Data from: The N’ terminus of Alpha-1 Giardin, a parasitic annexin orthologue, is essential for oligomerization and lipid-binding activity
Data files
Jun 22, 2026 version files 15.26 GB
-
README.md
8.86 KB
-
Supplementary_video_1.gif
468.95 KB
-
Supplementary_video_10.gif
660.97 MB
-
Supplementary_video_11.gif
604.38 MB
-
Supplementary_video_12.gif
633.92 MB
-
Supplementary_video_13.gif
936.46 MB
-
Supplementary_video_14.gif
1.05 GB
-
Supplementary_video_15.gif
936.45 MB
-
Supplementary_video_16.gif
1.01 GB
-
Supplementary_video_17.gif
733.20 MB
-
Supplementary_video_18.gif
676.06 MB
-
Supplementary_video_19.gif
737.44 MB
-
Supplementary_video_2.gif
806.19 KB
-
Supplementary_video_20.gif
618.23 MB
-
Supplementary_video_21.gif
805.25 MB
-
Supplementary_video_22.gif
705.05 MB
-
Supplementary_video_23.gif
570.36 MB
-
Supplementary_video_24.gif
1.41 GB
-
Supplementary_video_3.gif
1.40 MB
-
Supplementary_video_4.gif
1.11 MB
-
Supplementary_video_5.gif
518.80 MB
-
Supplementary_video_6.gif
768.52 MB
-
Supplementary_video_7.gif
675.28 MB
-
Supplementary_video_8.gif
607.28 MB
-
Supplementary_video_9.gif
596.11 MB
Abstract
Annexins are Ca²⁺- and membrane-binding proteins conserved across eukaryotes and involved in diverse membrane-associated processes, including unconventional protein secretion (UPS). In the intestinal parasite Giardia lamblia, annexins are represented by a family of 21 proteins termed alpha-giardins (αGs), all of which lack classical signal peptides yet are detected at the parasite surface and in the secretome. Here, we present a comprehensive functional analysis of αGs, with a particular focus on α1-giardin (α1G), a leading vaccine candidate against giardiasis and a putative virulence-associated UPS substrate. We show that αGs localize to peripheral endocytic compartments (PECs-PVs), specialized endo-lysosomal organelles implicated in protein release, and engage in multiprotein complexes enriched in UPS substrates. Using lipid-binding assays, cross-linking mass spectrometry (XL-MS), and structural modelling, we define conserved and divergent features of αG membrane interactions. Focusing on α1G, we combine site-directed mutagenesis, confocal microscopy, and mass photometry to demonstrate a critical role for its short N-terminal region in oligomerization and membrane association. Integration of these data with molecular dynamics simulations reveals how α1G oligomeric state and membrane occupancy could be regulated at PECs-PVs membranes. Taken together, our findings provide structural insights relevant to parasite virulence and vaccine development and contribute to a mechanistic model for annexin-mediated UPS in Giardia.
Dataset DOI: 10.5061/dryad.02v6wwqk4
Description of the data and file structure
README file for supplementary videos 1-24. Preprint DOI: https://doi.org/10.64898/2026.01.09.697555
Live cell imaging of mNeon-Green tagged alpha giardin reporters expressed in transgenic Giardia lamblia cells.
Experimental context: To interrogate αlpha Giardin (αG) behaviour in living Giardia lamblia cells, we engineered and transfected constructs for expression of: full-length α1G fused to mNeonGreen (mNG) at the C terminus (line G370), an mNG C-terminally tagged version of the α1G variant with a deletion of the first three amino acid residues (Δ3α1G; line G390), and an mNG C-terminally tagged α8G variant (line G391). A line engineered to express free mNG (line G369) was included as a control.
Live cell imaging files 1-4 show how, in contrast to the mNG fusion to full-length α1G which associates to endocytic compartments called PECs-PVs, and in line with previous data, the -3 N-terminal deletion construct Δ3α1G fused to mNG (line G390) loses PECs-PVs targeting and mimics more closely the deposition of both a free mNG (line G369) and the mNG fusion to α8G (line G391).
Approach: Transgenic cell lines were generated by electroporation, selected and grown in complete growth medium, and imaged in a widefield fluorescent microscope.
File: Supplementary_video_1.gif
Description: Line G369: live cell imaging of free mNG
File: Supplementary_video_2.gif
Description: Line G370: live cell imaging of full-length α1G fused to mNeonGreen (mNG) at the C terminus
File: Supplementary_video_3.gif
Description: Line G390: live cell imaging of an mNG C-terminally tagged version of the α1G variant with a deletion of the first three amino acid residues (Δ3α1G)
File: Supplementary_video_4.gif
Description: Line G391: live cell imaging of an mNG C-terminally tagged α8G variant
Molecular dynamics of oligomeric alpha1-giardin variants and membrane binding simulations
Experimental context: To investigate the membrane-binding behaviour of the potentially native α1G trimer, molecular dynamics (MD) simulations were performed across five distinct lipid compositions with the α1G trimer. The base membrane contained phosphatidylcholine (PC), phosphatidylserine (PS), and cholesterol, while the other systems included either 3% or 12% of phosphatidyl-inositol-3 or -5 -phospahte (PI(3)P or PI(5)P).
Videos of the simulation taken from different perspectives are shown in video files 5-24, specifying whether a base membrane was used in the simulation (base) or whether the base membrane with PC (base) was enriched with PIP residues, and at which percentage (3percPI3P, 3percPI5P, 12percPI3P, 12percPI5P).
Approach:
a) Protein Modelling and Initial Structure Preparation
To investigate the interaction between α1G and lipid bilayers, a trimeric model of α1G was constructed. The monomeric structure was predicted using the AlphaFold web server, excluding the N-terminal methionine (Met1) to reflect the mature protein form, resulting in all residue numbers in the molecular dynamics simulations being shifted by −1. A trimeric configuration was assembled using the HADDOCK v2.4 docking platform. The trimer was selected for simulation based on evidence from mass spectrometry indicating its oligomeric state.
Five distinct membrane systems were constructed using the CHARMM-GUI Membrane Builder webserver, with each consisting of approximately 350 lipid molecules per leaflet. All systems included PI(3)P and PI(5)P lipids at varying concentrations to assess their influence on α1G binding. Water molecules were removed from the unit cell, and the α1G trimer was positioned above the membrane at a minimal distance of 10 Å from the surface to allow for potential interaction while avoiding initial steric clashes.
b) Molecular dynamics simulations
The membrane system preparation was continued in CHARMM-GUI, with all components parameterized using the CHARMM36 force field, generating GROMACS-compatible files.
The TIP3P water model was used to solvate each system in a simulation box of averaged dimensions 170 × 140 × 140 Å, containing approximately 350000 atoms. Appropriate numbers of Na⁺ and Cl⁻ counterions were added to ensure electrostatic neutrality and physiological ionic strength (~0.15 mM).
Energy minimization was performed using the steepest descent algorithm, with a convergence Vmax of 1000 kJ·mol⁻¹·nm⁻¹. Partial equilibration was conducted on the CHARMM GUI server using the NVT and NPT ensembles to gradually relax positional restraints on the thermos-coupled components of the systems. Electrostatic interactions were treated using the particle mesh Ewald (PME) method, while van der Waals forces were computed via a force-switch cutoff scheme with a switching distance of 1.0 nm and a cutoff of 1.2 nm. Temperature coupling was applied separately to the solute, membrane, and solvent groups using the V-rescale thermostat with a reference temperature of 310 K and a coupling constant of 1.0 ps. Pressure was maintained at 1 bar using the C-rescale barostat with semi-isotropic coupling, a compressibility of 4.5×10⁻⁵ bar⁻¹, and a relaxation time of 5.0 ps. Finally, maintaining the same barostat, production runs of 200 ns were carried out using GROMACS 2023.3 with a leap-frog integrator and a time step of 2 fs, applying the LINCS algorithm to constrain all hydrogen-containing bonds. All simulations, including the final equilibration and production runs, were performed using 1 CPU core and GPU RTX4090 of University of Bern cluster UBELIX. Calculations were performed on UBELIX (https://www.id.unibe.ch/hpc), the HPC cluster at the University of Bern. Protein-lipid interaction was performed with the Python library PyLipID ^with a dual cutoff of 0.15 nm – 0.25nm and bootstrapping of Koff values. The dual cutoff values were selected based on the protocol in the PyLipID library documentation. Bootstrapping was automatically run by the PyLipID library based on duplicate simulations. Residues were selected based on Koff values suggested by Song et al. The duration and occupancy thresholds were set in proportion to the total simulation time and matched with the interaction behavior we wanted to focus on.
For all files below:
Colour legend to molecular dynamics simulation:
PIPs - Orange
PS - Yellow
Other lipids - Grey
Interacting residues - White
File: Supplementary_video_5.gif
Description: 5. a1gtrimerdel1Met_base_meme_1_side
File: Supplementary_video_6.gif
Description: 6. a1gtrimerdel1Met_base_meme_1_top
File: Supplementary_video_7.gif
Description: 7. a1gtrimerdel1Met_base_meme_2_side
File: Supplementary_video_8.gif
Description: 8. a1gtrimerdel1Met_base_meme_2_top
File: Supplementary_video_9.gif
Description: 9. a1gtrimerdel1Met_3percPI3P_1_side
File: Supplementary_video_10.gif
Description: 10. a1gtrimerdel1Met_3percPI3P_1_top
File: Supplementary_video_11.gif
Description: 11. a1gtrimerdel1Met_3percPI3P_2_side
File: Supplementary_video_12.gif
Description: 12. a1gtrimerdel1Met_3percPI3P_2_top
File: Supplementary_video_13.gif
Description: 13. a1gtrimerdel1Met_3percPI5P_1_side
File: Supplementary_video_14.gif
Description: 14. a1gtrimerdel1Met_3percPI5P_1_top
File: Supplementary_video_15.gif
Description: 15. a1gtrimerdel1Met_3percPI5P_2_side
File: Supplementary_video_16.gif
Description: 16. a1gtrimerdel1Met_3percPI5P_2_top
File: Supplementary_video_18.gif
Description: 18. a1gtrimerdel1Met_12percPI3P_1_top
File: Supplementary_video_17.gif
Description: 17. a1gtrimerdel1Met_12percPI3P_1_side
File: Supplementary_video_19.gif
Description: 19. a1gtrimerdel1Met_12percPI3P_2_side
File: Supplementary_video_20.gif
Description: 20. a1gtrimerdel1Met_12percPI3P_2_top
File: Supplementary_video_21.gif
Description: 21. a1gtrimerdel1Met_12percPI5P_1_side
File: Supplementary_video_22.gif
Description: 22. a1gtrimerdel1Met_12percPI5P_1_top
File: Supplementary_video_23.gif
Description: 23. a1gtrimerdel1Met_12percPI5P_2_side
File: Supplementary_video_24.gif
Description: 24. a1gtrimerdel1Met_12percPI5P_2_top
Code/software
any digital video player software will do, for example, VLC player
