Maximum likelihood phylogeny of Plasmodium interspersed repeat (pir) protein sequences.

The phylogram shown at centre was generated using IQTREE from an 837-character alignment of 4236 PIR protein sequences taken from 15 Plasmodium or Hepatocystis species genomes, supplemented with 17 SURFIN protein sequences, which are designated as the outgroup. A JTT+F+G model of amino acid substitution was applied with a Gamma shape (alpha) value of 3.16. Phylogenetic distance is indicated by a vertical scale. The phylogeny is subdivided into major clades thata are indicated by green segments and labelled as subfamilies A-W. Edge shading denotes branch robustness; black and red shading indicates edges that are subtended by nodes with bootstrap values >70 and <70 respectively. The phylogeny is surrounded by five tracks, from inside to outside: i) parasite species; ii) orthology (i.e. where a sequence belongs to an orthologous clade after reconciliation analysis (see text), the number of sequences in the clade is given); iii) number of conserved protein cores; iv) percentage of protein sequence comprising amino acid repeats; v) total protein length in amino acids. More intense shading represents higher numbers for orthology, cores and repeats respectively.

Reconciliation of pir and Plasmodium phylogenies

The pir gene tree was reconciled with the known relationships for the 15 species concerned here using TREERECS. The main output, showing the many genetic lineages fitted within the species tree, is shown in green. The common ancestor of all pir genes is shown by a red star. Gene duplications are indicated by red rectangles. Where deletion of a gene is inferred, this is shown with a black branch. For comparison, the gene trees of six specific pir clades are shown at top. The reconciled tree identifies points in species phylogeny that coincide with substantial reduction in pir diversity (a-d), coinciding with the origins of rodent malaria species (Vinckeia subgenus) and Hepatocystis, P. malariae, and the ancestor of P. inuii, P. knowlesi, P. coatneyi and P. fragile respectively. It also identifies moments of substantial expansion in pir diversity (e-g), coinciding, respectively, with diversification of rodent malaria species, origin of primate malaria species (Plasmodium subgenus) and diversification of the clade including P. knowlesi and P. coatneyi. (Inset) Expanded view of the pir gene tree root, showing basal duplication events that created the major pir subfamilies in the ancestor of Vinckeia and Plasmodium subgenera, (most of which were subsequently lost from Vinckeia).

Estimation of selection across the PIRC1 protein structure compared to related paralogs.

a) The protein structure of pirC1, as inferred by Alphafold (AF-K6V314); shading indicates four distinct regions of the sequence alignment (Fig. S1). b) Dn/Ds, the ratio of synonymous to non-synonymous amino acid substitutions, (ω) was estimated using codeML (Yang, 2007) for a codon alignment of pirC1 ortholog sequences taken from 15 species. Individual residues in the protein model are shaded by the ω value of their corresponding codons. Sites showing significant negative selection, as determined by the Bayes empirical Bayes method of codeML (Yang et al, 2005), are shown in dark blue. Regions not included in the sequence alignment are shaded dark grey. c-f) For comparison, selection across four alignments of pirC within-species paralogs, each applied to their Alphafold predicted structures, are shown for P. cynomolgi (AF-K6V3I4), Hepatocystis (AF-A0A653GW70), P. ovale (AF-A0A1D3JEM5) and P. vivax (AF-A5KDQ9) respectively. Sites showing significant positive selection are shown in dark red. g) Dn/Ds (ω) ratio values for pirC1 codon sequences (green line) compared with the four groups of pirC within-species paralogs (black lines).

Growth of murine parasites lacking PIRC1.

a) Replication of Plasmodium chabaudi control transgenic line PcASluc230p (Cunningham et al, 2017) and ΔPCHAS_0101200 parasites in C57BL/6 mice (n=8). Parasitaemia was monitored at 2-day intervals by enumeration of parasites on Giemsa-stained thin blood films. Note that the mutant displayed a longer patency and a lower maximal parasitemia; **p<0.01 (Mann-Whitney). b) Giemsa-stained thin smears of P. chabaudi-infected erythrocytes expressing pirC1 (top) or lacking pirC1 (bottom). See Fig. S13 for additional images. Bar indicates 5 μm. c) Live imaging of the localization of P. berghei PIRC1-mCherry in an erythrocyte infected with a trophozoite-stage parasite. The cells were stained with TER to visualize the erythrocyte membrane and Hoechst 33342 to visualize the parasites DNA – the lack of overlap of mCherry and TER indicates that little to no PIRC1-mCherry is exported. For images of additional intraerythrocytic stages, see Fig. S14. Bar equals 1 μm. d) Localization of PIRC1-mCherry P. berghei-infected hepatocytes. The parasites were visualized by staining with anti-AMA1 antibodies and PIRC1-mCherry was visualized using anti-mCherry antibodies. Host and parasite DNA was visualized by staining with Hoechst 33342.

Growth of Plasmodium knowlesi pirC1 mutant.

a) Replication of P. knowlesi wildtype parasites and parasites lacking PIRC1. Synchronized parasites were treated with DMSO or 100 nM rapamycin during cycle 1 and the parasitemia was determined in the subsequent cycles by staining an aliquot of the culture with SYBR Green followed by cytometry. b) Invasion of P. knowlesi into erythrocytes labeled with CellTrace. Synchronized parasites were treated with DMSO or rapamycin on day 1 and CellTrace-labeled erythrocytes were added on day 2 such that approximately half the erythrocytes were labeled. Parasitemia was determined with SYBR Green staining and cytometry in all cycles. Note the appearance of SYBR Green-positive CellTrace-labeled cells (upper right-hand quadrant) in both the wildtype and mutant culture on day 3. Note also the lack of increase in the parasitemia of the rapamycin-treated parasites. c) DNA content of parasites treated with DMSO or rapamycin in the presence or absence of ML10. Synchronized parasites were treated with DMSO or rapamycin on day 1 and ML10 or DMSO was added to the cultures on day 2. DNA content was determined using SYBR Green and cytometry on day 2 and day 3. Note the accumulation of parasites with a high DNA content in the ML10-treated cultures of the wildtype and mutant parasites.

Phenotypic analysis of P. knowlesi pirC1 mutant and localization of PIRC1.

a) Giemsa-stained thin smears of erythrocytes infected with P. knowlesi parasites expressing (DMSO)) or lacking pirC1 (Rapa). b) Development of P. knowlesi parasites with or lacking PirC1 after invasion. Synchronized parasites were treated with DMSO or rapamycin and synchronized again and blocked with ML10. Invasion of parasites was synchronized by removing ML10 and thin-film blood smears were produced at the indicated times. See Fig. S15 for additional images. c) Size of wildtype P. knowlesi parasites and parasites lacking PIRC1. Parasites were treated with DMSO or rapamycin and synchronized prior to passage to the subsequent intraerythrocytic cycle. Thin-film blood smears were stained with Giemsa and parasite size was determined by microscopy; ****p<0.001. d) Comparison of DNA content of the parasites producing or lacking PIRC1 shown in Fig. 5C. e) Localization of P. knowlesi PirC1 during the asexual intraerythrocytic cycle. PIRC1 was visualized using an anti-SPOT antibody and the PVM was visualized using an anti-EXP2 antiserum. Parasite DNA was visualized by staining with Hoechst.