Plasmodium berghei calcium-dependent protein kinase 3 is required for ookinete gliding motility and mosquito midgut invasion

Abstract

Apicomplexan parasites critically depend on a unique form of gliding motility to colonize their hosts and to invade cells. Gliding requires different stage and species-specific transmembrane adhesins, which interact with an intracellular motor complex shared across parasite stages and species. How gliding is regulated by extracellular factors and intracellular signalling mechanisms is largely unknown, but current evidence suggests an important role for cytosolic calcium as a second messenger. Studying a Plasmodium berghei gene deletion mutant, we here provide evidence that a calcium-dependent protein kinase, CDPK3, has an important function in regulating motility of the ookinete in the mosquito midgut. We show that a cdpk3- parasite clone produces morphologically normal ookinetes, which fail to engage the midgut epithelium, due to a marked reduction in their ability to glide productively, resulting in marked reduction in malaria transmission to the mosquito. The mutant was successfully complemented with an episomally maintained cdpk3 gene, restoring mosquito transmission to wild-type level. cdpk3- ookinetes maintain their full genetic differentiation potential when microinjected into the mosquito haemocoel and cdpk3- sporozoites produced in this way are motile and infectious, suggesting an ookinete-limited essential function for CDPK3.

Fig. 1. Targeted disruption of the cdpk3 gene.

A. Illustration of the cdpk3 gene locus and the disruption construct used. B. Diagnostic PCR showing absence of a cdpk3-specific product and presence of an integration-specific product in cdpk3^– clone 7.1. C. Southern blot analysis of EcoRI/HindIII-digested genomic DNA, showing presence of the tgdhfr/ts resistance gene in clone 7.1 and two bands diagnostic of the intact cdpk3 locus in wild type. The presence of two bands is due to HindIII restriction sites within the cdpk3 gene.

Fig. 2. Functional complementation of the cdpk3^– mutant.

A. Schematic illustration of the complementation plasmid, showing the pbcdpk3 gene fused to an N-terminal double c-myc epitope tag under the control of 2.7 kb upstream sequence and a generic 3′UTR from the pbdhfr/ts gene. B. Western blot analysis of protein extracts from 10⁵ ookinetes of the cdpk3^– clone 7.1 and a CDPK3 myc-complemented parasite population. The blot was first probed with the anti-myc mouse monoclonal antibody 9E10 (Sigma, UK). As a loading control the membrane was then stripped and re-probed with an rabbit polyclonal serum raised against T. gondii CDPK2 (Kieschnick et al., 2001) that recognizes CDPK4 in P. berghei (Billker et al., 2004). C. Oocyst numbers counted 10 days after mosquitoes were allowed to feed on blood containing 800 ookinetes per µl. Averages oocyst numbers and standard deviations from two experiments using independent ookinetes culture are shown. Twenty-five midguts were dissected in each replicate experiment.

Fig. 3

cdpk3^– ookinetes fail to associate with the mosquito midgut epithelium. Representative midgut epithelial sheets are shown that were dissected 2 h after mosquitoes had fed on equal numbers of cultured wild-type (A–C) or cdpk3^– (D–F) ookinetes. Midguts were fixed, permeabilized and immunostained with an antibody against A. stephensi annexin to visualized the epithelium (panels A and D), and against the ookinete surface antigen P28 (panels B and E). All images are stacks of confocal sections. A higher magnification inset in (B) shows an ookinete associated with the midgut epithelium.

Fig. 4. Motility of wild-type and cdpk3^– ookinetes and sporozoites.

A. Wild-type ookinetes purified from in vitro cultures disperse from aggregates after being incubated with insect cells. B. cdpk3^– ookinetes fail to show productive gliding motility. Each sequence of images is representative of at least three independent experiments with ookinetes cultured from different infected mice. C. Quantification of ookinete dispersal. The proportion of dispersed ookinetes was determined after 24 h of culture in the absence and presence of cytochalasin D. Arithmetic means ± standard deviations from three independent experiments are shown. D and E. Motility trails of wild-type (D) and cdpk3^– sporozoites (E) obtained from salivary glands of ookinete-injected mosquitoes. Trails were visualized using a monoclonal antibody directed against the circumsporozoite protein, which labels both the sporozoites (arrows) and the trails of surface proteins shed by gliding sporozoites. The upper row of each panel shows corresponding immunofluorescence and phase contrast images of the same sporozoites. The bottom row of each panel illustrates additional representative circumsporozoite trails.