Recent insights into apicomplexan parasite egress provide new views to a kill

Abstract

A hallmark of apicomplexan pathogens such as Plasmodium, Toxoplasma and Cryptosporidium is that they invade, replicate within, and then egress from their host cells. Egress usually results in lysis of the host cell, with deleterious consequences for the host. In the case of malaria, for example, much of the disease pathology is associated with cyclical waves of host erythrocyte destruction. This review highlights recent advances in mapping the signaling pathways that lead to egress and the parasite molecules involved in responding to and transmitting those signals. The review also discusses new findings for effector molecules that mediate disruption of the bounding membranes that enclose the intracellular parasite and the manner in which membrane rupture occurs to finally release invasive forms of the parasite.

Introduction

Egress is a crucial step in infection by all intracellular pathogens. This event liberates the infectious agent from an expiring or nutritionally exhausted abode so that it can enter a fresh host cell and return to the business of expanding its brood. While various classes of intracellular agents use a range of emancipative strategies including ones that don’t immediately kill the host cell (reviewed in [1], protozoa tend to use lethal cytolytic approaches that obliterate the cell. Cytolytic egress leads both to direct injury to tissues and also indirect collateral damage of the ensuing inflammatory and pyrogenic immune response leading to fever, a phenomenon exemplified by human malaria. Synchronous egress of malaria parasites from infected erythrocytes releases a bevy of pyrogens from the annihilated cells, causing the hallmark cyclical fevers of malaria. Malaria parasites, most notably Plasmodium falciparum, belong to the phylum Apicomplexa. This taxanomic group contains other important human pathogens including Toxoplasma, which causes opportunistic encephalitis and ocular disease, and Cryptosporidium, an agent of acute intestinal inflammation and diarrhea. Such parasites share an apical complex consisting of cytoskeletal structures and secretory organelles that are essential for motility, cell invasion and egress.

This review will focus on new developments in the mechanisms underlying egress by apicomplexan parasites, with special emphasis on cell exit by the pathogenic stages of the two most intensely studied genera, Plasmodium and Toxoplasma.

Kinases wire egress

Apicomplexan parasites harbor apical secretory organelles called micronemes that are crucial for parasite gliding motility, cell invasion and egress. Micronemes deploy transmembrane adhesive protein complexes that connect the substrate or host receptors with the parasite motor system (the glideosome) for propulsion and active invasion. For egress, micronemes discharge a perforin-like protein that, for T. gondii at least, disrupts host cell membranes to facilitated parasite escape and subsequent re-invasion of a new host cell [2,3]. While it has been known for some time that microneme secretion is promoted by elevation of cytosolic calcium within the parasite, only more recently have some of the molecular players in the calcium signaling pathway been identified.

Three simultaneously published studies provide new insight into the multiple modes of regulating cell egress by T. gondii [4–6]. Previous work by L. David Sibley’s group demonstrated that TgCDPK1, a member of a calcium dependent protein kinase family, is crucial for protein secretion from the micronemes [7]. Following up on this, they newly show that a second member of the family, TgCDPK3, is only required for cell egress and is not necessary for microneme secretion except when secretion is triggered by certain stimuli [4]. In parallel, Chris Tonkin’s group reported that TgCDPK3 is part of a calcium dependent signaling pathway that is triggered when the parasite senses a drop in environmental potassium occurring upon cell damage or permeabilization [5]. In a third parallel study, the Arrizabalaga and Boothroyd groups used whole genome sequencing of cell egress mutants to identify TgCDPK3 as an essential factor for calcium ionophore induced egress from host cells [6]. All three studies showed that TgCDPK3 is not required for gliding motility and cell invasion. The studies also established that TgCDPK3 localizes to the periphery of the parasite underlying the plasma membrane. The findings support a sensory hypothesis in which TgCDPK3 is part of a signaling pathway designed to sense changes in the environment that the parasite uses as cues for egress. If correct, this sensory role is most likely tied to situations where the parasite-infected cell has been damaged or is under direct assault from immune cells that recognize it as being infected.

The Sibley study also showed that activation of a third component of the egress-signaling cascade, cGMP-dependent protein kinase G (TgPKG), triggers egress[4]. Previous work using a selective apicomplexan PKG inhibitor termed compound 1 established that TgPKG was necessary for microneme secretion, but the new work revealed that activating TgPKG is also sufficient to induce microneme secretion and egress. TgPKG appears to function independent of calcium and it is proposed to act upon the same substrate(s) as TgCDPK3. Collectively, the findings illustrate the complex signaling pathways regulating T. gondii egress and suggest a model in which the parasite can respond to multiple signals to escape cells.

The TgCDPK3 findings generally mirror those of PfCDPK5, which was shown to be crucial for Plasmodium merozoite egress but not invasion [8]. Activation of PfCDPK5 may be dependent on release of calcium from internal parasite stores, since a gradual increase in cytoplasmic Ca²⁺ was observed in the hours leading up to egress in P. falciparum [9]. However, unlike TgCDPK3, PfCDPK5 appears to function downstream of PfPKG and it is crucial for natural egress. PfPKG is nonetheless required for egress, since both compound 1 and another PKG inhibitor called compound 2 potently block egress in P. falciparum blood-stages [10]. The identification of kinase substrates, such as those reported recently for TgCDPK1 [11], should provide important new insight into the signaling events accompanying apicomplexan egress. Progress in refining selective inhibitors [12] will also help realize the potential impact of extinguishing kinase activity on infection.

Do micronemes DOC2 the apical membrane?

While the identification of CDPKs revealed insight into upstream aspects of the calcium regulated signaling pathway governing microneme secretion, the extent to which fusion of micronemes with the parasite apical membrane depends on calcium regulation remained unknown. Recent work by the Gubbel and Duraisingh labs used whole genome sequencing to identify a putative double C2 domain protein (DOC2.1) as an essential factor for T. gondii microneme secretion, egress and invasion [13]. Similar proteins in other systems use the C2-domains for calcium dependent membrane association, thus it was proposed that DOC2.1 facilitates microneme fusion with the apical membrane, likely in association with other membrane fusion components including SNARES. In this model, DOC2.1 is predicted to redistribute to the apical membrane in response to elevated calcium, though low expression levels precluded testing this. Conditional knockdown of the P. falciparum ortholog also impaired microneme secretion and invasion. The authors concluded that the invasion defect was sufficient to explain the observed decrease in parasite propagation without involving an effect on egress. However, recent findings strongly suggest that calcium signaling regulates Plasmodium merozoite egress by triggering the secretion of exonemes, which contain an essential egress factor called SUB1 [14]. Thus, it will be interesting to address in future work whether DOC2.1 or another DOC2 protein is involved in exoneme secretion.

Rhoptry secretion is not required for egress

Rhoptries are club shaped secretory organelles that are discharged during invasion [15] and egress [16]. Although their roles in invasion and intracellular survival are well documented, the contribution of rhoptry contents to egress remained unknown until the recent characterization of T. gondii rhoptry secretion mutants [17,18]. The mutants, which fail to secrete the rhoptry contents due to a defect in rhoptry positioning, are invasion incompetent but show normal motility and they efficiently egress from host cells. While these findings suggest that rhoptry contents do not contribute to Toxoplasma egress, it remains to be determined if this is a general feature of other apicomplexans.

A new checkpoint for initiation of motility and egress

Micronemes are thought to fuse with the apical membrane at a site located within a thimble-like cytoskeletal structure termed the conoid (or apical prominence in Plasmodium merozoites). Recent work in T. gondii suggests that the conoid is also a resting site for at least one additional egress factor, a protein methyltransferase called TgAKMT required for the initiation of motility [19]. That TgAKMT null parasites are substantially defective in induced egress and invasion is consistent with the role of motility in these events. Interestingly, AKMT null parasites show normal microneme secretion, which likely explains the ability of such parasites to permeabilize the parasitophorous vacuole membrane (PVM) based on the release of TgPLP1 from the micronemes [2]. TgAKTM strikingly relocates from the conoid to the parasite cytoplasm upon elevation of intracellular calcium. Precisely how this redistribution affects the initiation of motility and the methylation of substrates awaits further study. Nonetheless, this study along with the DOC2.1 findings suggests that the redistribution of egress factors from and to the apical region likely contributes to the calcium dependent activation of the parasite glideosome and microneme secretion, respectively. It should be noted that, whilst activation of the glideosome must be required for invasion by Plasmodium merozoites, whether it is a prerequisite for egress of Plasmodium merozoites from hepatocytes or erythrocytes remains unclear, in part because active motility of these small zoites is difficult to visualize and quantify.

Action in the host cytoplasm

Whilst a multitude of signaling events in the parasite controlling egress are emerging, recent work suggests that an equally elaborate series of events unfolds in the cytoplasm of the infected host cell before egress. The finding that host calpain-1 is required for efficient egress of both T. gondii and P. falciparum suggested a role for the active enzyme in disrupting the host cytoskeleton [20]. Supporting this, a follow up study showed that inhibition or depletion of calpain-1 in malaria infected erythrocytes diminished proteolysis of several key cytoskeletal proteins including α/β-spectrins and ankyrin-1 during egress [21]. This work revealed evidence for perturbation of the host cytoskeleton beginning ~35 h post-infection and involving the loss of α-adducin from the cytoskeleton. Because α-adducin connects actin and spectrin filaments to the cytosolic face of the plasma membrane, its loss is expected to create lesions in the cytoskeleton. Consistent with this, atomic force microscopy revealed discontinuities in the cytoskeleton of infected erythrocytes [21]. Further work based on an RNAi interference screen of host signaling factors revealed a calcium-based signaling cascade required for egress of T. gondii and P. falciparum [22]. The pathway is proposed to involve a heterotrimeric G protein (including a Gαq component), which activates phospholipase C to generate diacylglycerol as an activator of protein kinase C. Protein kinase C phosphorylation inactivates α-adducin, causing cytoskeletal lesions. It was further suggested that these cytoskeletal perturbations trigger the opening of a plasma membrane mechanosensitive cation channel, TRPC6, thus activating calpain-1 via an influx of calcium from the extracellular medium. While a role for TRPC6 is well supported, other studies have shown depletion of extracellular calcium has little or no effect on egress [9,23], arguing against the calcium influx theory. Recent work by Glushakova and colleagues proposed that the PV is a key source of calcium that could be mobilized to the host cell during egress [9], and Garg et al. [24] also observed an increase in intracellular calcium levels in P. falciparum schizonts just prior to egress. Together, these studies raise the possibility that TRPC6 acts at the PVM to mobilize calcium to the host cytoplasm for activation of calpain-1 and possibly other calcium-responsive egress factors. In this model, the observed swelling of the PV that occurs several minutes before egress would activate TRPC6 for calcium release. Future work examining the subcellular distribution of TRPC6 and the use of genetically encoded calcium sensors in the PV could test this hypothesis.

Breaking down the barriers

Egress requires disruption of the PMV and host cell plasma membrane, as well as its underlying cytoskeleton. Parasite effector molecules directly responsible for rupture of these barriers at egress remain only partially defined. The T. gondii micronemal perforin-like protein TgPLP1 likely acts by punching holes in the PVM, a model supported by a recent detailed genetic and in vitro analysis [3] showing that TgPLP1 is sufficient for membrane disruption, with an essential requirement for the central MACPF domain and the β-sheet rich C-terminal domain, which has membrane-binding activity. Like many pore-forming proteins, TgPLP1 appears to oligomerize upon membrane binding to form large, multimeric membrane-embedded complexes. Whether the primary function of TgPLP1 is to weaken the PVM or to render the membrane permeable to components of the PV (e.g., ions, proteins) that subsequently activate downstream effector steps in egress, remains unknown.

There are 5 perforin-like genes in the Plasmodium genome (PPLP1-5) and so, in view of the TgPLP1 functional data, it is interesting to speculate on a possible equivalent role for a perforin-like protein in Plasmodium egress. Very recent work [24] has suggested a role for PPLP1 and/or PPLP2 in P. falciparum asexual blood-stage egress, showing that both proteins are expressed and that PPLP1 localizes to micronemes in merozoites. Secretion of PPLP1 appeared to be calcium-dependent, as was its subsequent binding to membranes of the parasitized host cell. Binding of PPLP1monomers to erythrocyte membranes led to PPLP1 oligomerization and lytic activity. Curiously however, Plasmodiumberghei parasites null for PPLP1 (Ishino et al 2005) or PPLP2 [25] displayed no phenotype in asexual blood stages, with PPLP2 disruptants showing only a block in gametocyte egress from their host erythrocytes. This suggests a different role for these perforin-like molecules in different Plasmodium species, or the existence of some functional redundancy. Ultimately, the importance of perforin-like proteins in Plasmodium asexual blood-stage egress may need to be confirmed in additional genetic studies.

In Plasmodium, pharmacological or genetic strategies have implicated the two classes of endoproteases: subtilisin-like serine protease SUB1 and the cysteine protease-like SERA proteins. Plasmodium SUB1 is stored in exonemes until its discharge into the PV lumen shortly before egress. A diverse range of parasite-derived substrates has been identified for SUB1, including merozoite surface proteins and PV-resident members of the SERA family [26–28]. Of the nine P. falciparum SERA genes, only SERA5 (serine-type active site) and SERA6 (cysteine-type active site) are essential in asexual blood stages [29]. Although SERA5 has not been shown to possess enzyme activity a recombinant protein corresponding to its presumed prodomain, or a short peptide thereof, showed egress-inhibitory activity in P. falciparum cultures [30]. Intriguingly, given the apparent essentiality of SERA5 in P. falciparum, disruption of the only two Ser-type genes in P. berghei had no phenotypic consequences throughout the parasite life cycle [31], though it resulted in increased expression levels of the Cys-type PbSERA3 (the orthologue of P. falciparum SERA6). To date, attempts to disrupt PbSERA3 have been unsuccessful, suggesting it is indispensable in blood stages. A genetic analysis of P. falciparum SERA6 [32] indicated an essential role for its putative active site Cys residue, further supporting suspicions that it is a protease. Mutations that block SERA6 cleavage by SUB1 were lethal, and cleavageby P. falciparum SUB1 of a recombinant form of the P. berghei orthologue activated an E64-sensitive autocatalytic processing activity. Collectively these observations suggest a model in which SUB1 processing activates SERA6 protease activity. However, a direct role for SERA6 or any other blood-stage SERA in egress has yet to be demonstrated, and – if these proteins are indeed proteases - their physiological substrates are unknown.

P. falciparum dipeptidyl peptidase 3 (DPAP3) emerged as a putative egress regulator through use of selective small molecule inhibitors and activity-based probes [33,34]. DPAP3 inhibition reduced expression of mature SUB1 and also reduced SERA5 processing. A proteomic analysis designed to identify proteolytic events associated with egress, combined with the use of a selective DPAP3 inhibitor, revealed that 64% of the proteolytic events occur downstream of DPAP3 activation [35]. However, as with the SERA family, direct substrates of DPAP3 have yet to be identified.

There is still no satisfactory mechanistic explanation for how proteases might mediate membrane disruption at egress, but evidence for their crucial role in the pathway continues to accumulate. An electron microscopic study of P. falciparum gametocyte egress [36] confirmed the ‘inside-out’ model implied by several previous studies (including the PPLP2 knockout study mentioned above), showing that the PVM ruptures at several sites within seconds of activation of egress, followed by eventual erythrocyte membrane rupture at a single point. Rupture of the two membranes was differentially sensitive to protease inhibitors, whilst PVM rupture was induced by the potassium ionophore nigericin suggesting that, as in T. gondii [37], Plasmodium senses a drop in potassium within the expiring host cell as a trigger for egress. It is unlikely that this is the only trigger for malaria parasites though since it was recently shown that egress of asexual blood-stage parasites is not affected by growing the parasite in medium containing potassium levels equivalent to those inside the host cell [38]. Finally, some caution should be exercised in comparisons of egress of Plasmodium asexual and sexual stages, since distinctions in the regulation of egress between different developmental stages likely exist.

There is now a general consensus that PVM rupture occurs before host cell membrane rupture. The PV is a relatively Ca²⁺-rich compartment, so it is likely that PMV rupture (and/or its prior permeabilization by pore-forming proteins) enables rapid upregulation of [Ca²⁺] in the host cell cytosol, which might facilitate activation of calpain-1. It is worth noting that SUB1 enzymatic activity is calcium-dependent, so high Ca²⁺ levels in the PVM are likely important for its role in egress.

The big event

Several groups have taken biophysical approaches to dissect the rapid dynamics of P. falciparum merozoite egress from the infected erythrocyte and the associated mechanical alterations to the host cell. A study combining diffraction phase microscopy, atomic force microscopy and optical tweezers showed in increased membrane stiffness of the infected erythrocytes blocked for egress at a stage following PVM rupture [39]. Elevated perturbations of the erythrocyte membrane following PVM rupture were interpreted as being due to the trapped mobile merozoites striking the membrane surface. It was concluded that the inhibitors prevented the cytoskeletal destabilization required for normal egress. Time-lapse microscopy has revealed that egress is preceded by a transient swelling of the PV and shrinkage of the residual host cell cytoplasmic compartment accompanied by overall swelling of the entire schizont, suggestive of a build-up of osmotic pressure. Increased mobility or disaggregation of the intracellular merozoites then occurs, presumed to be due to their release from the central residual body and breakdown of the PVM [40–42]. Shortly before final schizont rupture – which is an almost explosive event in this species - ‘poration’ (permeabilization) of the erythrocyte membrane occurs, elegantly shown by the accessibility of erythrocyte cytoskeleton to externally-provided fluorescent phalloidin in the moments leading up to rupture, as well as the loss of a GFP-tagged transgene product from the erythrocyte cytosol [24,40,41]. This mirrors analogous findings in T. gondii [43]. Intriguingly, poration of P. falciparum schizonts was not blocked by the cysteine protease inhibitor E64, indicating that it is independent of calpain-1 activity.

Host erythrocyte rupture itself is rapid, taking place in ~400 milliseconds. High-speed video microscopy was used to show that egress comprises three mains steps: (1) the appearance of a single ~1 μm pore allowing the osmotically-driven release of 1–2 merozoites; (2) an outward curling of the erythrocyte membrane from the pore epicentre; and (3) finally a rapid inversion (‘buckling’) of the entire erythrocyte membrane, acting to physically eject and disperse the remaining parasites [42,44]. The latter 2 steps were considered to rely on elastic instability of the infected erythrocyte membrane, perhaps aided by prior modification and/or degradation of the underlying erythrocyte cytoskeleton.

In contrast to the explosive nature of Plasmodium egress from the erythrocyte, video and electron microscopic examination of induced T. gondii egress has revealed a process that much more resembles invasion, with some distinctions. Upon induction of egress with calcium ionophore the intracellular tachyzoite rosettes disassemble, the parasites extend their conoids and begin to move within the PV, eventually rupturing it to move through the host cell cytoplasm and breach the plasma membrane [45]. The role of gliding motility in induced egress was very evident, although motility is not absolutely required for natural egress [46]. Clearly the major challenge now in both Plasmodium and Toxoplasma is to consolidate all these observations with an improved understanding of the molecular events leading up to egress.

Conclusions

Egress of apicomplexan pathogens is fundamental to their life cycle and important for pathogenesis. As this review illustrates, increased attention has produced substantial advances in our understanding of egress over recent years. Despite this, important questions remain unanswered. What is the natural trigger for egress in Plasmodium blood stages? How are the calcium-dependent and cGMP-dependent signals that lead to egress sensed and orchestrated? Are secreted microneme/exoneme proteins involved in egress (e.g., TgPLP1, SUB1) selectively discharged prior to those involved in invasion? Why does the malaria parasite possess apparently abundant PV-resident endoproteases for egress when T. gondii appears to lack these? How do proteases and pore-forming proteins rupture the membrane that enclose the intracellular parasite, and are additional effector components involved? Answers to these and other central questions will allow full exploitation of the resulting knowledge for disease control.

Highlights.

• Egress is a fundamental event in the biology of apicomplexan parasites

• Kinases, a methyl transferase and a putative vesicular fusion protein regulate egress

• Signal transduction in the host cytosol aid in dismantling the cytoskeleton for parasite exit

• A protease cascade in the malaria parasitophorous vacuole is required for egress

• Egress occurs via rupturing membranes from the inside out