Fig 1.
PfCen1-4 localize to centriolar plaque and can display liquid-like dynamics.
(A) Schematic of nuclear and cell division during asexual blood stage schizogony (B) Schematic of PfCen1-4 and HsCen2 indicating reported post-translational modifications and EFh domains. (C) Immunofluorescence staining of tubulin and PfCen1-4-GFP in parasite strains. All images are maximum intensity projections (MIP). DNA stained with Hoechst. Scale bars; confocal, 1 μm, STED, 100 nm. (D-E) STED time lapse of centriolar plaque region of parasites expressing PfCen1-Halo labeled with MaP-SiR-Halo dye. DNA stained with SPY505-DNA. Scale bar, 100 nm. Quantification of ratio (black line) between height and width (grey lines) of the PfCen1-Halo signal, as indicated in the small schematic.
Fig 2.
PfCen1, 3 and HsCen2 undergo calcium-dependent and reversible LLPS in vitro.
(A) Widefield image of highly concentrated recombinant centrin solutions to promote high droplet density of PfCen1-6His (200 μM), PfCen2-6His (120 μM), PfCen3-6His (193 μM), PfCen4-6His (120 μM), HsCen2 (200 μM) before, after 2 mM CaCl2 addition, and after 10 mM EDTA addition. Scale bars, 10 μm. Inlays show time lapse images of droplet fusion events. Movie metadata was used to estimate the time after calcium addition (bottom) (B) Turbidity measurements in centrin solutions at 30 μM (and higher concentration where indicated) during calcium addition followed by EDTA. Conditions: 50 mM BisTris (pH 7.1) at 37°C.
Fig 3.
Centrins from highly divergent eukaryotes can undergo phase separation in vitro.
(A) Turbidity measurements of recombinant T. brucei centrin A (TbCenA, 196 aa), C. reinhardtii centrin (CrCen, 169 aa) and S. cerevisiae centrin (Cdc31, 161 aa) at 30 μM (and higher concentrations where indicated) during calcium addition to 2 mM followed by 10 mM EDTA. Conditions: 50 mM BisTris (pH 7.1) at 37°C. (B) Using the IUpred 3.0 prediction we find a correlation between intrinsic disorder region probability (red line) and LLPS in vitro. (C) Widefield images of recombinant centrin solutions of TbCenA (100 μM), ScCdc31 (200 μM), CrCen (100 μM) after 2 mM CaCl2 addition show droplets and surface wetting, which dissolves quickly after addition of 10 mM EDTA. Scale bar, 10 μm.
Fig 4.
PfCen1 and PfCen3 interact through co-condensation.
(A-C) Turbidity measurements of recombinant PfCen1-6His and PfCen3-6His at individually subcritical concentrations either independently or as a mixture during addition of calcium followed by EDTA. Individual centrins were supplemented with BSA to a consistent total protein concentration of >20 μM. (D) Brightfield and fluorescence imaging of preformed 30 μM PfCen3-6His protein droplets in presence of Ca2+ before and 30 s after addition of recombinant PfCen1-GFP-6His at 30 μM. Focus was adjusted between timepoints. Scale bar, 10 μm. All conditions: 50 mM BisTris (pH 7.1) at 37°C.
Fig 5.
PfCen1 displays condensate-like properties in parasites.
(A) Schematic of pFIO plasmids during DiCre-dependent recombination. (B) Immunofluorescence of tubulin and GFP in parasites moderately overexpressing PfCen1-GFP (pFIO). Percentage of early schizonts containing ECCAs (arrows) indicated. (C) as in (B) but for cell strongly overexpressing PfCen1-GFP (pFIO+) (D) Time lapse images of parasites overexpressing PfCen1-GFP from pFIO+ and labeled with microtubule live stain SPY555-Tubulin (E) Normalized mean cytoplasmic PfCen1-GFP fluorescence intensity over time and share of PfCen1-GFP fluorescence signal contained within foci with standard deviation. n > 33. (F) Quantification of timepoint of PfCen1-GFP foci appearance relative to first mitotic spindle formation detected by SPY650-tubulin comparing pFIO/pFIO+; standard error of the mean, two-tailed t-test, n1 = 46, n2 = 45. (G) Proteostat and PfCen1-GFP live cell staining. (H) Schematic of pFIO+ version containing T2A skip peptide and expected translation products. (I) Immunofluorescence of rapamycin-induced and control parasites carrying pFIO+_GFP-T2A-PfCen1 plasmid labelled with anti-PfCen1 and anti-Tubulin antibodies. Arrow heads indicate foci above threshold. All images are MIP. DNA stained with Hoechst. Scale bars; 1 μm. (J) Quantification of cells having non-centrosomal foci (which includes unspecific background dots and ECCAs) in induced (n = 49) and control (n = 53) conditions. Absolute numbers given in columns. (K) Quantification of number of non-centrosomal foci per centriolar plaque.
Fig 6.
1,6-hexanediol treatment dissolves centrin foci in vivo.
(A) Turbidity measurements of recombinant PfCen1 solution at 30 μM in 50 mM BisTris (pH 7.1) at 37°C during addition of calcium followed by 5% 1,6-hexanediol and EDTA. (B) Immunofluorescence of non-overexpressing parasites treated with 5% 1,6-hexanediol or untreated control for 5 min stained with anti-PfCen1 and anti-Tubulin antibodies. Images are MIP. DNA stained with Hoechst. Scale bars; 1 μm. (C) Quantification of spindles in treated (n = 110) and control (n = 85) cells (early schizonts) with or without centrin focus at the spindle pole in percentage. Absolute numbers given in columns.
Fig 7.
PfCen1 phase separation depends on its N-terminus and calcium binding.
(A-B) Turbidity measurements in recombinant protein solutions at 30 μM of PfCen1 mutants lacking the N-terminus or with disabled EFh domains, respectively, during addition of calcium followed by EDTA. Conditions: 50 mM BisTris (pH 7.1) at 37°C. (C-D) Immunofluorescence staining of tubulin and GFP in parasites overexpressing N-terminal deletion and non-Ca2+-binding mutant of PfCen1-GFP from pFIO+. Percentage of cells containing ECCAs (arrows) indicated. All images are MIP. DNA stained with Hoechst. Scale bars; 1 μm. (E) Our findings suggest that centrin accumulation at the centriolar plaque from a pre-mitotic diffuse cytoplasmic pool depends on LLPS, which might be promoted by cellular signaling events like calcium levels and a nucleating factor at the centrosome. Parasite is shown without host red blood cell for clarity.