Genome-wide functional analysis of Plasmodium protein phosphatases reveals key regulators of parasite development and differentiation

Abstract

Reversible protein phosphorylation regulated by kinases and phosphatases controls many cellular processes. Although essential functions for the malaria parasite kinome have been reported, the roles of most protein phosphatases (PPs) during Plasmodium development are unknown. We report a functional analysis of the Plasmodium berghei protein phosphatome, which exhibits high conservation with the P. falciparum phosphatome and comprises 30 predicted PPs with differential and distinct expression patterns during various stages of the life cycle. Gene disruption analysis of P. berghei PPs reveals that half of the genes are likely essential for asexual blood stage development, whereas six are required for sexual development/sporogony in mosquitoes. Phenotypic screening coupled with transcriptome sequencing unveiled morphological changes and altered gene expression in deletion mutants of two N-myristoylated PPs. These findings provide systematic functional analyses of PPs in Plasmodium, identify how phosphatases regulate parasite development and differentiation, and can inform the identification of drug targets for malaria.

Graphical abstract

Figure 1

The Plasmodium Phosphatome Schematic phylogenetic tree and domain architectures for the PPs of P. berghei ANKA and P. falciparum 3D7 showing family and subfamily classification. Proteins encoded in only one species are highlighted. Deletion mutants obtained are shown in bold red text. Domain architecture for P. falciparum protein is shown unless no ortholog exists (PTP2, NIF1). See also Figure S1.

Figure 2

PP Expression and Phenotypic Analysis of 14 PP Mutants (A) Localization of representative PP-GFP classified into three categories: present only in nucleus (Nucleus), diffuse staining (Cytoplasm), or localized to specific cellular domain (Local). Scale bar, 5 μm. Green, GFP; blue, Hoechst; red, Cy3 P28 staining. (B) PP-GFP expression in five key developmental stages. N, nucleus; C, cytoplasm; L, local/heterogeneous. (C) Representation of phenotypic analysis. See also Figures S2 and S3, Table S1, Table S2, and Table S5.

Figure 3

PPM2 and PPM5 Expression, Phosphatase Activity and Phosphorylation, and N-Myristoylation Status (A) Wild-type RNA expression of ppm2 (upper panel) and ppm5 (lower panel). Error bar ± SEM, n = 3. AS, asexual blood stages; Sch, schizonts; NA, nonactivated gametocytes; AG, activated gametocytes; Ook, ookinetes; Spor, sporozoites. (B) Expression of PPM2-GFP (left) and PPM5-GFP (right). Merge is the composite of Hoechst to detect the nuclei, GFP and Cy3 P28 for sexual stages. Scale bar, 5 μm. (C) Anti-GFP western blot of soluble (S), peripheral membrane (PM), and integral membrane (IM) fractions from parasite lysates. (D) (Left) Phosphatase activity in parasite lysate immunoprecipitates. Error bar ± SEM, n = 3. (Right) Anti-GFP western blot from corresponding lysates. (E) In vivo phosphorylation. (Upper panel) [³²P]-phosphorylation of immunoprecipitated GFP-proteins from parasite lysates. (Lower panel) Corresponding western blot. Protein markers are to the left. (F) Parasite lysates labeled with YnMyr (+) and controls without labeling (−) were run directly (− pull-down) or following affinity purification (+) and detected with anti-GFP antibody. Protein markers are shown to the left. See also Figure S4 and Table S6.

Figure 4

Phenotypic and Ultrastructure Analysis of Δppm2 and Δppm5 (A) Exflagellation (left), gametocytaemia (middle), and gametocyte sex allocation (right) of WT-GFP, Δppm2, and Δppm5 parasites. Error bar ± SD; n = 3. (B) Ookinete conversion in WT-GFP, Δppm2, and Δppm5 parasites. Error bar ± SD; n = 3. (C) Average number of oocysts per mosquito gut. Scale bar, arithmetic mean; n = 60. Infection prevalence was 81% for wild-type, 0% for Δppm2, and 88% for Δppm5. Scale bar, 50 μm. (D) Ookinete differentiation. Morphologies used for scoring are given above the graph and are previously described (Janse et al., 1985). Error bar ± SD, n = 3. Scale bar, 5 μm. (E) Ookinete conversion after genetic crossing. Error bar ± SD; n = 3. (F) Genetic complementation. Scale bar, arithmetic mean; n = 60. (G) (Gi) TEM of a longitudinal section through a wild-type crescent shaped ookinete. (Gii) Section through a Δppm2 retort showing the bulbous shape of the parasite but with normal structures in the cytoplasm. Note the very few micronemes. (Giii) Example of a Δppm5 ookinete showing the crescent shape but no micronemes. (Giv) Δppm5 ookinete containing few micronemes. For all panels, A, apical membrane complex; N, nucleus; Cr, crystalline body; M, micronemes. Scale bar, 1 μm. See also Figure S5.

Figure 5

Global Transcriptional Analysis of Δppm2 and Δppm5 by RNA-Seq (A and B) Ratio-intensity scatterplots of normalized FPKM values for each stage of mutant development. Log₂ fold change between wild-type and mutant (y axis) and the average FPKM value (x axis). (C) Log₂ fold change in Δppm2 and Δppm5 at different life cycle stages. Functional groups were inferred from annotations available in GeneDB (http://www.genedb.org/Homepage). Genes are arranged in order of significance (in relation to regulation) in each sample and in the total data set. Full gene list and heatmap order are shown in Table S4. (D and E) qRT-PCR of a variety of genes (based on data from RNA-Seq) in (D) Δppm2 and (E) Δppm5 parasites compared to wild-type controls. Error bar ± SEM, n = 3 biological replicates. Schizonts, Sch; activated gametocytes, AG; ookinetes, Ook. Student’s t test, ^∗p < 0.1, ^∗∗p < 0.05, ^∗∗∗p < 0.001. (F and G) qRT-PCR validation of the RNA-Seq data using Log₂ values. See also Figure S6, Table S3, Table S4, and Table S6.

Figure 6

Summary of PP Function throughout the P. berghei Life Cycle PPs with essential functions in the mosquito are highlighted (blue). Protein kinases essential at similar stages (Tewari et al., 2010) are highlighted in red. See also Table S2.