henhaobio
Sign in

CellCell BiologyPeer-reviewed journal article

Charting the landscape of cytoskeletal diversity in microbial eukaryotes

Felix MikusiD, Armando Rubio-RamosiD, Hiral ShahiD, Jonas HellgothiD, Marine OlivettaiD, Susanne Borgers, Clémence Saint-Donat, Margarida AraújoiD, Chandni BhicktaiD, Paulina CherekiD, Jone BilbaoiD, Estibalitz TxurrukaiD, Yana EglitiD, Nikolaus LeischiD, Yannick SchwabiD, Filip HusníkiD, Sergio SeoaneiD, Ian ProbertiD, Paul GuichardiD, Virginie HameliD, Gautam DeyiD, Omaya DudiniD

PDFDOIPreprintCodeData
In-depth full-text digest available · Jump to digest ↓

Abstract

Microbial eukaryotes are small and often resistant to standard labelling and imaging techniques, and therefore remain understudied - despite their critical ecological importance - with the exception of a few established models. Here, we use Ultrastructure Expansion Microscopy (U-ExM) to carry out high-resolution volumetric imaging of over 200 cultured planktonic eukaryotes across major lineages. By combining U-ExM with pan- and specific immuno-labelling, we reveal novel microtubule and centrin-containing elements and assign molecular identities to enigmatic cytoskeletal structures observed previously only by electron microscopy. Our investigation represents the first systematic survey of the extensive cytoskeletal diversity on display across the eukaryotic tree, including the major species groups of dinoflagellates, haptophytes, ciliates, euglenids, cryptomonads, and green algae. Our U-ExM approach extends to mixed environmental samples, paving the way for environmental cell biology at ultrastructural resolution and unprecedented scale, a crucial step towards understanding and protecting complex ecosystems in the face of biodiversity loss.

In-depth summary

Claude Opus 5.5 · xhigh effort · 8 Nov 2026 · from the full text of version 1 (18 Oct 2024) · CC BY 4.0

Ultrastructure expansion microscopy applied to 221 cultured planktonic eukaryotes maps diverse microtubule and centrin-based cytoskeletons, including dinoflagellate apical centrin rings and cryptophyte cortical centrin meshes.

Microbial eukaryotes often resist standard immunolabelling and their subcellular detail lies beyond conventional light microscopy, while electron microscopy is slow, costly and cannot label target molecules, so cell biology has centred on a few models. A standardised, mainly chemically fixed ultrastructure expansion microscopy (U-ExM) workflow is shown to scale to hundreds of cultured species and mixed environmental cultures, assigning molecular identities to cytoskeletal structures.

Key results

  1. ShownOne standardised U-ExM workflow using chemical fixation expanded and labelled most of the cultured species tested, and no major eukaryotic lineage was entirely refractory. The images form an openly shared, atlas-scale resource.

    Of 221 species, 79% (175 species) expanded and 77% (168 species) showed at least partial staining with both NHS-ester pan-labelling and antibodies; 1,670 image volumes, approximately 4TB, ∼200 hours of microscopy within a 16-month fixation-to-imaging effort; thylakoid stacks resolved at 180-240 nm spacing in Urgorri complanatus. · Volumetric imaging of ultrastructure across the…; Discussion

  2. ShownVolumetric U-ExM shows dinoflagellate chromosome banding, previously visible only by TEM, and quantifies closed-mitosis spindles in unsynchronised cultures, including membrane-lined nuclear tunnels that contain microtubule bundles.

    Chromosome band spacing 0.16 ± 0.021 μm in Karenia papilionaceae (n = 116 band pairs, 2 nuclei) vs 0.193 ± 0.024 μm in Prorocentrum (n = 44 pairs); Protodinium sp. spindles held 8.25 ± 1.25 fibres (n = 16 spindles), angled 55.8 ± 17.8 to the cingulum (n = 6); BODIPY ceramide confirmed membrane tunnels around spindle bundles. · Structural features of dinoflagellate chromosome…; Figure legends

  3. ShownCentrin (20H5 antibody) marks varied dinoflagellate structures. These include apical rings, open-sided in Heterocapsa rotundata and closed in Brandtodinium nutriculum, from which cortical microtubule bundles emanate. The authors propose these rings may be conoid-like.

    Open and closed rings differ in diameter but share a perimeter averaging 1.8 µm; microtubule bundles per ring 9±1 in H. rotundata (n = 9 cells) vs 11±2 in B. nutriculum (n = 6 cells), joined via an unidentified NHS-ester-dense linker; centrin may also contain or sit alongside flagellar R-fibres and is anticipated in basal-body striated collars. · Centrin as a cytoskeletal organiser in…; Figure legends

  4. ShownCentrin fibres occur in many lineages. Examples include nucleus-to-basal-body connectors in Dunaliella salina and a periodic cortical centrin mesh in some cryptophytes (absent in Cryptomonas curvata), which show no evidence of cortical basal bodies; this points to a structural cytoskeletal role.

    Cryptophyte mesh period, longitudinal vs transversal: Rhodomonas sp. 192.6 ± 67 vs 525.8 ± 132 nm (n = 7 cells), Rhinomonas nottbecki 193 ± 81 vs 550.1 ± 72 nm (n = 5 cells); rootlet centrin striation 205.9 ± 35 nm in Pyramimonas diskoicola (n = 5 cells) and 191.3 ± 21 nm in Microglena reginae (n = 6 cells), close to 170-220 nm in Tetraselmis subcordiformis. · A centrin-based cytoskeleton is a defining…; Figure legends

  5. SuggestedU-ExM also worked on enriched mixed cultures from a single Tokyo Bay sample, giving 3D maps of ciliate feeding and contractile cytoskeletons. In Euplotes rariseta it showed previously unreported centrin fibres between the para-oral membrane and adoral zone of membranelles (AZM).

    At least four ciliates and one amoeba (Flabellula sp.) imaged; E. rariseta orthogonal tubulin bundles spaced on average 90 nm; Lacrymaria oral centrin ring approximately 4 µm in diameter (n = 3 cells); Vorticella oral centrin ring widened and thinned during feeding; only Flabellula sp. and E. rariseta confirmed by 18S sequencing. · Cytoskeletal diversification and specialisation…; Culturing of environmental samples and…; Figure legends

Methods, data and reproducibility

Cultures from two European collections were formaldehyde-fixed (a few cryofixed), anchored, embedded in swellable gels, denatured and expanded. They were then stained after expansion with an NHS-ester protein dye, DNA dyes and antibodies to tubulin, tubulin modifications and centrin (20H5), and imaged by confocal, Airyscan, STED or widefield microscopy.

  • Near-axenic planktonic eukaryote cultures from the Roscoff Culture Collection (RCC) and Basque Microalgae Culture Collection (BMCC) spanning major lineages · multiple microbial eukaryote lineages · n = 221 species (175 expanded; 168 at least partially stained)
  • Enriched mixed ciliate-dominated cultures from a shallow-water Tokyo Bay sample collected 13.06.2023 · ciliates (e.g., Euplotes rariseta) and amoeba Flabellula sp. · n = at least four ciliates and one amoeba
  • Public U-ExM image resource (MoBIE project in ome-zarr format) plus a curated Data S1 PDF covering 60 species · multiple microbial eukaryote lineages · n = 1,670 image volumes from 221 species
Methods & tools
Ultrastructure Expansion Microscopy (U-ExM), NHS-ester pan-protein labelling (DyLight 405, Alexa Fluor 594), Anti-centrin 20H5, anti-α/β-tubulin and tubulin glutamylation/glycylation antibodies, BODIPY TR ceramide membrane stain; Hoechst 33342 or DAPI DNA stain, High-pressure freezing (Leica EM ICE) and plunge freezing with freeze substitution, Nikon CSU-W1 SoRa and Olympus IXplore SpinSR spinning-disk confocal microscopes, Zeiss LSM980 Airyscan and Leica SP8 STED 3X microscopy, Fiji with PatternJ toolset, Imaris 10.1.0 and Amira for 3D reconstruction and segmentation, MoBIE plugin, mobie-utils-python and bioformats2raw, 18S locus amplicon sequencing for species identification, RStudio, Excel and GraphPad Prism for quantification
Software & environment
Fiji with PatternJ; Imaris 10.1.0; Amira v.2019.3 or 2020.1; GraphPad Prism 9 or 10.3.1; RStudio; OME-Zarr conversion via bioformats2raw (chunk size 96 x 96 x 96), MoBIE python API and MinIO client upload.
Experimental conditions
Computational hardware not reported; about 4TB of data came from ∼200 hours of microscopy, within a 16-month effort covering fixation, expansion and imaging.
Reproducibility gaps
The public repository holds only MoBIE upload scripts and metadata. Quantification scripts (a line-plot script drafted with ChatGPT-4 and finished in RStudio, plus Excel peak analysis) are not deposited, and the 18S sequences have no stated accession.
Limitations (6)
  • The specificity of the anti-centrin 20H5 antibody cannot be validated directly, because these species lack genetic tools to deplete endogenous centrin. Support rests on the antibody's prior use in other protists and on centrin sequence conservation. · stated by the authors
  • Formaldehyde fixation was chosen because it scales, but fixative choice affects which macromolecules are preserved: methanol suits microtubules and cryofixation better preserves some architecture. Species also respond variably to dyes, antibodies and fixation. · stated by the authors
  • Expansion and staining efficiency varied by species, likely because of cell wall properties. Incomplete expansion in Pseudourostyla sp. and some Vorticella sp. cells produced artefacts. · stated by the authors
  • The authors cannot exclude that the dinoflagellate apical centrin rings are unrelated to the alveolate conoid, so the proposed evolutionary link is still a hypothesis. · stated by the authors
  • The environmental demonstration uses one Tokyo Bay sample grown as enriched mixed cultures rather than fixed directly from the field, and most taxa were identified by morphology alone. This limits claims about scaling to natural communities. · noted in this summary
  • Many measurements come from a few cells of a single isolate (e.g., n = 3 and 4 Protodinium spindles), and a second Rhodomonas sp. culture showed a different centrin pattern, so generalising to whole species is uncertain. · noted in this summary

Read this if you study protist or plankton cell biology or cytoskeletal evolution (centrin, microtubules, conoid origins), or want to scale expansion microscopy to non-model or environmental eukaryotes.

Citations
26
FWCI
6.90
Field percentile
top 2%

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

Similarity Map
SelectedSame fieldCross-field
THISFlori '25Gorilak '21Bertiaux '21Gambarotto '18Hinterndorf '22Mercey '26
66% semantic matchCell Biology
Serena Flori, Felix Mikus, Eliott Flaum+14 more·Current Biology·Oct 31, 2025· 10 citations