Molecular CellCell BiologyPeer-reviewed journal article
Towards community-driven visual proteomics with large-scale cryo-electron tomography of Chlamydomonas reinhardtii
Ron KelleyiD, Sagar KhavnekariD, Ricardo D. Righetto, Jessica HeebneriD, Martin ObriD, Xianjun ZhangiD, Saikat Chakraborty, Grigory Tagiltsev, Alicia K. MichaeliD, Sofie van DorstiD, Florent WaltziD, Caitlyn L. McCafferty, Lorenz LammiD, Simon ZuffereyiD, Philippe Van der StappeniD, Hugo van den HoekiD, Wojciech WietrzyñskiiD, Pavol Harar, William WaniD, John A.G. Briggs, Jürgen M. PlitzkoiD, Benjamin D. EngeliD, Abhay KotechaiD
Abstract
In situ cryo-electron tomography (cryo-ET) has emerged as the method of choice to investigate structures of biomolecules in their native context. However, challenges remain in the efficient production of large-scale cryo-ET datasets, as well as the community sharing of this information-rich data. Here, we applied a cryogenic plasma-based focused ion beam (cryo-PFIB) instrument for high-throughput milling of the green alga Chlamydomonas reinhardtii, a useful model organism for in situ visualization of numerous fundamental cellular processes. Combining cryo-PFIB sample preparation with recent advances in cryo-ET data acquisition and processing, we generated a dataset of 1829 reconstructed and annotated tomograms, which we provide as a community resource to drive method development and inspire biological discovery. To assay the quality of this dataset, we performed subtomogram averaging (STA) of both soluble and membrane-bound complexes ranging in size from >3 MDa to [~]200 kDa, including 80S ribosomes, Rubisco, nucleosomes, microtubules, clathrin, photosystem II, and mitochondrial ATP synthase. The majority of these density maps reached sub-nanometer resolution, demonstrating the potential of this C. reinhardtii dataset, as well as the promise of modern cryo-ET workflows and open data sharing towards visual proteomics.
- organism: plant (other)
- organism: C. elegans
- assay: imaging
- assay: protein structure
- assay: cryo-EM
- research focus: database / resource
- approach: experimental assay
- approach: deep learning
In-depth summary
Claude Opus 5.5 · xhigh effort · 5 Nov 2026 · from the full text of version 1 (28 Dec 2024) · CC BY 4.0
Automated cryogenic plasma-focused ion beam milling and cryo-electron tomography of Chlamydomonas reinhardtii yield 1,829 annotated cellular tomograms and sub-nanometer in situ structures of six macromolecular complexes.
In situ visual proteomics requires large, high-quality cellular cryo-electron tomography (cryo-ET) datasets, but traditional gallium ion milling is low-throughput and public cellular cryo-ET repositories remain scarce. Combining automated plasma-focused ion beam (cryo-PFIB) milling with high-throughput cryo-ET produces a 28 TB open community dataset spanning diverse eukaryotic organelles.
Key results
ShownAutomated xenon and argon cryo-PFIB milling paired with cryo-ET generated 1,829 curated, annotated tomograms across Chlamydomonas reinhardtii organelles.
Preparing lamellae took ~23 min with xenon (4 min site preparation, 19 min milling) and ~26 min with argon; 2,991 tilt-series collected across 42 sessions (55 grids, >300 lamellae, 1.96 A pixel size) yielded 1,829 quality tomograms (median thickness 157 nm) covering chloroplasts (823, including 84 pyrenoids), mitochondria (581), nuclei (387), Golgi (246), and centrioles (16). · Cryo-PFIB milling to enable large-scale cryo-ET; Towards visual proteomics in Chlamydomonas; Figure legends; Tables
ShownSubtomogram averaging (STA) of 80S ribosomes reaches 4.0 A resolution, resolves functional translation and ER-bound states, and quantifies PFIB surface radiation damage and optimal lamella thickness.
Averaging ~140,000 ribosome subvolumes from 600 tomograms reached 4.0 A global resolution (3.9 A in the large subunit) and separated 5 translational states plus ER-bound sec61-TRAP complexes; Rosenthal-Henderson B-factors were higher at 15-30 nm depth (B-f = 166) than 30-45 nm (B-f = 118) and in lamellae >225 nm thick, indicating optimal STA at <225 nm thickness and 30-100 nm depth. · Analysis of radiation damage from PFIB using 80S…; STA: 80S Ribosome; Figure legends
ShownIn situ STA resolves soluble and filamentous complexes ranging from ~200 kDa to >3 MDa (including Rubisco, nucleosomes, microtubules, and clathrin coats) at sub-nanometer resolution.
STA yielded density maps at 7.5 A for pyrenoid Rubisco (~520 kDa; 6% of picked particles), 9.6 A for nuclear nucleosomes (~200 kDa; 23,878 of 223,507 particles, ~11%, revealing an extra H2B alpha-helical density), 4.7 A for 13-protofilament cytoplasmic microtubules, and 8.7 A for cytosolic clathrin triskelia (17,997 symmetry-expanded particles). · Subtomogram averaging of diverse molecular…; Rubisco in the chloroplast; Nucleosomes in the nucleus; Microtubules in the cytosol
ShownMembrane-embedded complexes are resolved in situ at 5.2 A for mitochondrial ATP synthase peripheral stalks and 19 A for chloroplast photosystem II (PSII) using neural-network particle picking.
Co-refining mitochondrial ATP synthase dimers (~1.3 MDa, 261 tomograms) with 80S ribosomes resolved the peripheral stalk to 5.2 A (~88,000 particles, local resolution 4.8 A) and rotary F1 head states to 8.5-8.6 A (~11,000 particles), revealing absence of the Polytomella ASA1 bridge; a 2.5D U-net picked 26,827 thylakoid PSII particles (21 tomograms) to yield a 19 A map (3,122 particles). · Photosystem II in the chloroplast; ATP synthase in mitochondria; STA: Photosystem II; STA: ATP synthase and multispecies refinement…
Methods, data and reproducibility
Plunge-froze Chlamydomonas reinhardtii (strain mat3-4, CC3994) on EM grids, milled lamellae with xenon or argon cryo-PFIB (Arctis), and collected 2,991 dose-symmetric tilt-series on a Krios G4 with Selectris X and Falcon 4i. Reconstructed and denoised 1,829 tomograms, segmented organelles and membranes using U-nets and MemBrain, and performed STA across 7 macromolecular complexes.
- Curated cryo-ET tilt-series and reconstructed, denoised tomograms of Chlamydomonas reinhardtii mat3-4 (CC3994) cells · plant (other) (Chlamydomonas reinhardtii) · n = 1829 tomograms from 2991 tilt-series (55 grids, >300 lamellae, 28 TB) · EMPIAR-11830
- Annotated reconstructed tomograms and membrane/particle segmentations on CZII CryoET Data Portal · plant (other) (Chlamydomonas reinhardtii) · n = 1829 tomograms · DS-10302
- Cryo-PFIB/SEM slice-and-view volume imaging of a whole vitreous Chlamydomonas reinhardtii cell · plant (other) (Chlamydomonas reinhardtii) · n = 1 whole cell (335 slices, 20 nm slice thickness) · EMPIAR-11275
- Subtomogram average density maps of 80S ribosome, nucleosome, photosystem II, Rubisco, microtubules, clathrin, and ATP synthase in EMDB · plant (other) (Chlamydomonas reinhardtii) · n = 7 subtomogram maps (EMD-51847, EMD-19906, EMD-51731, EMD-51848, EMD-51804, etc.) · EMD-51847
- Methods & tools
- TOMOMAN, MotionCor2, CTFFIND4, AreTomo, IMOD, cryo-CARE, novaCTF, STOPGAP, PyTom, Warp, RELION-4, MemBrain
- Code
- https://github.com/Chromatin-Structure-Rhythms-Lab/ChlamyAnnotationshttps://github.com/turonova/Fourier3Dhttps://gitlab.mpcdf.mpg.de/bturo/gapstop_tmhttps://github.com/SBC-Utrecht/pytom-match-pick
- Data & accessions
- https://cryoetdataportal.czscience.com/datasets/10302/
- Software & environment
- Dragonfly 2022.2; Amira 2023.2 and v.6.2.0; MemBrain v2; RELION v3.0, v3.1.4, and RELION-4; Warp v1.0.9; M v1.0.9; Matlab R2022b (S21, S22, S25-S29).
Limitations (3)
- Substantial tilt-series discard rates (only 1,829 of 2,991 tilt-series reconstructed) occurred primarily due to inconsistent sample vitrification and alignment issues across grids (S3). · stated by the authors
- For small complexes in crowded environments such as Rubisco and nucleosomes, 3D classification discarded 89-94% of particles initially picked by template matching, preventing a complete spatial census of these complexes within cellular volumes (S7, S8, S14). · stated by the authors
- Strong membrane density and contrast transfer function (CTF) fringing artefacts from straight stacked thylakoid membranes caused standard template matching to fail for photosystem II and limited its subtomogram average resolution to 19 A (S11, S14). · stated by the authors
Read this if you develop cryo-electron tomography particle-picking, segmentation, or subtomogram averaging algorithms, or study native organelle and macromolecular architecture in Chlamydomonas reinhardtii.
- Citations
- 59
- FWCI
- 18.60
- Field percentile
- top 1%
Reproducibility ledger
Nobody has reported trying to reproduce this paper yet. Did you replicate an experiment or run the code? Your report saves the next lab a week.
Sign in to file a reproducibility report.
Related papers
- Optimizing data quality and completeness in visual proteomics experiments
Joseph M. Dobbs, Julia Mahamid·bioRxiv·Apr 14, 2026
- Making plant tissue accessible for cryo-electron tomography
Matthias Pöge, Marcel Dickmanns, … Wolfgang P. Baumeister·bioRxiv·Feb 14, 2025
- Cryo-electron tomography pipeline for plasma membranes.
Willy W. Sun, Dennis J. Michalak, … Justin W. Taraska·Nature CommunicationsPeer-reviewed·Jan 20, 2025
- Streamlined Structure Determination by Cryo-Electron Tomography and Subtomogram Averaging using TomoBEAR
N. Balyschew, A. Yushkevich, … M. Kudryashev·bioRxiv·Jan 10, 2023
- A collection of yeast cellular electron cryotomography data
L. Gan, C. T. Ng, … S. Cai·GigaSciencePeer-reviewed·Jun 27, 2019
- Fully automated, sequential focused ion beam milling for cryo-electron tomography
Tobias Zachs, Andreas Schertel, … Martin Pilhofer·eLifePeer-reviewed·Mar 9, 2020