Skip to main page content
U.S. flag

An official website of the United States government

Dot gov

The .gov means it’s official.
Federal government websites often end in .gov or .mil. Before sharing sensitive information, make sure you’re on a federal government site.

Https

The site is secure.
The https:// ensures that you are connecting to the official website and that any information you provide is encrypted and transmitted securely.

Access keys NCBI Homepage MyNCBI Homepage Main Content Main Navigation
[Preprint]. 2024 Jun 6:2024.06.05.597487.
doi: 10.1101/2024.06.05.597487.

Scalable electron tomography for connectomics

Affiliations

Scalable electron tomography for connectomics

Aaron T Kuan et al. bioRxiv. .

Abstract

We demonstrate limited-tilt, serial section electron tomography (ET), which can non-destructively map brain circuits over large 3D volumes and reveal high-resolution, supramolecular details within subvolumes of interest. We show accelerated ET imaging of thick sections (>500 nm) with the capacity to resolve key features of neuronal circuits including chemical synapses, endocytic structures, and gap junctions. Furthermore, we systematically assessed how imaging parameters affect image quality and speed to enable connectomic-scale projects.

PubMed Disclaimer

Conflict of interest statement

Harvard University filed a patent application regarding GridTape (WO2017184621A1) on behalf of the inventors including W.C.A.L., and negotiated licensing agreements with interested partners. The other authors declare no competing interests.

Figures

Extended Data Fig. 1.
Extended Data Fig. 1.. Quantifications of ET reconstruction quality at 300 kV.
Quantifications of mean FSC resolution (top 2 rows) and correlations with reference images (bottom 2 rows) for limited-tilt ET reconstructions (Methods). Plots are arranged by sample thickness (columns: left: 500 nm, middle: 750 nm, right: 1000 nm) and magnification (1st and 3rd row: 3.3k, 2nd and 4th row: 11k). The number of projections is indicated by line and marker color (legend at bottom).
Extended Data Figure 2.
Extended Data Figure 2.. ET operating point image matrix for connectomics.
A mosaic of ET images across different tilt ranges (x-axis) and number of tilt angles (y-axis, evenly sampled over the tilt range) of an excitatory synapse in the cerebellar molecular layer. (top) X-Z and (right) Y-Z virtual slices through the tomographic volume at different operating points. The parameter set (Δθ = ±30, N = 21) outlined in purple is relevant to connectomic. Scale bar, 250 nm.
Extended Data Fig 3.
Extended Data Fig 3.. ET operating point parameterization at lower voltage and differing section thicknesses.
(a) Parameterization mosaic of ET images across different tilt ranges (x-axis) and number of tilt angles (y-axis, evenly sampled over the tilt range) of an unmyelinated axon bundle. Scale bars, 250 nm (b) Similar to Extended Data Fig. 2a with 500 nm thick sections. (c) Similar to Extended Data Fig. 2a with 1000 nm thick sections. (d) Parameterization mosaic of ET images of an excitatory synapse (similar to Extended Data Fig. 2a) at 200 kV.
Extended Data Fig. 4.
Extended Data Fig. 4.. Biological features in ET with differing section thickness.
(a) Similar to Fig. 1d with 250 nm thick sections, full (top row) and limited (bottom row) single-axis tomography. Scale bars: Unmyelinated neurites and Chemical synapses, 250 nm; Gap junctions, 100 nm; Spinules, 75 nm. (b) Similar to (a) from 500 nm thick sections. (c) Similar to (a) from 750 nm thick sections. (d) Similar to (a) from 1000 nm thick sections.
Figure 1.
Figure 1.. Overview of scalable electron tomography (ET).
(a) Schematic of ET data acquisition. 2D image projections (images, right) of the sample at a range of tilt angles (black and gray bars) are acquired using a transmission electron microscope. (b) 3D volume of the sample (tomogram) computationally reconstructed from multiple 2D tilt projections. (c) A comparison of ET images of a 750 nm semi-thick section from the molecular layer of an adult mouse cerebellar cortex. Different subportions of the image matrix show data reconstructed using different tilt ranges (x-axis) and number of tilt angles (y-axis, evenly sampled over the tilt range). X-Z (top bar) and Y-Z (right bar) virtual slices through the tomographic volume are shown from different operating points. Note, differences in data quality with different tilt ranges and angles. Scale bar, 500 nm. (d) Biological features including unmyelinated neurites, chemical synapses, gap junctions, and spinules from tomographic volumes acquired with full (left column) and limited-tilt (right column) tomography. Scale bars: Unmyelinated neurites and Chemical synapses, 250 nm; Gap junctions, 100 nm; Spinules, 50 nm. (e) Plot of Fourier shell correlation (FSC) resolution as a function of the number of imaged projections equally spaced across an angular range (Δθ) of ± 30º and at a magnification (mag) of 3.3k. Left y-axis indicates FSC resolution in nm, whereas the right y-axis indicates resolution in pixels. Colored lines denote nominal section thickness (color legend in (f), blue: 1000 nm, green: 750 nm, orange: 500 nm). Error bars indicate mean ± IQR. (f) Similar to (e) at 11k mag. (g) Plot of FSC resolution as a function of angular range (Δθ) for a constant number of imaged projections (n = 21). Colored lines denote nominal section thickness as in (e,f). Note, angular range corresponding to the highest resolution is less than ±30º and depends on thickness. Error bars indicate mean ± IQR.
Figure 2.
Figure 2.. Examples of ET volume stitching, segmentation, and multi-resolution embedding.
(a) ET images from 15 sequential 600 nm thick sections of mouse cerebellar cortex aligned in Z, showing orthogonal x,y, and z re-slices. Single axis tomography was conducted at low magnification (3.3k). (b) Segmentation of a granule cell parallel fiber axon (green) synapsing onto two Purkinje cell spines (brown), astrocytes (purple and blue) and spinules. Boxes outline the areas where multi-axis tomography at higher magnification (11k) was implemented. (c) Axon from (a-b, green) segmented (green outline) across 15 aligned 600 nm thick sections projected in an X-Z slice through the volume. Scale bar, 500 nm. (d) Multi-axis, high magnification (11k) tomogram featuring spinules (white arrowheads). Scale bar, 500 nm. (e) Multi-axis, high magnification (11k) tomogram featuring gap junction (purple arrowhead) and spinules (white arrowheads). Scale bar, 500 nm. (f) Multi-axis, high magnification (11k) tomogram featuring the synapse from (a-c). Green outline – parallel fiber axon. Red outlines – Purkinje cell spines. Scale bar, 500 nm.

References

    1. Eichler K. et al. The complete connectome of a learning and memory centre in an insect brain. Nature 548, 175–182 (2017). - PMC - PubMed
    1. Nguyen T. M. et al. Structured cerebellar connectivity supports resilient pattern separation. Nature 613, 543–549 (2023). - PMC - PubMed
    1. Ohyama T. et al. A multilevel multimodal circuit enhances action selection in Drosophila. Nature 520, 633–639 (2015). - PubMed
    1. Kuan A. T. et al. Synaptic wiring motifs in posterior parietal cortex support decision-making. Nature (2024) doi:10.1038/s41586-024-07088-7. - DOI - PMC - PubMed
    1. Azevedo A. et al. Tools for comprehensive reconstruction and analysis of Drosophila motor circuits. bioRxiv 2022.12.15.520299 (2022).

Publication types