Figures and data

Comparative analysis of SUN protein functions in common eukaryotic model systems.
SUN proteins bridge the outer (ONM) and inner (INM) membranes of the nuclear envelope (NE) to link the cytoskeleton (i.e. actin, microtubules and its organising centres) to various heterochromatic domains (i.e. centromere, telomeres) and the nuclear lamina (five different functions with a unique colour). Established roles of SUN proteins in common model organisms are depicted by coloured circles. ‘-‘ no functional connection for SUN was found and/or structures are not present. ‘?’ denotes possible roles for SUN proteins as NE connections with such structures have been established in these lineages. LECA: Last Eukayotic Common Ancestor.

Location of SUN1 during male gametogenesis.
A. The upper panel schematic illustrates the process of male gametogenesis. N, genome ploidy. Live cell images show the location of SUN1-GFP (green) at different time points(1-8min) during male gametogenesis. DNA(blue) was stained with Hoechst. White arrows indicate the loop/folds. Representative images of more than 50 cells with more than three biological replicates. Scale bar: 5 μm. B. Serial block face-scanning electron microscopy (SBF-SEM) data slice of a gametocyte highlighting the complex morphology of the nucleus (cyan). Representative of more than 10 cells. Scale bar: 1 μm. C. Two 3D models of gametocyte nuclei showing their contorted and irregular morphology. Representative of more than 10 cells. Scale bar: 1 μm. D. SIM images of SUN1-GFP male gametocytes activated for 8 min and fixed with paraformaldehyde. Arrows indicate the SUN1-GFP signals with high intensity after fixation. Representative image of more than 10 cells from more than two biological replicates. Scale: 1 µm. E. SIM images of SUN1-GFP male gametocytes activated for 8 min and fixed with methanol. Arrows indicate the SUN1-GFP signals with high intensity after fixation. Representative image of more than 10 cells from more than two biological replicates. Scale: 1 µm. F. Expansion microscopy (ExM) images showing location of SUN1 (green) detected with anti-GFP antibody and BB/MTOC stained with NHS ester (grey). Hoechst was used to stain DNA. Scale bar: 5 µm. Inset is the area marked with the red box around the BB/MTOC highlighted by NHS-ester staining. Scale bar: 1LJµm. Representative images of more than 10 cells from two biological replicates. G. ExM images showing location of SUN1 (green) and α-tubulin (magenta) detected with anti-GFP and anti-tubulin antibodies, respectively. Hoechst was used to stain DNA(blue). N=Nucleus; S=Spindle; A=Axoneme. Scale bar: 5 µm. Inset is the area marked with the white box on Fig. 1E middle panel around the BB/MTOC. Scale bar: 1LJµm. Representative images of more than 10 cells from two biological replicates. H. Live cell imaging showing location of SUN1-GFP (green) in relation to the BB and axoneme marker, kinesin-8B-mCherry (magenta) at different time points(1-5min) during gametogenesis. Blue in merged image is DNA stained with Hoechst. Representative images of more than 20 cells from more than three biological replicates. White arrows indicate the loops/folds labelled with SUN1 where BB/axonemes are assembled outside the nuclear membrane. Scale bar: 5LJµm. I. Live cell imaging showing location of SUN1-GFP (green) in relation to the spindle marker, EB1-mCherry (magenta) at different time points during gametogenesis. Blue in merged image is DNA stained with Hoechst. White arrows indicate the loops/folds labelled with SUN1. Representative images of more than 20 cells from more than three biological replicates. Scale bar: 5LJµm. J. Live cell imaging showing location of SUN1-GFP (green) in relation to the kinetochore marker, NDC80-mCherry (magenta) at different time points during gametogenesis. Blue in merged image is DNA stained with Hoechst. White arrows indicate the loops/folds labelled with SUN1. Representative images of more than 20 cells with more than three biological replicates. Scale bar: 5LJµm.

Deletion of sun1 affects male gamete formation and blocks parasite transmission.
A. Exflagellation centres per field at 15 min post-activation. n = 3 independent experiments (>10 fields per experiment). Error bar, ± SEM. B. Percentage ookinete conversion from zygote. n= 3 independent experiments (> 100 cells). Error bar, ± SEM. C. Total number of GFP-positive oocysts per infected mosquito in Δsun1 compared to WT-GFP parasites at 7-, 14-and 21-days post infection. Mean ± SEM. n= 3 independent experiments. D. The diameter of GFP-positive oocysts in Δsun1 compared to WT-GFP parasites at 7-, 14- and 21-days post infection. Mean ± SEM. n= 3 independent experiments. E. Mid guts at 10x and 63x magnification showing oocysts of Δsun1 and WT-GFP lines at 7-, 14- and 21-days post infection. Scale bar: 50 μm in 10x and 20 μm in 63x. F. Total number of midguts sporozoites per infected mosquito in Δsun1 compared to WT-GFP parasites at 14- and 21-days post infection. Mean ± SEM. n= 3 independent experiments. G. Total number of salivary gland sporozoites per infected mosquito in Δsun1 compared to WT-GFP parasites at 21-days post infection. Mean ± SEM. n= 3 independent experiments. H. Bite back experiments showing no transmission of Δsun1, while WT-GFP parasites show successful transmission from mosquito to mouse. Mean ± SEM. n = 3 independent experiments. I. Rescue experiment showing Δsun1 phenotype is due to defect in male sun1 allele. Mean ± SEM. n= 3 independent experiments. J. Representative images of male gametocytes at 8 min post activation stained with DAPI and tubulin (left). Fluorometric analyses of DNA content (N) after DAPI nuclear staining (right). The mean DNA content (and SEM) of >30 nuclei per sample are shown. Values are expressed relative to the average fluorescence intensity of 10 haploid ring-stage parasites from the same slide. K. Representative images of flagellum (male gamete) stained with Hoechst for DNA (left). The presence or absence of Hoechst fluorescence was scored in at least 30 microgametes per replicate. Mean ± SEM. n = 3 independent experiments.

Ultrastructural analysis of Δsun1 gametocytes showing defect in spindle formation and BB segregation
A. Deletion of sun1 blocks first spindle formation as observed by ExM of gametocytes activated for 1.5 min. α -tubulin: magenta, amine groups/NHS-ester reactive: green. Basal Bodies: BB; spindle: S. Insets represent the zoomed area marked by the white boxes shown around BB/MTOC highlighted by NHS-ester and tubulin staining. Scale: 5LJµm.
B. ExM images showing defect in BB/MTOC segregation in Δsun1 gametocytes activated for 15 min. α -tubulin: magenta, amine groups/NHS-ester reactive: green. Basal Bodies: BB; Insets represent the zoomed area shown around BB/MTOC highlighted by NHS-ester and tubulin staining. More than 30 images were analysed in more than three different experiments. Scale bars:LJ5LJµm.
C. The schematic illustrates structures associated with mitosis and axoneme formation showing the first spindle is not formed, and BB are not separated in Δsun1 gametocytes.
D. Electron micrographs of WT-GFP microgametocytes (male) at 8 min (a-c) and 15 min (g, h) plus Δsun1 gametocytes at 8 min (d-f) and 15 min (i-l). Bars represent 1 µm (a, d, i), and 100 nm (b, c, e, f, g, h, insert, j, k, l).
a. Low power magnification of WT-GFP microgametocyte showing the central nucleus (N) with two nuclear poles (NP) with a basal body (BB) adjacent to one. The cytoplasm contains several axonemes (A).
b. Enlargement of enclosed area (box) in a showing the basal body (BB) adjacent to one nuclear pole (NP).
c. Detail showing the close relationship between the nuclear pole (NP) and the basal body (BB). Note the cross-sectioned axonemes (A) showing the 9+2 microtubular arrangement.
d. Low power magnification of Δsun1 cell showing a cluster of electron dense basal structures (enclosed area) in the cytoplasm adjacent to the nucleus (N). A – axonemes.
e. Detail from the cytoplasm (boxed area in d) shows a cluster of 4 basal bodies (arrowheads) and portions of axonemes (A) around a central electron dense structure of nuclear pole (NP) material.
f. Detail from a nucleus showing kinetochores (K) with no attached microtubules.
g. Periphery of a flagellating microgamete showing the flagellum and nucleus protruding from the microgametocyte.
h. Detail of a longitudinal section of a microgamete showing the spiral relationship between the axoneme (A) and nucleus (N). Insert. Cross section of a microgamete showing 9+2 axoneme and adjacent nucleus (N).
i. Section through a microgametocyte with a central nucleus (N) undergoing exflagellation.
j. Enlargement of the enclosed area (box) in i showing one cytoplasmic protrusion containing a single axoneme forming a flagellum (F), while the other has multiple axonemes (A).
k. Longitudinal section through a flagellum (F) with a basal body (B) at the anterior end but note the absence of a nucleus.
l. Cross section showing a cytoplasmic process contain 5 axonemes (A) but no associated nucleus.
E-H. Quantification of Δsun1 phenotypes compared to WT-GFP during male gametogenesis. N = Nucleus; BB = Basal Body; NP = Nuclear pole; A = Axonemes.

Reciprocal co-immunoprecipitation of PbSUN1-GFP and ALLAN-GFP during male gametogony
A. Projection of the first two components of a principal component analysis (PCA) of unique peptides derived from two SUN1-GFP (and WT-GFP) immunoprecipitations with GFP-trap (peptide values: Supplementary Table S2). A subset of proteins is highlighted on the map based on relevant functional categories. B. Similar to panel A, but now for the allantoicase-like protein ALLAN (PBANKA_1304400). C. Selected proteins, their size and corresponding gene ID and representation by the number of peptides in either WT-GFP, PbSUN1-GFP or ALLAN-GFP precipitates.

Location of ALLAN-GFP during male gametogenesis
A. The schematic on the upper panel illustrates the process of male gametogenesis. N, ploidy of nucleus. Live cell images showing the location of ALLAN-GFP (green) at different time points(1-15min) during male gametogenesis. Representative images of more than 50 cells with more than three biological replicates. Scale bar: 5 µm. B. ExM images showing location of ALLAN-GFP (green) detected by anti-GFP antibody compared to nuclear pole (NP)/MTOC and BB stained with NHS ester (grey) in gametocytes activated for 8 min. Scale bar: 5 µm. Representative images of more than 20 cells from two biological replicates. Insets represent the zoomed area shown around NP/MTOC and BB highlighted by NHS-ester. Scale bar: 1 µm. C. ExM images showing the location of ALLAN-GFP (green) compared to spindle and axonemes (magenta) detected by anti-GFP and antI-tubulin staining respectively in gametocytes activated for 8 min. Representative images of more than 20 cells from two biological replicates. Scale: 5 µm. Insets represent the zoomed area shown around spindle /axonemes highlighted by tubulin and GFP staining. Basal Bodies: BB; Spindle: S; Axonemes: A; Nucleus: N. Scale bar: 1 µm. D, E, F. Live cell imaging showing location of ALLAN-GFP (green) in relation to the BB and axoneme marker, kinesin-8B-mCherry (magenta) (D); spindle marker, EB1-mCherry (magenta) (E); and kinetochore marker, NDC80-mCherry (magenta) (F) during first mitotic division (1-3 min) of male gametogenesis. Arrows indicate the focal points of ALLAN-GFP. Representative images of more than 20 cells with three biological replicates. Scale bar: 5LJµm.

Deletion of ALLAN impairs male gametogenesis by blocking BB segregation
A. Exflagellation centres per field at 15 min post-activation in Δallan compared to WT-GFP parasites. n≥3 independent experiments (>10 fields per experiment). Error bar, ± SEM. B. Percentage ookinete conversion from zygote. n≥3 independent experiments (> 100 cells). Error bar, ± SEM. C. Total number of GFP-positive oocysts per infected mosquito in Δallan compared to WT-GFP parasites at 7-, 14-and 21-days post infection. Mean ± SEM. n≥3 independent experiments. D. Total number of sporozoites in oocysts of Δallan compared to WT-GFP parasites at 14- and 21-days post infection. Mean ± SEM. n≥3 independent experiments. E. Bite back experiments reveal successful transmission of Δallan and WT-GFP parasites from mosquito to mouse. Mean ± SEM. n = 3 independent experiments. F. ExM images of gametocytes activated for 8- and 15-min showing MTOC/BB stained with NHS ester (green) and axonemes stained with anti-tubulin antibody (magenta). Axonemes: A; Basal Bodies: BB; Microtubule organising centre: MTOC. Insets represent the zoomed area shown around BB/MTOC highlighted by NHS-ester and tubulin staining. More than 30 images were analysed in more than three different experiments. Scale bars:LJ5LJµm. G. Electron micrographs of WT-GFP microgametocytes at 8 mins (a-c) and 15 mins (d, e) and the Δallan at 8 min (f-h) and 15 min (I, j). Bars represent 1 µm (a, f) and 200 nm in all other images.
a. Low power image of a microgametocyte showing the nucleus (N) with two NP complexes (arrows) consisting of the basal body, NP and attached kinetochores. Axonemes (A) are present in the cytoplasm.
b. Enlargement showing the nuclear pole (NP), associated basal body (BB) and axonemes (A).
c. Detail of the nuclear pole (NP) showing kinetochores (K) attached the spindle microtubules. Note the basal body (BB) adjacent to the nuclear pole (NP).
d. Longitudinal section of a microgamete showing the nucleus (N) closely associated with the axonemes (A).
e. Cross section through a microgamete showing the nucleus (N) and axoneme (A) enclosed plasma membrane.
f. Lower magnification of a microgametocyte showing a nucleus (N) with an adjacent clump of electron dense structures (enclosed area) and axonemes (A) in the cytoplasm.
g. Enlargement of the enclosed area in f showing multiple basal bodies (BB) and unseparated nuclear poles (NP) enclosed by portions of nuclear membrane (NM). N – nucleus.
h. Detail from a nucleus showing several kinetochores (K) with no associated spindle microtubules.
i. Longitudinal section of an exflagellating cytoplasmic process consisting of two axonemes (A) but no nucleus.
j. Cross section through an exflagellating cytoplasmic process showing the presence of multiple axonemes (A) but the absence of any nucleus.
H to K. Quantification of Δallan phenotype compared to WT-GFP during male gametogenesis. N = Nucleus; BB = Basal Body; NP = Nuclear pole; A = Axonemes.

Evolution and Structure of the SUN1-ALLAN interaction
A. Domain analysis shows two proteins with allantoicase domain and two proteins with SUN-domain in P. berghei. The SUN domain and two domains comprising allantoicases are part of the same galactose-binding domain family, with a strikingly similar fold. B. Phylogenetic profiles showing the presence of SUN-, ALLAN- and KASH-domain and lamin proteins in Apicomplexa and a selection of other eukaryotes, including two model species Homo sapiens and Arabidopsis thaliana. C. AlphaFold3-modelled interaction between ALLAN and SUN1 based on separate domains (no full structure could be modelled). The SUN1 C-terminus forms a trimeric complex (pTM:0.37) similar to a trimeric ALLAN complex (grey) with the N-terminus of SUN1 interacting with ALLAN (pTM:0.55). This N-terminal domain is unique to Haemosporida. D. Overview in similar style as Fig. 1 of main interactors for putative localization at the nuclear envelope for ALLAN and SUN1 during male gametogenesis. Structures in grey have not been found to be associated with SUN1.