlooks the complex physical constraints and forces that govern vesicle trafficking to the active zone (AZ). While two-dimensional transmission electron microscopy (TEM) has hinted at more complex morphologies, the inherent limitations of 2D imaging make it difficult to distinguish true 3D shapes from projection artifacts. This study addresses these gaps by quantifying the 3D shape, orientation, and spatial distribution of SVs using high-resolution isotropic imaging and advanced spatial statistics.
Methodology
The researchers utilized Focused Ion Beam Scanning Electron Microscopy (FIB-SEM) to obtain 3D volumes of the mouse hippocampus (CA1 region) with an isotropic resolution of 5 nm. This high resolution is critical for accurate 3D reconstruction without the distortions common in anisotropic datasets.
Data Annotation and Modeling
The study involved the manual annotation of six synapses: five excitatory (asymmetric) and one inhibitory (symmetric). Approximately 500 vesicles were segmented for the excitatory synapses and 131 for the inhibitory one. Each vesicle was modeled as an ellipsoid using a least-squares fit to its boundary.
Spatial Statistical Analysis
The synapse was treated as a marked point pattern, where the "points" represent vesicle centers and "marks" represent physical attributes like size, shape, and orientation. To detect non-random spatial interactions, the observed data were compared against a "hard core" null model generated via Markov Chain Monte Carlo (MCMC) simulations. This model accounts for physical exclusion (vesicles cannot overlap) and structural boundaries (e.g., mitochondria). Spatial density was analyzed using $L$-functions, while mark variograms were used to assess correlations in orientation between neighboring vesicles.
Key Insights and Results
The analysis revealed that synaptic vesicles are significantly non-spherical. In excitatory synapses, vesicles typically take an oblate (flattened) shape, whereas inhibitory vesicles tend to be prolate (cigar-shaped). In excitatory synapses, the average ratio of the longest to shortest radii ($a/c$) was 1.96, a marked departure from the 1.0 ratio of a perfect sphere.
Systematic Orientation and Alignment
Vesicles demonstrated a systematic orientation relative to the active zone. In excitatory synapses, the shortest axis of the ellipsoid often aligns perpendicular to the path toward the AZ, potentially minimizing drag or physical resistance during transport. Furthermore, "spatial alignment" was observed: neighboring vesicles tend to share similar orientations, suggesting a shared response to local cytoskeletal elements or organizing forces.
Evidence of Spatial Repulsion
A major finding was the presence of "extra" repulsion in excitatory synapses. The $L$-function for these synapses fell below the 2.5% quantile of the null model at distances up to 50 nm. This indicates that vesicles are spaced further apart than can be explained by simple volume exclusion, likely due to electrostatic repulsion between their negatively charged membranes. Interestingly, this additional repulsion was not observed in the inhibitory synapse sample.
Conclusion
By moving beyond the spherical model of synaptic vesicles, this research provides a more nuanced understanding of the physical environment within the presynaptic terminal. The discovery of shape-specific orientations and extra-physical repulsion suggests that vesicle trafficking is a highly organized process governed by both geometric and electrostatic constraints. Future work utilizing automated deep-learning segmentation will be necessary to validate these findings across larger, more diverse synaptic populations.
Summary updated