A cascade of DNA-binding proteins for sexual commitment and development in Plasmodium

Abstract

Commitment to and completion of sexual development are essential for malaria parasites (protists of the genus Plasmodium) to be transmitted through mosquitoes. The molecular mechanism(s) responsible for commitment have been hitherto unknown. Here we show that PbAP2-G, a conserved member of the apicomplexan AP2 (ApiAP2) family of DNA-binding proteins, is essential for the commitment of asexually replicating forms to sexual development in Plasmodium berghei, a malaria parasite of rodents. PbAP2-G was identified from mutations in its encoding gene, PBANKA_143750, which account for the loss of sexual development frequently observed in parasites transmitted artificially by blood passage. Systematic gene deletion of conserved ApiAP2 genes in Plasmodium confirmed the role of PbAP2-G and revealed a second ApiAP2 member (PBANKA_103430, here termed PbAP2-G2) that significantly modulates but does not abolish gametocytogenesis, indicating that a cascade of ApiAP2 proteins are involved in commitment to the production and maturation of gametocytes. The data suggest a mechanism of commitment to gametocytogenesis in Plasmodium consistent with a positive feedback loop involving PbAP2-G that could be exploited to prevent the transmission of this pernicious parasite.

Figure 1. Identification of mutations in PBANKA_143750 that account for the repeated spontaneous loss of commitment to gametocytogenesis

A. Gametocyte production during a year of continuous mechanical passage of P. berghei. Best-fit polynomial trend (thick) lines of gametocytaemia on individual weekly observations (thin lines). B. Open reading frames (ORF, in yellow) of Pbap2-g (PBANKA_143750) and Pbap2-g2 (PBANKA_103430) with point mutations in new GNP lines from panel A and the long-established line 2.33. Predicted DNA Binding Domains (DBD, light blue) and DBD recognition motifs for PBAP2-G upstream of each ORF (brown bars) are indicated. Dark blue arrows show integration sites for selectable marker cassettes as used for genetic complementation of GNPs (COMP-DOWN) or to disrupt the promoter (COMP-UP). Numbering is relative to position 1 of the ORF. C. FACS analyses of male and female gametocyte numbers (red circled areas) expressed as a percentage of the total parasitized cell counts. From left: P. berghei ANKA HP line (which lacks GFP/RFP reporters thus having no fluorescent signal and from which all subsequent lines reported in this study were derived) served as a negative control. Line 820 is the reporter line from which GNP mutants and a targeted KO (using vector PbGEM-072446) were derived. 820REP and GNPm7REP were generated with the COMP-DOWN complementation vector. D. Giemsa-stained gametocytes in GNP line 2.33 (G756) repaired by the COMP-DOWN construct and after a single transmission through mosquitoes. Scale bar is 6 μm. E. Gametocyte quantification from manual counting in Giemsa-stained blood smears of an independently produced pbap2-g deletion mutant before and after complementation with the DS vector and of two independent pbap2-g2 ko mutants. Error bars show standard deviations from 3 replicates. The loss of gametocytes from the KO mutants was significant (p < 0.05). F. Relative growth kinetics of GNPm9, pbap2-g⁻ and pbap2-g2⁻ lines determined by flow cytometry. i) Cloned GNPm9 constitutively expressing CFP (line GNPm9-CFP) was mixed in a 1:1 ratio with wild type (PBANKA HP) producer line constitutively expressing RFP (line WT-RFP). The daily % of the population expressing either RFP (red), CFP (blue) or both (purple) reflecting mixed-multiply infected cells was calculated. ii) Deletion vectors for ap2-g, ap2-g2 or p28 (control gene for neutral growth rate) were transfected in GFP or mCherry expressing lines (blue and red bars, respectively) and the relative abundance of each mutant determined in mixed infections of uncloned parasites. Error bars show standard deviations from 3 biological replicates. The competitive advantage was significant for the ap2-g⁻ (p < 0.01) but not ap2-g2⁻ parasites (two tailed Student’s T-test for change in relative abundance).

Figure 2. Characterisation of the DNA binding specificity, expression and subcellular localisation of PBAP2-G

A. Top. Protein Binding Microarray (PBM) determination of the DNA binding recognition preference of recombinant DNA Binding Domain (DBD) of PBAP2-G. Bottom. Electrophoretic Mobility Shift Assay (EMSA) in which a shift indicates whether AP2-G DBD binds to double stranded DNA containing wild type (W) or mutated (M) motifs (panels a1-d1and a2-d2, respectively) from the upstream regions of Pbap2-g itself, Pbap2-g2, and position −610 of a hypothetical gene spm1 (Sub-Pellicular Microtubule protein-1, PBANKA_081070). B. Expression analysis by RT-PCR of ap2-g in targeted and spontaneous ap2-g mutants and the wild type control line, 820. The 1.15 kb product indicates lack of transcript only in the targeted KO line. RT = reverse transcriptase. Primer positions were as shown in the schematic. See Fig. S7 for ap2-g transgene expression data. (N=3) C. Localisation of pbap2-g mini gene-product to the nucleus of P. berghei female gametocytes. CFP was sandwiched between the N-terminal 300 bp and at the C-terminal 800 bp of pbap2-g including the DBD and expressed from 2 kb of the pbap2-g promoter in line 820. Expression was only detected in the nuclei of female gametocytes (>50 observations in 3 experiments). It is the C terminal segment that determines the nuclear localisation of PBAP2-G (Fig. S8). Scale bar = 6 μm. Cartoon is not to scale.

Figure 3. Pbap2-g acts upstream of gametocyte gene transcription

A. Volcano plot of log₂ fold change in gene expression in schizonts of ap2-g KO1 (whole ORF deletion) vs. WT line 820 against significance of change (−log10 t-test). Red triangles indicate genes upregulated in gametocytes compared to schizonts. Black and yellow shapes are genes detailed in figure 3A and 3C, respectively. B. Reporter gene expression constructs were transfected into the GNPm9 and 820 control clones to confirm gametocyte gene specific promoters. Reporters contained 2 kb of upstream sequence from the indicated genes driving CFP expression with a constitutive 3′ UTR. Bar plots show CFP measured by flow cytometry over three days in the 820 line. Life cycle stages (asexual, male and female) are separated based on GFP or RFP expression. Mean of three measurements (geometric mean CFP fluorescence) +/− SD, *P<0.05, ** P<0.01 ***P<0.001 by 2-tailed t-test. Flow cytometry plots are shown for CFP expression of reporters in 820 (parental) (left) or GNPm9 (right) lines. Plots show GFP (X-axis) vs. RFP (Y axis) expression for all infected red blood cells and CFP expression in magenta. Numbers on each plot represent the percentage of events within each gate that are positive for CFP (see also Figure S8). C. Deletion studies in the pbap2-g promoter provide support for a role of PBAP2-G binding motifs in the positive feedback regulation of pbap2-g expression. Top: DNA constructs containing a selectable marker were integrated into the promoter region of ap2-g in PBANKA 820. The constructs either deleted 207 bp surrounding the two instances of the AP2-G binding motif at the positions indicated (G901 and G902) or did not (G903 and G904). Two sites of selectable marker integration were tested, 2 and 3 kbp upstream of the ORF of ap2-g. Additionally interruption at −1288 upstream of the ORF of ap2-g was shown to disrupt gametocytogenesis (Fig. S4F). Control line G905 was transfected with a reporter construct targeted to the 230p locus and known not to affect gametocytogenesis. Bottom: Gametocytaemia was measured on consecutive days by flow cytometry once the parasitaemia reached >1%. Mean +/− SD shown, * P<0.05 compared to 820 parental (2-tailed t-Test). Data shown are pooled from three days’ observations and representative of three independent experiments.