Actomyosin forces and the energetics of red blood cell invasion by the malaria parasite Plasmodium falciparum

Abstract

All symptoms of malaria disease are associated with the asexual blood stages of development, involving cycles of red blood cell (RBC) invasion and egress by the Plasmodium spp. merozoite. Merozoite invasion is rapid and is actively powered by a parasite actomyosin motor. The current accepted model for actomyosin force generation envisages arrays of parasite myosins, pushing against short actin filaments connected to the external milieu that drive the merozoite forwards into the RBC. In Plasmodium falciparum, the most virulent human malaria species, Myosin A (PfMyoA) is critical for parasite replication. However, the precise function of PfMyoA in invasion, its regulation, the role of other myosins and overall energetics of invasion remain unclear. Here, we developed a conditional mutagenesis strategy combined with live video microscopy to probe PfMyoA function and that of the auxiliary motor PfMyoB in invasion. By imaging conditional mutants with increasing defects in force production, based on disruption to a key PfMyoA phospho-regulation site, the absence of the PfMyoA essential light chain, or complete motor absence, we define three distinct stages of incomplete RBC invasion. These three defects reveal three energetic barriers to successful entry: RBC deformation (pre-entry), mid-invasion initiation, and completion of internalisation, each requiring an active parasite motor. In defining distinct energetic barriers to invasion, these data illuminate the mechanical challenges faced in this remarkable process of protozoan parasitism, highlighting distinct myosin functions and identifying potential targets for preventing malaria pathogenesis.

Fig 1. Development of a conditional knockout and complementation system for Plasmodium myosins.

A Schematic of the modified p230p locus. The p230p-BIP-sfGFP repair template introduces sfgfp under a constitutive BIP promoter. The BIP promoter is exchanged for the Pfmyoa promoter, forming p230p-prMA-sfGFP, with sfgfp exchanged for 3xHA-tagged Pfmyoa forming PfMyoA-comp or PfMyoA-K764E. B Myosins produce force during a powerstroke, where conformational changes from ATP hydrolysis are communication by a relay (yellow) and SH1 (red) helices to the converter, to swing the lever arm. The SH1 helix in PfMyoA is unusually immobile and additional interactions are needed to stabilise the rigor-like state. Stabilising residues include those between phospho-S19 in the N-terminal extension (NTE, brown) and K764 in the converter (green). PPS structure from [21]; Rigor-like structure (PDB: 6I7D, neck region and light chain outlines added as schematic, by extension of the last helix of the converter). C Genotyping PCR of PfMyoA-comp and PfMyoA-K764E lines confirmed that the WT p230p locus (green half arrow) is completely lost in PfMyoA-comp, while the integrated locus (IN, grey half arrow) is present. D PfMyoA-comp parasites grow at the same rate as the parental PfMyoA-cKO line, while PfMyoA-K764E parasites grow slightly slower over 96 h. Lines show mean parasitaemia, N = 3, each experiment in triplicate.

Fig 2. Conditional complementation and mutagenesis of PfMyoA.

A RAP treatment of PfMyoA-comp and PfMyoA-K764E show that the complementing line has no growth defect, while the RAP-treated K764E line grows at around 55% of DMSO-treated control per cycle under static conditions. B Genotyping PCR of the endogenous Pfmyoa locus in PfMyoA-comp and PfMyoA-K764E after RAP treatment shows that the replacement of the unexcised, integrated allele (IN) with the excised allele (EX) is almost complete. C Western blot of WT, PfMyoA-cKO, PfMyoA-comp and PfMyoA-K764E schizonts confirms that PfMyoA-FLAG, expressed from the endogenous Pfmyoa locus, is almost completely lost in favour of truncated PfMyoA-GFP. PfMyoA-3xHA, expressed at the ectopic locus, is unaffected by RAP treatment. D Parasites were treated with DMSO or RAP and cultured under static or suspension conditions. Parasitaemia was then measured in the following cycle by flow cytometry and normalised to DMSO control for each line and condition, showing that suspension conditions partially alleviate the growth defect caused by K764E mutation, from 67% of DMSO-treated to 77%. Bars show mean parasitaemia, N = 4 (or 2 for PfELC-cKO, tested separately), each experiment in triplicate. Significance assessed by paired t test, two tailed.

Fig 3. Disruption of PfMyoB produces a mild parasite growth defect.

A Schematic showing generation of a PfMyoB-cKO targeting construct. A region of Pfmyob encoding the C-terminal 204 residues was synthesised with re-optimised codons (rcz) and a 3xHA tag, and is placed between two loxPint modules, with sfgfp out-of-frame downstream. Guide RNA sites (scissors) and homology regions were chosen to start as close as possible to the start and end of the modified region. B A structural model of PfMyoB indicating the region excised in PfMyoB-cKO (in black). C Genotyping PCR confirms that transfectants contain only the integrated locus (IN, purple half arrow), while the WT locus (blue half arrow) is completely lost. D Growth of PfMyoB-cKO parasites over 96 h is no different to the parental, DiCre-expressing, B11 line. Line shows mean parasitaemia, N = 3, each experiment in triplicate. E Western blot analysis of WT, PfMyoB-cKO or Cas9-3xHA-expressing controls (where Cas9 was the only 3xHA-tagged protein or PfMyoA-3xHA was also expressed). In all lanes with PfMyoB-cKO or Cas9-3xHA-expressing controls a band around the expected size of Cas9-3xHA (168 kDa) and a presumed Cas9-3xHA breakdown product (~95 kDa) is observed. In PfMyoB-cKO+DMSO, but not +RAP, a slightly larger band is detected around the expected size for PfMyoB-3xHA (97 kDa), confirming that PfMyoB-3xHA is properly expressed and lost after RAP treatment. The PfMyoB-3xHA band runs at a similar size to PfMyoA-3xHA control (96 kDa). F Genotyping PCR shows the loss of much of the integrated, unexcised locus (IN, purple half arrow) after RAP treatment and detection of the excised locus (EX, green half arrow). G Measuring the parasitaemia of PfMyoB-cKO parasites in each of the three cycles following RAP treatment shows a small, steady growth defect, of 93% on average. Lines show mean parasitaemia, normalised to DMSO for each line/cycle. N = 3, each experiment in triplicate.

Fig 4. Video microscopy of merozoite invasion.

A Schematic of merozoite invasion. Invasion comprises attachment to the RBC, deformation of the RBC membrane, internalisation of the merozoite and RBC echinocytosis. Invasion attempts are classified as successful (Type A) or by the phase of failure (Type B-D). B Each event was assigned a score based on the intensity of RBC deformation, from 0 (no deformation) to 3 (severe deformation). C Event types in videos from each line after DMSO treatment. While the distribution of events is not significantly different in PfMyoA or PfELC-cKO lines, PfMyoB-cKO shows significantly more successful invasion. Videos pooled from two independent experiments, each in duplicate. Numbers indicate total videos, each video corresponds to one merozoite invading one RBC (or the first merozoite if multiple invaded the same RBC). Significance assessed by chi-square test, either each PfMyoA/PfELC line separately, or PfMyoB-cKO vs others pooled. D Deformation scores from all DMSO-treated lines separated by successful invasion (INV) or any type of failure (FAIL) reveal a significant increase in strong deformation in failed events. Numbers indicate total videos. Significance assessed by chi-square test. E Examples of each event type, with the numbers indicating the phase of invasion shown in that image (numbers from A). (For Type B event, see S3 Video; for Type D event, see S1 Video). F For Type A and Type B events, the start and end of deformation, internalisation and echinocytosis were timed, leading to the five intervals shown. The median times from Type A events across all lines after DMSO treatment is shown [interquartile range in square brackets], with overall agreement to published values ([8] in brackets).

Fig 5. In the absence of PfMyoA or PfELC, merozoites cannot strongly deform or internalise.

A Comparison of event types from PfMyoA-cKO, PfMyoA-comp and PfELC-cKO lines after DMSO and RAP treatment. PfMyoA-cKO parasites show neither deformation nor internalisation after RAP treatment (p<0.0001, significance assessed by Fisher’s exact test comparing pooled failures to Type A events). In contrast, PfMyoA-comp shows only a slight drop in the rate of successful invasion after RAP treatment (p = 0.047). PfELC-cKO parasites also shows a complete loss of successful invasion, but almost half of the events did involve deformation. Significance assessed by chi-square test. B The distribution of deformation scores does not change significantly for PfMyoA-comp events after RAP treatment. The mean deformation score for PfELC-cKO merozoites is greatly reduced by RAP treatment. Significance assessed by chi-square test. C Comparing the duration of each phase of invasion for PfMyoA-comp parasites after DMSO or RAP treatment shows no significant differences. D Examples of RAP-treated PfMyoA-cKO merozoite, showing attachment only, with no further progress (S1 Video), and a PfELC-cKO merozoite undergoing a Type C failure, showing deformation but no internalisation (S2 Video). Time indicated in seconds, scale bar 2 μm. E Schematic based on Fig 4A showing invasion attempts by PfMyoA-cKO merozoites arrest after attachment, while invasion attempts by PfELC-cKO merozoites arrest after deformation, though many arrest before deformation as well.

Fig 6. PfMyoA drives invasion pore closure, while PfMyoA and PfMyoB both help initiation of internalisation.

A RAP treated PfMyoA-K764E parasites showed a significantly shifted distribution of event types, with more Type B and Type C events. This is consistent with these parasites having insufficient motor function to overcome a third barrier, at completion of internalisation. Significance assessed by chi-square test. PfMyoB-cKO parasites showed a significant increase in Type C failures, consistent with an impairment at initiation of internalisation. Significance assessed by Fisher’s exact test, comparing successful invasion to pooled invasion failures. B PfMyoA-K764E parasites show a slight weakening of deformation, though not significant. There is no difference between the deformation scores in PfMyoB-cKO parasites after RAP treatment. Significance assessed by chi-square test. C For PfMyoA-K764E parasites undergoing Type A events (black bars and data points) RAP treatment induces a longer pause pre-internalisation. Only in Type B events (cyan bars) after RAP treatment is internalisation significantly slower and the pause post-internalisation shorter. Bars show median and interquartile range, or median only for Type B events. Significance assessed between Type A DMSO and RAP treatments by Mann-Whitney test, shown when p<0.5. D The duration of pre-internalisation pause is significantly increased in PfMyoB-cKO parasites, suggesting that PfMyoB plays a role in initiation of internalisation. The post-internalisation pause is reduced by a similar amount. Bars show median and interquartile range. Significance assessed by Mann-Whitney test, shown when p<0.5. E Example of a RAP-treated PfMyoA-K764E merozoite, undergoing a Type B failure, showing apparent completion of internalisation before subsequent ejection (S3 Video), and a PfMyoB-cKO merozoite undergoing successful invasion (S4 Video). Time indicated in seconds, scale bar 2 μm. F Schematic based on Fig 4A showing that PfMyoA-K764E merozoites can proceed to internalisation, but are frequently ejected, while PfMyoB-cKO merozoites invade successfully, but with delayed initiation of internalisation.

Fig 7. A stepwise model for Plasmodium myosin force generation during merozoite invasion.

A PfMyoA produces force at three sequential energetic barriers to drive invasion. Successively more disruptive mutations bring the PfMyoA force production capacity closer to, or below, the energetic barriers, represented by orange bars with blurred tops indicating variability in individual parasites and RBCs. Some PfMyoA-K764E parasites fail at initiation of internalisation (Type C failures), some fail at completion of internalisation (Type B failures) and relatively few complete invasion. PfELC-cKO parasites often fail at the barrier of deformation (Type D failures), though some can deform and fail at initiation of internalisation (Type C failures). PfMyoA-cKO parasites never overcome the energetic barriers to deformation or internalisation. PfMyoB-cKO parasites are only impaired at initiation of internalisation, suggesting that PfMyoB may act to reduce the energetic barrier at initiation of internalisation. B Overview of actomyosin functions at each of the three steps, including using deformability to select suitable host cells, driving initial wrapping of the merozoite and twisting shut the invasion pore. Parasite effectors reduce the energetic barriers by modulating RBC biophysical properties, including creation of the tight junction (TJ) to introduce a line tension, insertion of membrane material (darker membrane) and binding of adhesins like EBA-175 that may affect RBC lipid packing C PfMyoA is phosphorylated before egress and may be dynamically dephosphorylated at some point before internalisation to tune the motor for maximal force production.