Molecular architecture of glideosome and nuclear F-actin inPlasmodium falciparum

Obtaining sporozoites P. falciparum sporozoites (strain: NF54-ΔPf47-5’csp-GFP-Luc: expressing a GFP-Luciferase fusion protein under the control of the csp promoter, genomic integration, no selection marker) were prepared at TropIQ (Nijmegen, Netherlands). Gameto- cytes were fed to 2 day old female Anopheles stephensi mos- quitoes. Mosquito infection was confirmed 7 days post feeding by midgut dissection. At 7 days post infection, mosquitoes re- ceived an extra non -infectious blood meal to boost sporozoite production. Two weeks post infection, sporozoites were isolated using salivary gland dissection and shipped at room temperature in Leibovitz medium with 10% heat inactivated human serum. Cryo-grid preparation P. falciparum sporozoites were checked under the fluorescent microscope and then diluted 1:4 into RPMI medium (without phenol red). 3 μl of parasites were applied onto a freshly plasma-cleaned UltrAufoil R1.2/1.3 300 mesh EM grid (Quanti- foil) in a humidity controlled facility. Excess liquid was manually back-blotted and grids were plunged into a reservoir of ethane/propane using a manual plunger. Grids were stored un- der liquid nitrogen until imaging. Cryo FIB milling Grids were clipped into autogrids modified for FIB preparation17 and loaded into either an Aquilos or an upgraded Aquilos2 cryo- FIB/SEM dual -beam microscope (Thermofisher Scientific). Overview tile sets were recorded using MAPS software (Ther- mofisher Scientific) before being sputter coated with a thin layer of platinum. Good sites with parasites were identified for lamella preparation before the coincident point between the electron beam and the ion beam was determined for each point by stage tilt. Prior to milling, an organometallic platinum layer was depos- ited onto the grids using a GIS (gas-injection-system). Lamellae were milled manually until under 300 nm in a stepwise series of decreasing currents. Milling was performed at the lowest possi- ble angles to increase lamella length in thin cells. Finally, polish- ing of all lamella was done at the end of the session as quickly as possible but always within 1.5 h to limit ice contamination from water deposition on the surface of the lamellae. Before re- moving the samples, the grids were sputter coated with a final thin layer of platinum. Grids were stored in liquid nitrogen for a maximum of 2 weeks before imaging in the TEM. Tilt-series collection Cryo-EM FIB-milled grids were rotated by 90° and loaded into a Titan Krios microscope (Thermofisher) equipped with a K3 direct electron detector and (Bio -) Quantum energy filter (Gatan). Tomographic data was collected with SerialEM with the energy- selecting slit set to 20 eV. Datasets were collected using the dose-symmetric acquisition scheme at a ± 65° tilt range with 3° tilt increments. For all datasets, 5 - 10 frames were collected and aligned on the fly using SerialEM and the total fluence was kept to less than 120 e −Å2. Defoci between 3 and 8 μm underfocus were used to record the tilt series’. Tomogram reconstruction Frames were aligned on the fly in SerialEM 18; CTF estimation, phase flipping and dose -weighting was performed in IMOD 19. Tilt-series’ were aligned in IMOD either using patch-tracking or by using nanoparticles (likely gold or platinum) on lamella sur- faces as fiducial markers. Tomograms were binned 4x and fil- tered in IMOD or by using Bsoft20. Subvolume averaging Subvolume averaging was performed using PEET 21 as de- scribed previously 22. Model processing was done using TEMPy23, Scipy 24, Scikit -learn25, Matplotlib 26 and Numpy 27 in Python 3. Initial models were generated manually by picking line segments using pairs of IMOD model points and then inter- polating particles at 1 voxel (1.3 nm) increments. The initial Y axes were aligned with the line segments and initially Y axis ro- tation angles were randomised. The initial reference was gener- ated by averaging particles with the starting orientations, thus generating a featureless cylinder. A small subset of particles (~700) were refined to create a reference with F -actin features which was then used for alignment of ~ 70k initial positions. Du- plicate and low scoring particles were removed. In order to im- prove model completeness and allow separation of particles into two independent halves, the subvolume positions were then fit- ted to a spline-smoothed helical model allowing for small varia- tion in helical pitch (Figure S2). Subvolume positions were then generated based on the best fitting model parameters. These were split into two halves and aligned independently. Overlap- ping particles between the two half-maps were removed before generating final half-maps. Fourier Shell Correlation was meas- ured using Bsoft, suggesting 27 Å resolution at the 0.143 cutoff. Particles from the two half -datasets (11487 total) were then combined and aligned together. The final volume was sharp- ened using Bsoft with an arbitrarily chosen B-factor of -3000 for fitting and visualisation. Segmentation and visualisation Membrane segmentation was performed in IMOD, using draw- ing tools followed by linear interpolation. These were then resampled using open3d to achieve an isotropic coordinate dis- tribution, which were then used to generate a volume using IMOD imodmop. F-actin, microtubules, apical polar ring and pre- conoidal rings were backplotted: average volumes were placed into 3D volumes using coordinates determined by SVA. Actin and microtubule models were smoothed for backplotting. Sur- face visualisation was performed usin g UCSF ChimeraX or open3d. Volume sections were visualised using IMOD 3dmod. Plots were generated using Matplotlib. Length measurements Filament lengths for comparison of nuclear, cytosolic and pellic- ular filament lengths were derived from helical models based on subvolume averaging positions (see above). Filament lengths for comparison of apical, lateral and basal pellicular filament lengths were measured manually using 3dmod. F-actin concentration The number of actin subunits in observed F-actin was estimated from subvolume averaging (15,058) and manual length meas- urements (17165, assuming 38 nm per 13 subunits). The sub- volume averaging-derived value is likely an underestimate due to cross-correlation based particle cleaning; it is the number of particles after the first alignment step of the two independent da- tasets. The two estimates were used to calculate the experi- mental error, expressed as standard deviation. The total ob- served volume of 29 tomo grams with an average thickness of 244 nm was 4.2 x 10 -17 m3, of which cells made up approxi- mately 7/12. 9.7 x 10-19 mol in 2.4 x 10-14 L corresponds to 4.0 x 10-5molL-1 .CC-BY 4.0 International licensemade available under a (which was not certified by peer review) is the author/funder, who has granted bioRxiv a license to display the preprint in perpetuity. It is The copyright holder for this preprintthis version posted April 22, 2024. ; https://doi.org/10.1101/2024.04.22.590301doi: bioRxiv preprint Molecular architecture of glideosome and nuclear F-actin in Plasmodium falciparum Pražák et al. 2024 (preprint) 6

1. Cowman, A. F., Healer, J., Marapana, D. & Marsh, K. Malaria: Biology and Disease. Cell 167, 610–624 (2016). 2. Amino, R. et al. Quantitative imaging of Plasmodium transmission from mosquito to mammal. Nat. Med. 12, 220–224 (2006). 3. Heintzelman, M. B. Gliding motility in apicomplexan parasites. Semin. Cell Dev. Biol. 46, 135–142 (2015). 4. Schmitz, S. et al. Malaria Parasite Actin Filaments are Very Short. J. Mol. Biol. 349, 113–125 (2005). 5. Vahokoski, J. et al. Structural Differences Explain Diverse Functions of Plasmodium Actins. PLoS Pathog. 10, e1004091 (2014). 6. Lu, H., Fagnant, P. M. & Trybus, K. M. Unusual dynamics of the diver- gent malaria parasite PfAct1 actin filament. Proc. Natl. Acad. Sci. U. S. A. 116, 20418–20427 (2019). 7. Kumpula, E.-P. et al. Apicomplexan actin polymerization depends on nucleation. Sci. Rep. 7, 12137 (2017). 8. Yee, M., Walther, T., Frischknecht, F. & Douglas, R. G. Divergent Plas- modium actin residues are essential for filament localization, mos- quito salivary gland invasion and malaria transmission. PLOS Pathog. 18, e1010779 (2022). 9. Del Rosario, M. et al. Apicomplexan F-actin is required for efficient nuclear entry during host cell invasion. EMBO Rep. 20, e48896 (2019). 10. Tosetti, N., Dos Santos Pacheco, N., Soldati -Favre, D. & Jacot, D. Three F-actin assembly centers regulate organelle inheritance, cell -cell communication and motility in Toxoplasma gondii. eLife 8, e42669 (2019). 11. Martinez, M. et al. Origin and arrangement of actin filaments for glid- ing motility in apicomplexan parasites revealed by cryo -electron to- mography. Nat. Commun. 14, 4800 (2023). 12. Bane, K. S. et al. The Actin Filament-Binding Protein Coronin Regu- lates Motility in Plasmodium Sporozoites. PLOS Pathog. 12, e1005710 (2016). 13. Douglas, R. G. et al. Inter-subunit interactions drive divergent dynam- ics in mammalian and Plasmodium actin filaments. PLOS Biol. 16, e2005345 (2018). 14. De Niz, M. et al. Progress in imaging methods: insights gained into Plasmodium biology. Nat. Rev. Microbiol. 15, 37–54 (2017). 15. Vahokoski, J. et al. High-resolution structures of malaria parasite ac- tomyosin and actin filaments. PLOS Pathog. 18, e1010408 (2022). 16. Harding, C. R. et al. Alveolar proteins stabilize cortical microtubules in Toxoplasma gondii. Nat. Commun. 10, 401 (2019). 17. Schaffer, M. et al. Cryo-focused Ion Beam Sample Preparation for Im- aging Vitreous Cells by Cryo -electron Tomography. Bio-Protoc. 5, e1575 (2015). 18. Mastronarde, D. N. Dual -Axis Tomography: An Approach with Align- ment Methods That Preserve Resolution. J. Struct. Biol. 120, 343–352 (1997). 19. Kremer, J. R., Mastronarde, D. N. & McIntosh, J. R. Computer visualiza- tion of three-dimensional image data using IMOD. J. Struct. Biol. 116, 71–76 (1996). 20. Heymann, J. B. Bsoft: Image and Molecular Processing in Electron Mi- croscopy. J. Struct. Biol. 133, 156–169 (2001). 21. Heumann, J. M., Hoenger, A. & Mastronarde, D. N. Clustering and vari- ance maps for cryo-electron tomography using wedge-masked differ- ences. J. Struct. Biol. 175, 288–299 (2011). 22. Ferreira, J. L. et al. Variable microtubule architecture in the malaria parasite. Nat. Commun. 14, 1216 (2023). 23. Cragnolini, T. et al. TEMPy2: a Python library with improved 3D elec- tron microscopy density -fitting and validation workflows. Acta Crys- tallogr. Sect. Struct. Biol. 77, 41–47 (2021). 24. Virtanen, P. et al. SciPy 1.0: fundamental algorithms for scientific com- puting in Python. Nat. Methods 17, 261–272 (2020). 25. Buitinck, L. et al. API design for machine learning software: experi- ences from the scikit -learn project. Preprint at https://doi.org/10.48550/arXiv.1309.0238 (2013). 26. Hunter, J. D. Matplotlib: A 2D Graphics Environment. Comput. Sci. Eng. 9, 90–95 (2007). 27. Harris, C. R. et al. Array programming with NumPy. Nature 585, 357– 362 (2020).

We thank Lindsay Baker for helpful discussions and Carolyn Moores for her continued support and critical reading of the manuscript. Thank you to the CSSB EM facility team for their support. We gratefully acknowledge funding by HFSP long-term postdoctoral fellowship LT000024/2020-L (JLF), Infrastructures for the control of vector -borne diseases (Infravec2) funded by the EU’s Horizon 2 020 programme (grant agreement No 731060) (JLF), Wellcome Career Development award 227774/Z/23/Z (JLF), LOEWE Centre DRUID (“Novel Drugs Targets against Poverty-related and Neglected Tropical Infec- tious Diseases”) within the Hessian Excellence Program (RGD). For the purpose of open access, the author has applied a Crea- tive Commons Attribution (CC BY) licence to any Author Ac- cepted Manuscript version arising. Biorxiv template from: www.github.com/chrelli/bioRxiv-word-template Author contributions VP and JLF designed the study and experiments. JLF generated samples and acquired data. JLF and VP processed and ana- lysed data. VP, JLF, RGD, DV, KG performed critical analysis of findings. JLF, VP, RGD wrote the manuscript. VP generated fig- ures. Data availability All data are available on request. Subvolume average was de- posited on the EMDB. Code availability Scripts are available on request. Competing interests The authors declare no competing interests. .CC-BY 4.0 International licensemade available under a (which was not certified by peer review) is the author/funder, who has granted bioRxiv a license to display the preprint in perpetuity. It is The copyright holder for this preprintthis version posted April 22, 2024. ; https://doi.org/10.1101/2024.04.22.590301doi: bioRxiv preprint