The conserved plant sterility gene HAP2 functions after attachment of fusogenic membranes in Chlamydomonas and Plasmodium gametes

Abstract

The cellular and molecular mechanisms that underlie species-specific membrane fusion between male and female gametes remain largely unknown. Here, by use of gene discovery methods in the green alga Chlamydomonas, gene disruption in the rodent malaria parasite Plasmodium berghei, and distinctive features of fertilization in both organisms, we report discovery of a mechanism that accounts for a conserved protein required for gamete fusion. A screen for fusion mutants in Chlamydomonas identified a homolog of HAP2, an Arabidopsis sterility gene. Moreover, HAP2 disruption in Plasmodium blocked fertilization and thereby mosquito transmission of malaria. HAP2 localizes at the fusion site of Chlamydomonas minus gametes, yet Chlamydomonas minus and Plasmodium hap2 male gametes retain the ability, using other, species-limited proteins, to form tight prefusion membrane attachments with their respective gamete partners. Membrane dye experiments show that HAP2 is essential for membrane merger. Thus, in two distantly related eukaryotes, species-limited proteins govern access to a conserved protein essential for membrane fusion.

Figure 1.

HAP2 is required for fertilization in Chlamydomonas and is phylogenetically conserved in many eukaryotes. (A) Dark-field images of wild-type plus gametes 10 min and 4 h after mixing with wild-type or 63B10 minus gametes. Clusters of gametes undergoing flagellar adhesion are present in both samples at 10 min. A large aggregate of zygotes is visible in the wild-type mixture at 4 h, whereas the 63B10 minus gametes failed to form zygote aggregates with wild-type plus gametes, and the small clusters persisted. (B) Differential interference contrast (DIC) microscopy images of a quadriflagellated zygote formed from fusion of a wild-type plus gamete with a wild-type minus gamete (left panel) and a wild-type plus gamete undergoing flagellar adhesion with a 63B10 minus gamete, but failing to fuse (right panel). (C) Southern blotting of EcoRI- and HindIII-digested Chlamydomonas genomic DNA from 63B10 and wild-type (B215) strains with a pSI103/PvuII DNA fragment as a probe. Arrowheads indicate the pSI103 plasmid in both samples. (D) Structure of the HAP2 gene and location of the aphVIII gene pSI103 plasmid. (E) Diagnostic PCR on genomic DNA showing the presence of disrupted HAP2 (primers p1–p17) in 63B10 gametes and the absence of intact HAP2 in 63B10 gametes and its reappearance in those 63B10 gametes that were rescued for fusion with the wild-type HAP2 gene (primers p1/p2). (F) Large aggregates of zygotes were present only in mixtures of wild-type plus and minus gametes and wild-type plus and 63B10 minus gametes rescued with the HAP2-HA construct (63B10-C9). (G) Phylogenetic tree based on trees generated using MOLPHY and TREE-PUZZLE illustrating the relationships of HAP2 proteins from several species (see Supplemental Material for accession numbers and methods).

Figure 2.

HAP2 is essential for sexual development and mosquito transmission of P. berghei. (A) Structure of the Plasmodium HAP2 gene and gene replacement construct. Short arrows indicate oligonucleotides used for PCR genotyping. (B) Southern hybridization of EcoRI-digested genomic DNA using the 5′ targeting sequence as a probe. Arrowheads indicate diagnostic 2.8-kb (wild-type) and 5.0-kb (HAP2) bands. (C) Diagnostic PCR with genomic DNA templates and primers GFko1/GFko2 to test for the presence of HAP2, and primers GFint/70 to detect a unique 1-kb product across the integration site. (D) RT–PCR detection of HAP2 transcript in parasite lines and stages (the expected larger product from genomic DNA includes one intron). (E) Representative images of midguts from A. stephensi mosquitoes 10 d after feeding on wild-type and hap2-infected mice (bar, 100 μm) and bar chart showing average numbers of oocysts per gut (error bar indicates SEM; n = 47 wild-type or hap2-exposed mosquitoes from three independent experiments). The overall prevalence of infection was 87% for wild type, and 0% for hap2. (F) Immunofluorescence images of live 20-h Plasmodium cultures immunostained for the female gamete/zygote marker P28. Elongate ookinetes (asterisks) were absent from the hap2 mutant (bar, 10 μm), which possessed only round female gametes. The bar chart shows ookinete conversion rates for wild type, hap2 clone 8, and the hap2 mutant that was complemented with HAP2-GFP. Conversion rate is expressed as the percentage of P28-positive parasites that had progressed to the ookinete stage (error bar indicates SD; n = 3). (G) Western blot analysis with anti-GFP, showing expression of a HAP2-GFP fusion protein (arrow) of the expected motility (120 kDa predicted) only in the complemented parasite line. Nonspecifically recognized bands confirm equal loading.

Figure 3.

HAP2 has a sex-restricted function in both Chlamydomonas and Plasmodium. (A) Chlamydomonas hap2 plus gametes can fuse with wild-type minus gametes. (Top panel) Genotyping by Southern hybridization of one mating partner. The second mating partner was always wild type (not shown). The 5.3 kb Not1 fragment is diagnostic of wild-type HAP2 and the 1.3-kb fragment is diagnostic of the disrupted allele. (Bottom panel) Percentage of the indicated gametes that fused when mixed with wild-type gametes of the opposite sex. (B) In vitro malaria ookinete conversion analysis demonstrates that the Plasmodium hap2 mutant shows productive cross-fertilization with the nek4 sterility mutant, which produces functional males only, and not with cdpk4, which produces functional females only (error bar indicates SD; n = 3).

Figure 4.

HAP2 functions after gamete activation, is present at the surface of the minus mating structure in Chlamydomonas (A–E), and is distributed along the length of the male gamete in Plasmodium (F). (A) 63B10 gametes were activated by incubation with wild-type plus gametes, flagella isolated from wild-type plus gametes, or db-cAMP. The percentage of cells that was activated was determined by measuring cell wall loss. (B, left panel) Fluorescence image of activated plus gametes in a mixture of wild-type plus gametes and 63B10 minus gametes. The arrowheads in the fluorescence (left panel) and DIC (middle panel) images indicate the site of the actin-filled mating structures. The right panel is an electron micrograph of an activated mating structure (fertilization tubule) on a wild-type plus that had been mixed with hap2 minus gametes. The actin filaments within the fertilization tubule originate from the cone-shaped, electron-dense doublet zone (arrowheads) at the base of the process (Detmers et al. 1983). Fringe (f; arrow) is visible on the surface of the fertilization tubule. Bar, 200 nm. (C) Immunoblotting with an anti-HA antibody documents that 63B10 cells rescued with HA-tagged HAP2 expressed HAP2-HA protein only in the gamete phase of their life cycle. (D) Immunoblotting with anti-HA antibody indicates that only a single form of HAP2-HA remained after treatment with 0.01% trypsin for 20 min at room temperature. (E) Anti-HA immunostaining combined with DIC microscopy of HAP2-HA gametes demonstrates that HAP2-HA is localized between the two flagella at the site of the minus mating structure. (F) Anti-GFP immunostaining and DIC microscopy of HAP2-GFP gametes of P. berghei shows HAP2-GFP localized along the length of the male gamete (bar, 2 μm).

Figure 5.

HAP2 functions in the gamete fusion reaction downstream from gamete membrane adhesion. (A) Activated live 63B10 gametes, like activated live wild-type minus gametes, adhered via their mating structures to activated, fixed, fluorescently tagged sag1-1 plus gametes, which are incapable of flagellar adhesion (top panel shows differential interference microscopy; bottom panel shows fluorescence; arrowheads indicate the sag1-1 plus gametes). The percent (±SEM) of sag1-1 gametes forming pairs when mixed with an excess of activated 63B10 or wild-type minus gametes is shown below the figure (average from two independent experiments; n = 150–200 sag1-1 cells examined in each). Similar results were obtained when the agglutinin mutant sag1-2 was used (not shown). Between 0 and 6% pairs were detected in controls in which activated live sag1-1 gametes were mixed with the fixed sag1-1 gametes (not shown). (B) Efficiency of exflagellation, gamete adhesion, and gamete fusion in wild-type, p48/45, and hap2 clones of Plasmodium (error bar indicates SD; n = 3 experiments, each examining 100 gametocytes).

Figure 6.

HAP2 is essential for membrane merger in Chlamydomonas. (A) The plasma membranes of activated plus gametes were labeled with the fluorescent lipid PKH26 (arrowheads), mixed with wild-type minus gametes (top panels) or 63B10 minus gametes (bottom panels), and the live cells were examined by DIC (left top and bottom panels) and epifluorescence microscopy. (B) A pair of wild-type plus and minus gametes just after fusion is shown in transmission electron microscopy at low magnification (left panel; bar, 200 nm) and at higher magnification (right panel; bar, 50 nm). The actin filaments of the fertilization tubule of the plus gamete (cell on the left in both images) have become incorporated into the minus gamete as the two cells merge. (C, left panel) In a mixture of wild-type plus gametes and hap2 minus gametes, the tip of the fertilization tubule on a wild-type plus gamete (cell on the left) is tightly associated with the apex of the minus mating structure (bar, 200 nm; arrowheads show the doublet zone). (Right panel) A higher-magnification view (bar, 50 nm) shows that the membranes of the two mating structures are separated from each other by ∼10 nm.