An inner membrane complex protein IMC1g in Plasmodium berghei is involved in asexual stage schizogony and parasite transmission

Abstract

The inner membrane complex (IMC), a double-membrane organelle underneath the plasma membrane in apicomplexan parasites, plays a significant role in motility and invasion and confers shape to the cell. We characterized the function of PbIMC1g, a component of the IMC1 family member in Plasmodium berghei. PbIMC1g is recruited to the IMC in late schizonts, activated gametocytes, and ookinetes. Pairwise yeast two-hybrid assays demonstrate that PbIMC1g interacts with IMC1c, a component of the PHIL1 complex, and the core sub-repeat motif "EKI(V)V(I)EVP" in PbIMC1g is essential for this interaction. Localization of PbIMC1g to the IMC was dependent on its IMCp domain, while its C-terminus and palmitoylation sites were required for the full efficiency of proper IMC targeting. PbIMC1g is required for asexual stage development, and its conditional knockdown resulted in a defect in schizogony. Additionally, PbIMC1g was also important for male gametogenesis and ookinete development. As an IMC component that assists in anchoring the glideosome to the subpellicular network, PbIMC1g was also involved in ookinete motility and mosquito midgut invasion. IMC1g from the human parasite Plasmodium vivax could functionally replace PbIMC1g in P. berghei, confirming the evolutionary conservation of IMC1g proteins in Plasmodium spp. Together, this work reveals an essential role of IMC1g in the parasite life cycle and suggests that IMC1 family members likely contribute to parasite gliding and invasion.

Importance: The malaria parasite's inner membrane complex is critical to maintain its structural integrity and motility. Here, we identified the function of the IMC1g protein, a member of the IMC1 family, in invasive and proliferative stages of P. berghei. We found that the IMCp domain of PbIMC1g is critical for proper IMC targeting, and PbIMC1g interacts with PbIMC1c. Conditional knockdown of PbIMC1g expression affects schizogony, gametogenesis, and ookinete conversion. PbIMC1g interacts with IMC1c to firmly anchor the glideosome to the subpellicular network. Additionally, we confirmed that IMC1g is functionally conserved in Plasmodium spp. These data reveal the function of IMC1g protein in anchoring the glideosome, providing further insight into the mechanism of the glideosome function.

Keywords: IMC1; Plasmodium berghei; gliding motility; inner membrane complex; malaria.

Fig 1

Expression and localization of PbIMC1g. (A) Expression of the PbIMC1g-HA protein in schizonts (SZ), gametocytes (GC), activated gametocytes (GCa), and ookinetes (Ook). The arrow indicates PbIMC1g-HA protein (~37.8 kDa) detected using anti-HA mAb. Proteins extracted from uninfected RBCs (EC) were used as a negative control. β-Actin was used as a loading control. (B) The relative PbIMC1g-HA/β-actin signal intensity ratios calculated with the Image J software are shown. ***, P < 0.001. (C) The subcellular localization of PbIMC1g-HA protein during parasite development. Troph, trophozoite; SZ, schizont; MG, male gametocyte; MGa, activated male gametocyte; Ex, exflagellated MG; FG, female gametocyte; FGa, activated female gametocyte; Zyg, zygote; Ook, ookinete. Differential interference contrast (DIC), the DAPI-stained nuclei (blue), PbIMC1g-HA (green), and co-localization markers (red) of the parasites are shown. TubII, α-tubulin II antibody (male-gametocyte/gamete marker); CDPK1, anti-CDPK1 sera (ring/trophozoite stage cytosolic marker); G377, α-Pbg377 sera (female-gametocyte/gamete marker); PSOP25, anti-PSOP25 sera (ookinete surface marker). Scale bar, 5  µm. (D) Co-localization of PbIMC1g-HA with glideosome components (GAP45 and GAP50) in schizonts (SZ), merozoites (MZ), and ookinetes (Ook). Nuclei were stained with DAPI. Numbers indicate Pearson’s correlation coefficient values (R). Scale bars, 5  µm. (E) Western blot analysis of freeze-thaw (FZ), 1% Triton X-100 detergent (Tx), and SDS fractionations shows that PbIMC1g is primarily associated with the cytoskeleton. PbIMC1g-HA band was visualized using mouse anti-HA mAb. GAPDH (cytosolic protein) and GAPM2 (integral membrane protein) were used as FZ and Tx fraction loading controls. The black arrowhead indicates PbIMC1g-HA protein.

Fig 2

The core sub-repeat motif of PbIMC1g binds to the PbIMC1c protein. (A) Schematic model of IMC1g and the main components of PHIL1 complex in IMC. (B) Tenfold dilutions of yeast two-hybrid reporter strains containing the indicated bait (BD) and prey (AD) plasmids. Plating onto a synthetic complete medium lacking leucine and tryptophan (–Leu –Trp) selects for the bait and prey plasmids, and plating onto a medium lacking leucine, tryptophan, histidine, and adenine (–Leu –Trp –His –Ade) additionally selects for physical interaction of the bait and prey fusion proteins. Lam, Gal4 BD fused with lamin. T, the SV40 large T antigen. (C) Diagram of the ΔIMCp and ΔC truncations of PbIMC1g and PbIMC1c used for pairwise Y2H assays. Residues included in each fragment are indicated. (D) Diagram of the four fragments within IMCp and core sub-repeat motif “EKI(V)V(I)EVP” mutated version of PbIMC1g that were individually generated for PbIMC1c binding assay. The core sub-repeat motif within the IMCp domain of PbIMC1g, including the “EKIVEVP” (58–64 aa), “EKIIEVP” (106–112 aa), and “EKVVEVP” (118–124 aa), was mutated to alanine simultaneously, to generate PbIMC1g-mFL construct. IMCp-1, 29–79 aa; IMCp-2, 80–130 aa; IMCp-3, 131–184 aa; IMCp-1+2, 29–130 aa. (E) Schematic model of the interaction region between PbIMC1g and PbIMC1c. The core sub-repeat motifs responsible for PbIMC1g and PbIMC1c interaction were marked in red.

Fig 3

The IMCp domain contributed to the IMC targeting of PbIMC1g. (A) Schematic drawing of mcherry-tagged full-length and deletion construct of PbIMC1g protein. (B) Fluorescent microscopy of mCherry-tagged PbIMC1g (red) and MSP1 (green) in schizonts of the FL, ΔC, and ΔIMCp parasites. The nucleus (nuc) was stained with DAPI (blue). Scale bar, 5 µm. Plot profiles of signal intensities evaluated by Image J software are shown on the right side of each immunofluorescence panel. The mis-localized signals were marked with arrows. (C) Solubility assay detecting membrane association of FL (top), ΔC (middle), and ΔIMCp (bottom) using different detergents. Cytosolic soluble proteins are in freeze-thaw (F), integral membrane proteins in Triton X-100 buffer (T), and peripheral membrane proteins in 2% SDS buffer (S). GAPDH, GAPM2, and β-actin were used as loading controls for F-, T-, and S-fractions, respectively. FL-mcherry, 77.9 kDa; ΔC-mCherry, 65.0 kDa; ΔIMCp-mCherry, 59.5 kDa. (D) Click chemistry method detecting palmitoylation of PbIMC1g in PbIMC1g^HA parasites in the schizont and ookinete stages. Structure and residues for potential palmitoylated cysteine residues of PbIMC1g were shown in the upper PbIMC1g schematic. Localization of PbIMC1g-HA protein on schizont and ookinetes is shown in the lower panel. The alkynyl palmitic acid (Alk-C16) metabolically labeled parasites were stained with Alexa Fluor 488-conjugated streptavidin (Strepv, green), parasite plasma membrane marker MSP1 (cyan) or P25 (cyan), and anti-HA mAb (α-HA, red). Merged images for Strepv, PPM marker, and α-HA are shown in the right column. Scale bar, 5  µm. (E) Immunoblot assay of palmitoylated PbIMC1g proteins in ookinetes using CuAAC-click chemistry. The captured palmitoylated proteins labeled with (+) or without (−) Alk-C16 were analyzed by western blotting using anti-HA mAb. S, IP supernatant; P, IP elution. The arrowhead indicates PbIMC1g-HA protein. (F) Indirect immunofluorescence microscopy of the PbIMC1g^HA parasites at schizont and ookinete stages after treatment with 100 μM 2‐BP. Scale bars, 5 µm. Arrows indicate mis-localized signals of PbIMC1g-HA protein. (G) Plot profiles of signal intensities evaluated by Image J software at merozoite and ookinete stages are shown. The mis-localized signals were marked with arrows. (H) Pictorial representations of fusion protein constructs used in the experiments are shown. The mutated palmitoylation sites are labeled in the schematics of the Palm-N and Palm-C proteins, respectively. (I) Fluorescent microscopy of mCherry-tagged PbIMC1g (red) and MSP1 (green) or P25 (green) at the merozoite and ookinete stages of the Palm-mutN and Palm-mutC parasites. Nuclei (nuc) were stained with DAPI (blue). Scale bar, 5 µm. Plot profiles of signal intensities evaluated by Image J software are shown on the right side of each merozoite stage transgenic strain’s immunofluorescence panel. The mis-localized signals were marked with arrows. (J) Solubility assay detects membrane association of Palm-mutN (mutN) and Palm-mutC (mutC) using different detergents. F, freeze-thaw extraction; T, Triton X-100 extraction; S, SDS extraction. GAPDH and GAPM2 were used as loading controls for F- and S-extractions, respectively.

Fig 4

Editing of PbIMC1g to generate a C2/10/11A mutant strain (Nmut). (A) Schematic illustrating PbIMC1g with putative N-terminal palmitoylation sites that have been predicted by CSS-Palm 4.0 software. C2, C10, and C11 are shown in red-colored letters. (B) Illustration depicting the strategy to mutate C2, C10, and C11 of PbIMC1g to alanine. The location of PCR primers used for genotyping is indicated in the figure. (C) A PCR product encompassing the modified locus was sequenced. An image of the chromatogram reveals successful editing of the locus of interest that resulted in the C2/10/11-to-A mutation. (D) Comparison of the proliferation curves of wild-type Pb. ANKA (WT) and Nmut parasites. Parasitemia in Nmut and WT parasite-infected mice was determined daily by light microscopy of Giemsa-stained blood smears. Data are the mean ± SD of three independent experiments (n = 10 mice/group). Statistical significance between WT and Nmut parasites was determined by two-tailed Student’s t-test. (E) The effects of N-terminal palmitoylation site mutation on the survival of infected mice. Kaplan-Meier survival curves of mice infected with the WT and Nmut parasites. Data are the mean ± SD of three independent experiments (n = 10 mice/group). (F) Gametocytemia of the WT and Nmut parasites at 3 days post-infection (dpi). (G) Male/female ratios of the WT and Nmut parasites at 3 dpi. (H) Stage distribution of WT and Nmut gametocyte at 3 dpi. (I) The number of exflagellation centers per 10 fields in WT and Nmut parasites at 3 dpi. (J) Ookinete numbers per well at 24 h post in vitro culture assay. (K) Cross-fertilization of Nmut line with Δp47 and Δp48/45. All bar graphs in this figure show mean ± SD from three biological replicates. *, P < 0.05; ***, P < 0.001.

Fig 5

PbIMC1g is required for asexual stage cytokinesis. (A) Schematic diagram outlining the methodology used for the in vivo assay of PbIMC1g depletion on merozoite invasion. (B) Parasitemia of PbIMC1g KD parasite-infected mice at 4 hpi. [+] TMP, PbIMC1g^cKD schizont injected directly after purification; [−] TMP, purified PbIMC1g^cKD schizont injected after 2-h incubation with medium without additional TMP supplement. ns, not significant. (C) The intraerythrocytic development of PbIMC1g-depleting parasites. PbIMC1g^cKD schizonts (10⁸) were purified and intravenously injected into mice. The infected mice were supplied with ([+]) or without ([−]) 1 mg/mL of TMP in drinking water. After 4, 8, 16, and 22 h post-infection, the percentages of rings (grey), trophozoites (pink), and schizonts (blue) were determined. Data represents mean ± SD of two independent experiments in duplicates (Tukey’s multiple comparison tests were applied between [+] and [−] TMP groups at the same time points). (D) Number of merozoites per schizont obtained after in vitro culture of PbIMC1g^cKD [+]/[−] TMP parasites. Data represents mean ± SD of two independent experiments (n = 60). ***, P < 0.001 (Mann–Whitney U test). (E) TEM images of matured schizont stage PbIMC1g^cKD [+]/[−] TMP parasites from in vitro schizont cultures. In both images, the different membranes are indicated as follows: erythrocyte (EM, black arrowhead), parasite vacuole (PVM, black arrowhead), parasite plasma (PPM, white arrow), and inner membrane complex (IMC, white arrow). N, nucleus; R, rhoptry. Representative of two experiments. Experiment 1, n = 20; experiment 2, n = 15. Scale bars, 500 nm (white), 200  nm (black). Schematics of IMC and PPM formation during schizogony of PbIMC1g^cKD [+]/[−] TMP parasites in vitro are shown in the lower panel. (F) Real-time PCR analysis of inner membrane complex located proteins in PbIMC1g^cKD [+]/[−] TMP parasites. (G) IFA of PbIMC1g^cKD schizonts cultured in the presence ([+]) or absence ([−]) of 1 µM TMP. The IFA images were detected using anti-HA, anti-GAPM1, anti-GAPM2, anti-GAPM3, and anti-IMC1c antibodies. Scale bar, 5 µm. (H) Western blot analysis of expression levels of PHIL1 complex in PbIMC1g KD parasites. The immunoblot membrane was probed with anti-HA, anti-GAPM1, anti-GAPM2, anti-GAPM3, and anti-IMC1c antibodies, respectively. +, [+] TMP; −, [−] TMP. β-Actin was used as a loading control. Relative band intensity (normalized to the signal intensity of β-actin) of each protein in [−] TMP compared to [+] TMP that indicates that the reduction of protein expression level was shown in the right panel.

Fig 6

PbIMC1g is required for male gametogenesis. (A) Gametocytemia of the PbIMC1g^cKD parasites in the absence (−TMP) or presence (+TMP) of 1 mg/mL TMP at 3 days post-infection (dpi). (B) Male/female ratios of the PbIMC1g^cKD [+]/[−] TMP parasites at 3 dpi. (C) The percentage of activated PbIMC1g^cKD [+]/[−] TMP microgametocyte at 3 dpi. (D) The proportions of macrogametocytes or microgametocytes forming gametes (%) in PbIMC1g^cKD [+]/[−] TMP parasites at 3 dpi. Gametocytes, Ter119 positive and α-tubulin II/G377 positive; gametes, Ter119 negative and α-tubulin II/G377 positive. (E) Comparison of exflagellation centers formation from PbIMC1g^cKD [+]/[−] TMP parasites. The numbers of exflagellation centers (mean ± SD) were determined from two independent experiments. (F) IFA images of PbIMC1g^cKD [+]/[−] TMP male gametocytes showing the Tubulin II (tub2, red) at different time points after activation (left panel). Scale bars, 5 µm. The quantified percentages of exflagellated PbIMC1g^cKD [+]/[−] TMP parasites after activation are shown in the right panel. (G) TEM micrographs of male gametogenesis of PbIMC1g^cKD [+]/[−] TMP parasites at 8 mpa. N, nucleus; A, axonemes. Scale bars, 500 nm (black), 200  nm (white). Schematics of axonemes of PbIMC1g^cKD [+]/[−] TMP parasites at 8 mpa were shown in the right panel. Each group examined 40 parasites (n = 40) for TEM analysis. Statistical analysis for panels A–F was done using the Student’s t-test. ns, not significant; *, P < 0.05.

Fig 7

Transcriptome analysis for activated gametocyte of PbIMC1g knockdown parasites by RNA-seq. (A) Hierarchical clustering of all genes with significant changes in expression. The normalized FPKM values are shown on the vertical axis and strain information ([+] TMP, PbIMC1g protein with normal expression level; [−] TMP, PbIMC1g knockdown parasites) on the horizontal axis. Clustering is based on Spearman correlation coefficients and plotted using an R program. Duplicates from each experimental group clustered independently (upper dendrogram). Refer to Table S1A for the data sets used to generate this figure. (B) Volcano plot showing the extent and significance of upregulated (red) and downregulated (blue) genes in the [−] TMP parasites compared to [+] TMP (absolute log₂ fold change  > 1). Refer to Table S1B. (C) Gene ontology enrichment analysis of significant twofold regulated genes in [−] TMP parasites compared to [+] TMP parasites. GO terms representing the biological processes (BP), cellular component (CC), and molecular function (MF) are presented in blue, green, and red bars, respectively. Refer to Table S1C. (D) Heatmaps showing differential expression of selected genes in the PbIMC1g^cKD [+]/[−] TMP parasites. Refer to Table S1D. (E) qRT-PCR validation of the expression of cytokinesis-related genes in [−] TMP-treated parasites compared with [+] TMP-treated parasites. The up- and downregulated genes are presented in orange and blue bars, respectively. The PbIMC1g^cKD [+]/[−] TMP gametocytes were purified using 48% Nycodenz and activated at 25℃ for 15 min. Error bars indicate SD from three biological replicates.

Fig 8

Transmission electron microscopy analysis of PbIMC1g^cKD [+]/[−] TMP parasites at the ookinete stage. (A) Ookinete developmental stages for PbIMC1g^cKD [+]/[−] TMP parasites. Ookinetes were identified using P25 and defined as those cells differentiated successfully into elongated “banana-shaped” forms. DNA is stained with Hoechst dye (blue). Scale bar, 5 µm. (B) Ookinete conversion rate (%) in PbIMC1g^cKD [+]/[−] TMP parasites. The conversion rate is the percentage of Pbs21-positive parasites successfully differentiated into elongated “banana-shaped” ookinetes. Statistical analysis was done using the Student’s t-test. ***, P < 0.001. (C) Cells of different morphologies (%). (D) TEM images of PbIMC1g^cKD [+] TMP (a, c, e, g) and [−] TMP (b, d, f, h) parasites at the ookinete stage. (a–d) Longitudinal section of a crescent-shaped [+] TMP ookinete. AC, apical complex; M, micronemes; Cr, crystalline body; P1, polar ring 1; P2, polar ring 2. (e–h) Cross section through the periphery of the anterior complex of a [+] TMP (e, g) and a [−] TMP (f, h) parasite showing similar substructure consisting of the outer parasite plasma membrane and the underlying inner membrane complex. N, nucleus; DG, dense granule; Mt, subpellicular microtubules. Scale bars, 1 µm (black), 200  nm (white). (E) qRT-PCR validation of the expression of kinesin superfamily genes in [−] TMP-treated parasites compared with [+] TMP-treated parasites. Error bars indicate SD from three biological replicates. (F) The distance between each microtubular at the ookinete stage was analyzed using Image J software. N = 36 for both [+] and [−] TMP groups. Statistical analysis was done using the Student’s t-test. ns, not significant.

Fig 9

Effect of PbIMC1g knockdown on ookinete motility and infectivity. (A) Effect of PbIMC1g KD on the average gliding speed of mature ookinetes in Matrigel. The speeds of PbIMC1g^cKD [+]/[−] TMP ookinete were expressed in μm/s. Images were captured every 10 s over a 30-min period. The plot is based on pooled data from two independent experiments and >10 ookinetes analyzed. Statistical analysis was done using the Student’s t-test. ***, P < 0.001. (B) Prevalence of infection of PbIMC1g^cKD [+] TMP parasite-infected mosquitoes. **, P < 0.01 (C) Direct mosquito feeding assay (DFA) in mice infected with the PbIMC1g^cKD [+] TMP parasites followed by sucrose feeding in [+]/[−] TMP conditions at 24 h post-feeding. Data points represent midgut oocyst numbers of individual mosquitoes in each group. Results from three independent experiments are shown. Error bars indicate mean ± SD (n  =  3). Statistical significance was determined using the Mann–Whitney U test. *, P < 0.05; ***, P < 0.001. (D) Real-time PCR analysis of PHIL1 complex and glideosome proteins in PbIMC1g^cKD [+]/[−] TMP parasites. (E) Immunofluorescence analysis of PbIMC1g^cKD [+]/[−] TMP ookinetes. The IFA images were detected using anti-HA, anti-GAPM1, anti-GAPM2, anti-GAPM3, and anti-IMC1c sera, respectively. Scale bar, 5 µm. (F) Immunoblotting of expression levels of PHIL1 complex in PbIMC1g-depleting parasites. The immunoblot membrane was probed with anti-HA, anti-GAPM1, anti-GAPM2, anti-GAPM3, and anti-IMC1c sera, respectively. β-Actin was used as a loading control. Relative band intensity (normalized to the signal intensity of β-actin) of each protein in [−] TMP compared to [+] TMP that indicates the reduction of protein expression level is shown in the right panel.

Fig 10

Phenotypic analysis of the PvIMC1g^TR-transgenic parasites in asexual and sexual stages. (A) Schematic figure of predicted palmitoylated sites on PvIMC1g protein. (B) Click chemistry method detecting palmitoylation of PvIMC1g-Myc protein in PvIMC1g^TR parasites at ookinete stages. The PbIMC1g^HA ookinetes were metabolically labeled with (+) or without (−) 25 µM of Alk-C16 for 4 h, biotinylated by click reaction, followed by precipitation with streptavidin agarose resin. The captured palmitoylated proteins were analyzed by western blotting using anti-Myc mAb. S, supernatant; P, pellet. The arrowhead indicates PvIMC1g-Myc protein. (C) IFA of 100 µM 2-BP-treated PvIMC1g^TR parasites at schizont and ookinete stages. Parasites were co-stained with anti-Myc mAb and anti-MSP1 (schizont) or anti-P25 (ookinetes) sera. The nuclei (nuc) were stained with DAPI. Scale bar, 5 µm. (D) Growth curves of the P. berghei ANKA (wild type [WT]) and transgenic PvIMC1g^TR (TR) parasites. Each mouse was inoculated with 1 × 10⁶ iRBCs, and parasitemia was monitored daily. Data are the mean ± SD of three independent experiments. (E) Kaplan-Meier survival curve of mice infected with the WT and TR parasites. Each group has five mice. The graph is representative of three independent experiments. (F) Gametocytemia of WT and TR transgenic parasites at 3 dpi. (G) Male/female gametocyte ratios of WT and TR parasites at 3 dpi. (H) Exflagellation centers formation in the WT and TR parasites at 3 dpi. (I) Ookinete conversion rates (%) of the WT and TR parasites. Data were representative from three independent experiments for figures (F–I). n = 5 each. (J) Prevalence of infection of WT and TR parasite-infected mosquitoes. (K) Midgut oocysts in mosquitoes infected with WT and TR strain at 10 dpi. Each dot indicates the oocyst number of a mosquito, while the horizontal bars indicate the mean  ±  SD. The graph shows two independent feeding experiments. For each feeding experiment, 40 mosquitoes were dissected for the WT and TR parasites, respectively.

Fig 11

Diagram of the hypothesized mechanism of PbIMC1g in governing ookinete motility. (A) PbIMC1g under normally expressed condition (PbIMC1g^cKD [+] TMP). PbIMC1g anchors the glideosome to the SPN by interacting with PHIL1 complex. The motor complex provides the needed power for ookinete’s gliding and mosquito midgut invasion. (B) PbIMC1g under KD condition (PbIMC1g^cKD [−] TMP). PbIMC1g KD consequently reduces the stability of PHIL1 complex in IMC, which decreases the strength of the motor complex’s anchoring to the SPN, leading to reduced gliding motility of the ookinete.