Selective inhibition of a two-step egress of malaria parasites from the host erythrocyte. Melbourne 3050, Australia. Melbourne 3050, Australia

Transcription

1 JBC Papers in Press. Published on July 11, 2003 as Manuscript M Selective inhibition of a two-step egress of malaria parasites from the host erythrocyte Mark E. Wickham 1, Janetta G. Culvenor 2 and Alan F. Cowman 1 1 The Walter and Eliza Hall Institute of Medical Research *Correspondence to: Melbourne 3050, Australia 2 Department of Pathology, The University of Melbourne Melbourne 3050, Australia Alan F. Cowman The Walter and Eliza Hall Institute of Medical Research 1G Royal Parade Melbourne 3050 Australia Telephone: Facsimile: cowman@wehi.edu.au Running title: Inhibition of P.falciparum host cell exit Copyright 2003 by The American Society for Biochemistry and Molecular Biology, Inc.

2 2 SUMMARY Escape from the host erythrocyte by the invasive stage of the malaria parasite Plasmodium falciparum is a fundamental step in the pathogenesis of malaria of which little is known. Upon merozoite invasion of the host cell, the parasite becomes enclosed within a parasitophorous vacuole, the compartment in which the parasite undergoes growth followed by asexual division to produce daughter merozoites. These daughter cells are released upon parasitophorous vacuole and erythrocyte membrane rupture. To examine the process of merozoite release we have used P.falciparum lines expressing GFP-chimeric proteins targeted to the compartments from which merozoites must exit; the parasitophorous vacuole and the host erythrocyte cytosol. This allowed visualization of merozoite release in live parasites. Here we provide the first evidence in live untreated cells that merozoite release involves a primary rupture of the parasitophorous vacuole membrane followed by a secondary rupture of the erythrocyte plasma membrane. We have confirmed that parasitophorous vacuole membrane rupture occurs prior to erythrocyte plasma membrane rupture in untransfected wild-type parasites using immunoelectron microscopy. We have also demonstrated selective inhibition of each step in this two-step process of exit using different protease inhibitors, implicating the involvement of distinct proteases in each of these steps. This will facilitate the identification of the parasite and host molecules involved in merozoite release.

3 3 INTRODUCTION During the erythrocytic phase of the P.falciparum lifecycle, the merozoites released initially from infected hepatocytes attach to and invade human erythrocytes in the bloodstream. Upon invasion of the host cell, the parasites become enclosed in a parasitophorous vacuole in which, as ring and trophozoite stages, they undergo growth followed by asexual division (schizogony) to produce daughter merozoites. It is from both this vacuole and the host erythrocyte that the newly formed merozoites in the schizont must escape. While little is known about the molecules that mediate this process, proteases have been implicated in both parasite exit from the erythrocyte and the subsequent invasion into erythrocytes through the use of protease inhibitors that halt these processes (1-3). Following the demonstration that P. knowlesi schizonts incubated in the presence of the protease inhibitors chymostatin and leupeptin show decreased reinvasion due to inhibition of schizont maturation (3,4), Lyon and Haynes demonstrated that P.falciparum schizonts cultured in the presence of the inhibitors chymostatin, leupeptin, antipain and pepstatin also fail to rupture properly (2). Recently it has been demonstrated that the protease inhibitors E-64 and E-64d inhibit schizont maturation (5,6). The role of the proteases involved in schizont rupture is presumably to degrade both the parasitophorous vacuole and the erythrocyte membrane skeleton, thereby facilitating release. However, the proteases involved have not yet been fully characterized. A number of inhibitor-sensitive parasite proteases have been implicated in schizont rupture, including the aspartic protease plasmepsin II which has been demonstrated to digest spectrin, actin and protein 4.1 at neutral

4 4 ph (7), falcipain-2; a cysteine protease that cleaves erythrocyte ankyrin and protein 4.1 (8-11), and the cysteine protease-like SERA (serine repeat antigen) family, the members of which possess a centrally located papain-like protease domain (1,12-14). Interestingly, inhibition of the proteolytic processing of SERA is possible using protease inhibitors that block schizont rupture (15). The relevance of these protease activities to the process of merozoite release remains to be determined, and the availability of the genome sequence of P.falciparum has facilitated the identification of further protease-like molecules that may play a role (12,16). Ultrastructural evidence suggests that during schizogony the parasite plasma membrane invaginates to surround the merozoites forming within the confines of the parasitophorous vacuole (17), and that late in schizogony, the parasitophorous vacuole membrane may be absent with the fully formed merozoites free within the host erythrocyte (17,18). This suggests that escape from the parasitophorous vacuole occurs prior to exit from the erythrocyte, but this possibility remains uninvestigated. An alternate model for schizont rupture has been proposed which involves escape from the host erythrocyte of the merozoites enclosed within the parasitophorous vacuole, followed by extraerythrocytic rupture of the vacuole and release of invasive merozoites (Figure 1 A) (5). Here we examine the mechanism of parasite exit from the host erythrocyte in both wild-type and transgenic P.falciparum lines. Presented here is the first evidence in live untreated cells that rupture of the parasitophorous vacuole membrane occurs while the parasite is within the erythrocyte (Figure 1 B). The

5 5 mixing of vacuolar and erythrocytic contents subsequent to this intraerythrocytic vacuolar rupture is also detected in wild-type parasites. Each step in the process of escape is selectively inhibitable by different protease inhibitors, indicating that different proteases mediate each event. EXPERIMENTAL PROCEDURES Parasite lines. To generate transgenic Plasmodium falciparum-infected erythrocytes expressing KAHRP-GFP chimeric proteins we made transfection constructs in the plasmid vector phh2 (19,20) as described (21). A region of the KAHRP gene encoding the first 60 amino acids, which includes a putative hydrophobic signal sequence of eleven amino acids, was joined upstream of the GFP coding sequence in the transfection vector phh2 (20). Parasites stably transfected with this construct traffic the GFP fusion into the parasitophorous vacuole throughout the asexual lifecycle (21). These parasites are designated 3D7-His. A region of the KAHRP gene encoding the first 123 amino acids of KAHRP was joined upstream of the GFP coding sequence. This region of KAHRP contains both the putative hydrophobic signal sequence required for transit to the parasitophorous vacuole, and the histidine rich region that contains the signal required for translocation into the erythrocyte (21). Transgenic parasites expressing this construct traffic the GFP fusion into the erythrocyte cytosol throughout the asexual lifecycle. These parasites are designated 3D7+His. The 3D7 cloned P.falciparum parasites were transfected by electroporation and drug selected using 0.25 nm WR99210 as previously described (22). Fractionation of infected erythrocytes and western blotting.

6 6 For permeablisation of infected red cells with streptolysin-o (SLO), haemolytic activity of SLO was determined and 2x10 7 parasites were incubated in RPMI containing 3-4 haemolytic units of SLO as described (23,24). In brief, P.falciparum-infected erythrocytes were incubated in 3-4 haemolytic units of SLO in RPMI for 6 min at room temperature, the supernatant resuspended in Laemmli sample buffer, the pellet washed in RPMI and subsequently resuspended in Laemmli sample buffer. For saponin lysis, 2x10 7 parasites were incubated in 1.5 volumes of 0.15% saponin for 10 min on ice, centrifuged and the supernatant resuspended in Laemmli sample buffer, the pellet washed in PBS and subsequently resuspended in Laemmli sample buffer. Proteins were separated by SDS-SAGE and transferred to PVDF and visualised by ECL using mouse anti-gfp antiserum (1:1000). Protease Inhibitor Treatment. Parasitised cells were synchronised by two consecutive sorbitol treatments 4 hr apart, cultured until middle-stage schizonts, then treated with 10 mm of L- transepoxy-succinyl-leucylamido-(4-guanidino)butane (E-64) (Sigma) as described (5) or a combination of 10 mg/ml each of leupeptin and antipain, or leupeptin and chymostatin (Sigma) as described (2). Indirect Immunofluorescence Assay. Indirect immunofluorescence assays were performed on control and protease treated P.falciparum-infected erythrocytes smeared on glass slides and fixed with methanol. Slides were incubated sequentially with rabbit anti-kahrp (1:200). The slides were then incubated with anti-rabbit antibodies conjugated to FITC (1:1000) in the presence of the nuclear stain 4,6-diamino-2-phenylindole (DAPI)

7 7 at final concentration of 2 mg/ml. Samples were viewed with a Carl Zeiss Axioskop with a PCO SensiCam and Axiovision 3 software. Fluorescence Microscopy. Ring stage parasites were synchronised using two consecutive sorbitol treatments 4 hours apart, cultured until early stage schizonts, and samples taken hourly during the process of schizongony and merozoite release. Protease inhibitor treated and control parasites were cultured in DAPI at final concentration of 2 mg/ml for 30 min at 37ºC immediately prior to imaging. Fluorescence from GFP and DAPI was observed and captured in live cells at 20ºC within 20 min of mounting the sample in culture medium under a coverslip on a glass slide using a Carl Zeiss Axioskop with a PCO SensiCam and Axiovision 3 software. Immunoelectron Microscopy. Control and protease inhibitor treated cultures were fixed in 0.25% glutaraldehyde in 0.1 M phosphate buffer ph 7.4 for 10 min at room temperature, followed by addition of 0.05 M NH 4 Cl and washed in 0.1 M phosphate buffer. Samples were dehydrated in 70% ethanol and embedded in L. R. White resin and polymerised for 4 hr at 50ºC or 5 days at 37ºC. Thin sections were incubated in rabbit anti-s-antigen antibodies in 0.05 M phosphate, ph 7.4/0.1% Tween 20/1% BSA, washed then protein A-5 or 10 nm gold in 0.05 M phosphate, ph 7.4/0.1% Tween 20/1% BSA, then stained in uranyl acetate.

8 8 RESULTS Intraerythrocytic parasitophorous vacuole rupture in GFP-expressing parasites To investigate the sequence of events involved in schizont rupture transgenic P.falciparum lines have been generated which traffic GFP fusions to the compartments the parasite must traverse upon exit; the parasitophorous vacuole and the host erythrocyte cytosol. To traffic GFP to the parasitophorous vacuole, a portion of the KAHRP gene that encodes the first 60 amino acids of the protein was joined upstream of the GFP coding sequence in the transfection vector phh2 (19,20). This region of KAHRP includes a putative hydrophobic signal sequence of eleven amino acids flanked by lysine residues. Transgenic parasites stably transfected with this construct traffic the fusion via the canonical secretory pathway into the parasitophorous vacuole throughout the asexual lifecycle. These parasites are designated 3D7-His. To traffic GFP into the host erythrocyte, a region of the gene encoding the first 123 amino acids of KAHRP was joined upstream of the GFP coding sequence. This region of KAHRP contains both the putative hydrophobic signal sequence required for transit to the parasitophorous vacuole, and the histidine rich region that contains the signal required for translocation into the erythrocyte. Transgenic parasites expressing this construct traffic the GFP fusion into the erythrocyte cytosol throughout the asexual lifecycle via an extension of the parasite s canonical secretory pathway in the host cell. These parasites are designated 3D7+His. The compartmentalization of GFP in these transgenic P.falciparum lines during asexual division (schizogony) and escape from the host erythrocyte was

9 9 examined. The transgenic parasites were made synchronous with regard to cell cycle phase by sorbitol lysis and followed by fluorescence microscopy. Early in schizogony the transgenic lines were found to exhibit GFP fluorescence in the compartments to which the GFP fusions are trafficked; the parasitophorous vacuole in 3D7-His parasites and the host erythrocyte cytosol in 3D7+His parasites (Figure 2 A and B). In late trophozoites and schizonts, some GFP can be seen in association with the food vacuole and hemozoin; GFP appears to reenter the parasites with erythrocyte cytoplasm ingestion (19,21). However, in both transgenic lines late in schizogony GFP consistently localised to compartments to which it is not targeted (Figure 2 C and D). In 3D7-His parasites, the GFP fusion that lacks the signal required for translocation into the host erythrocyte, was found present in the erythrocyte cytosol (Figure 2 C, white arrow). A possible explanation for this localization is that the parasitophorous vacuole membrane has lysed, and the GFP fusion has flooded into the erythrocyte by free diffusion. This intraerythrocytic vacuolar rupture and subsequent diffusion of GFP from the vacuole into the host erythrocyte has also been observed during confocal sectioning of late schizonts of an unrelated transgenic line that also targets GFP to the parasitophorous vacuole (Melanie Rug, personal communication). In 3D7+His parasites, the GFP fusion was observed both in the erythrocyte cytosol, the compartment to which it is targeted, and immediately surrounding the fully formed merozoites (Figure 2 C, black arrow). However, in early schizonts the GFP fusion is excluded from the vacuole surrounding the merozoites. This again supports an intraerythrocytic rupture of the vacuolar membrane, allowing diffusion of GFP from the erythrocyte to surround the

10 10 merozoites. Fractionation of GFP-expressing parasites To examine this intraerythrocytic rupture of the vacuolar membrane at a population level, cell fractionation using saponin and streptolysin O was utilized. Saponin fractionation allows analysis of the export of proteins from the parasite as it lyses both the erythrocyte plasma and parasitophorous vacuole membranes, leaving the parasite plasma membrane intact. Upon fractionation with saponin, proteins retained within the parasite will be present in the pellet and those exported from the parasite will be detected in the supernatant. Streptolysin O fractionation allows analysis of the export of proteins from the parasite since it permeabilises only the erythrocyte plasma membrane, leaving the parasitophorous vacuole membrane intact. Proteins trafficked into the vacuole will be present in the pellet and those trafficked into the host erythrocyte present in the supernatant. In 3D7-His trophozoites the GFP fusion is trafficked from the parasite, indicated by the predominance of the GFP fusion in the saponin supernatant and into the parasitophorous vacuole but not beyond, indicated by the presence of the fusion in the streptolysin O pellet but not supernatant (Figure 2 E). Fractionation of late 3D7-His schizonts (the lifecycle stage in M phase) shows that the GFP fusion is again trafficked from the parasite predominantly into the parasitophorous vacuole, but is also present in the erythrocyte cytosol, indicated by GFP fusion present in the streptolysin O supernatant. This localisation at a population level (2x10 7 parasites) agrees with that observed by microscopy on individual parasites and is consistent with intraerythrocytic rupture of the parasitophorous vacuole membrane during schizogony and diffusion of the GFP fusion into the

11 11 erythrocyte cytosol. Intraerythrocytic parasitophorous vacuole rupture in wild-type parasites. To confirm that parasitophorous vacuole rupture occurs within the erythrocyte in wild-type parasites, immuno-electron microscopy using antibodies to S- Antigen, a P.falciparum protein that localizes to the parasitophorous vacuole (25,26) was performed. In untransfected trophozoites S-Antigen labelling was observed in the parasitophorous vacuole (Figure 3 A). However, late in schizogony the vacuolar marker is observed throughout the erythrocyte cytosol (Figure 3 B), consistent with intraerythrocytic lysis of the parasitophorous vacuole membrane and subsequent diffusion of the vacuolar contents, including S-antigen, into the erythrocyte. This is consistent with the apparent flooding of cytosolic material from the erythrocyte into the vacuole observed previously, presumably following vacuolar lysis (27). Intraerythrocytic lysis of the vacuole membrane in untransfected P.falciparum lines rules out the possibility that intraerythrocytic lysis of the vacuole in transgenic lines is an artefact of GFP expression. Selective inhibition of parasitophorous vacuolar and erythrocyte plasma membrane rupture To establish that the altered localisation of GFP in late stage parasites is attributable to intraerythrocytic lysis of the vacuolar membrane, and to exclude the possibility that the GFP fusions are trafficked to the different compartments late in schizogony, selective inhibition of each step in the mechanism of escape from the host cell was attempted. It has been shown that the protease inhibitors E-64 and E-64d inhibit lysis of the parasitophorous vacuole membrane but not

12 12 the lysis of the erythrocyte plasma membrane (5,6). Addition of E-64 to both 3D7-His and 3D7+His transgenic lines caused an accumulation of the parasitophorous vacuole membrane-enclosed merozoite structures observed previously (Figure 4 A and schematically, B). The compartment surrounding the merozoites can be identified as the parasitophorous vacuole since these structures of the 3D7-His line, which traffics GFP to the vacuole, exhibit GFP fluorescence (Figure 4 A, schematically, B). The lack of GFP fluorescence in merozoite structures of the 3D7+His line, which exports GFP to the erythrocyte cytosol, indicates that lysis of the erythrocyte plasma membrane has occurred, releasing the GFP fusion into the culture supernatant, and that the limiting membrane observed is of vacuolar origin. The lack of inhibition of erythrocyte plasma membrane lysis in the presence of E-64 implicates distinct proteases in the lysis of the two membranes. Salmon et al. (2001) made the important observation that morphologically similar clusters to those seen upon E-64 treatment are observed at low frequency in untreated cultures, and proposed a two-step model for host cell exit. The current model for schizont rupture involves a primary lysis of the erythrocyte plasma membrane and release of the merozoites still enclosed within the parasitophorous vacuole membrane, followed by extraerythrocytic proteolysis of the vacuole membrane and merozoite dispersal (5). It has also been shown that schizonts cultured in the presence of the protease inhibitors leupeptin and antipain, like E-64, fail to rupture and remain surrounded by a limiting membrane. Using polyclonal serum raised against human erythrocytes, this membrane was identified as the erythrocyte plasma membrane (2). However, since the parasitophorous vacuole may contain

13 13 components of the erythrocyte plasma membrane either acquired upon invasion or later internalised (28,29), it is possible that this membrane is of vacuolar origin (5). Determination of Membrane Origin To determine both the origin of the limiting membrane of incompletely ruptured schizonts formed upon leupeptin and antipain treatment, the transgenic lines were treated with these inhibitors. Addition of leupeptin and antipain to the 3D7-His transgenic line caused an accumulation of the limiting membrane-enclosed merozoite structures observed previously (Figure 4 C and schematically, D). This limiting membrane is the erythrocyte plasma membrane since addition of leupeptin and antipain to both transgenic lines caused an accumulation of limiting membrane-enclosed merozoite structures that exhibit GFP fluorescence (Figure 4 C and schematically, D). A vacuolar origin of the limiting membrane would result in membrane-enclosed merozoite structures that did not exhibit GFP fluorescence in the 3D7+His line which traffics GFP beyond this membrane. As with (untreated) late schizonts of the 3D7+His transgenic line (Figure 2 C and D), GFP fluorescence was observed immediately surrounding the fully formed merozoites in the 3D7+His erythrocyte membrane-enclosed merozoite structures. Again, this supports an intraerythrocytic rupture of the vacuolar membrane, allowing diffusion of GFP from the erythrocyte to surround the merozoites. Therefore leupeptin and antipain inhibit the lysis of the erythrocyte plasma membrane without inhibiting lysis of the parasitophorous vacuole. The lack of inhibition of parasitophorous vacuole membrane lysis in the presence of leupeptin and antipain indicates that the process of escape is a two-step event involving

14 14 distinct proteases, both steps of which can be selectively inhibited. To confirm the origin of the limiting membrane in protease inhibitor treated cells, immunoelectron microscopy and indirect immunofluorescence assays using antibodies to parasite proteins exported from the parasite were performed. Immunoelectron microscopy was performed using antibodies to the vacuolar protein S-Antigen on thin sections of leupeptin and chymostatin treated wild-type parasites. In early schizonts we observe S-antigen labelling in the parasitophorous vacuole (Figure 3 C). However, late in schizogony the vacuolar marker is observed throughout the host erythrocyte cytosol (Figure 3 D) consistent with selective inhibition of erythrocyte plasma membrane, but not vacuolar, lysis. Using indirect immunofluorescence, PfEMP-3, a parasite protein that localizes under the host erythrocyte plasma membrane is present in merozoite clusters produced by leupeptin and antipain, but not E-64, treatment (Figure 4 E). This indicates that the origin of the limiting membrane in leupeptin and antipain-treated cells is the erythrocyte plasma membrane. DISCUSSION These data provide the first direct evidence that P.falciparum merozoites escape from the host erythrocyte is a two-step process involving a primary exit from the vacuole it acquires upon entry, followed by a secondary exit from the erythrocyte itself (Figure 1 B). It has been shown ultrastructurally that early in schizogony the parasite plasma membrane invaginates to surround the merozoites forming within the confines of the parasitophorous vacuole (17). Late in schizogony the parasitophorous

15 15 vacuole membrane may be absent with the fully formed merozoites free within the host erythrocyte (17,18). This is consistent with an intraerythrocytic rupture of the parasitophorous vacuole membrane. Indeed late stage schizonts have been observed with the parasitophorous vacuole either intact or partially intact with material of the same density either side of the parasitophorous vacuole membrane (Ross Waller, personal communication). While it is difficult to argue on the basis of density of the material either side of the vacuolar membrane that the parasitophorous vacuole has become permeable to the host erythrocyte cytosol in these cells, this in combination with the localisation of the vacuolar marker S-Antigen to the host erythrocyte cytosol (Figure 3), provides ultrastructural evidence for intraerythrocytic rupture of the vacuole membrane. This sequence of events clearly occurs at the population level, as detected by the release of parasitophorous vacuolar contents into the host erythrocyte cytosol. Both steps in this two-step process can be selectively inhibited; the primary vacuolar lysis by E-64 and the secondary erythrocyte membrane rupture by leupeptin and antipain, indicating that independent proteases mediate each step. Since E-64 is an irreversible inhibitor of cysteine proteases that does not inhibit serine proteases, it is likely that the protease(s) involved in parasitophorous vacuole lysis is a cysteine protease. The components of the parasitophorous vacuole that such a protease(s) would proteolytically process remain to be identified. The protease inhibitors shown to inhibit erythrocyte plasma membrane rupture are leupeptin, antipain, chymostatin and pepstatin (2,3,30). Leupeptin and antipain, used in this study at 10 mg/ml, each inhibit both trypsin-like serine proteases and cysteine proteases. Pepstatin is a nonselective inhibitor of aspartic proteases, and chymostatin inhibits both serine

16 16 proteases and cysteine proteases both of these inhibit schizont maturation at 10 mg/ml (2). It is therefore possible that the protease(s) mediating erythrocyte plasma membrane rupture is a cysteine or serine protease. However, it is likely that the rupture of each membrane involves a cascade of events that may be inhibited at different points using inhibitors with different specificities. As such it remains possible that erythrocyte plasma membrane rupture involves both an aspartic protease and a cysteine or serine protease. It is important to note that E-64, which inhibits parasitophorous vacuolar lysis, appears to do so only when added to middle stage schizonts (15) - no inhibition is observed with late stage schizonts (5). One explanation for this lack of inhibition of vacuolar rupture in late stage schizonts is that the parasitophorous vacuole membrane degradation has already commenced in these parasites. That leupeptin and antipain do not inhibit parasitophorous vacuole rupture may suggest that these inhibitors do not access the parasitophorous vacuole of infected erythrocytes. However, it has been demonstrated that leupeptin added to parasite cultures inhibits the processing of SERA, which localises to the parasitophorous vacuole, from a 56 kda fragment to a 50 kda fragment (15), indicating that leupeptin is indeed able to access the parasitophorous vacuole of P.falciparum infected erythrocytes. Additionally, the more membrane permeable analogue of E-64, E-64d, that lacks charged groups, has been demonstrated to block schizont development (6) again suggesting that the site of action of these inhibitors does not result from permeability differences. The inhibitor-sensitive proteases identified so far which may be involved in

17 17 erythrocyte plasma membrane rupture are the aspartic protease plasmepsin II, which cleaves erythrocyte spectrin, actin and protein 4.1 (31), the cysteine protease falcipain-2, which cleaves erythrocyte ankyrin and protein 4.1 (8,32), the putative serine protease ABRA (33) or the SERA/SERPH family of serine protease-like molecules. These molecules have been shown to localize to the parasitophorous vacuole and some, such as ABRA and the SERA/SERPH family, have been shown to be weakly associated with the merozoite surface (34). It has been recently demonstrated that SERA is associated with the parasitophorous vacuole membrane (15). Additionally, processing of baculovirus-expressed SERA is can be inhibited by a number of protease inhibitors, including those shown to inhibit rupture of the vacuole and erythrocyte plasma membrane (such as E-64 and leupeptin), and that the proteases responsible for this processing also appear membrane associated (15). It is possible that degradation of the parasitophorous vacuole membrane is required for these proteases to gain access to their substrates at the erythrocyte plasma membrane, but the degradation of the erythrocyte plasma membrane in the presence of E-64 suggests that the proteases responsible may be actively trafficked across the vacuolar membrane. The identification of the molecules mediating the process of schizont rupture is complicated by correlating inhibitor studies based on cell-free assays with experiments on P.falciparum parasite culture. For instance, leupeptin inhibits schizont rupture at concentrations of up to 68 mg/ml (15) through the inhibition of erythrocyte plasma membrane rupture (2). At higher concentrations (678 mg/ml), leupeptin inhibits both the proteolytic processing of SERA in cell-free assays (15), and the cleavage of ankyrin by falcipain-2 (9). Likewise E-64, which

18 18 clearly inhibits parasitophorous vacuolar rupture at 10 mm (5), also inhibits SERA and ankyrin cleavage at higher concentrations (100 mm and mm, respectively) (9,15). Consequently, the roles of these candidate protease activites in the process of escape remain to be elucidated. There has already been some success in the design of inhibitors selective for the aspartic proteases plasmepsin I and II (35-37) and the papain-family cysteine proteases known as the falcipains (6,11,38-42). Most recently, a screening of chemical libraries has identified inhibitors of falcipain-1 that have facilitated the identification of the role of falcipain-1 in red blood cell invasion (6). Parasites expressing GFP will prove an invaluable tool both in the search for the parasite molecules mediating the process of escape, and for the screening of inhibitors or combinations of inhibitors targeting this process. ACKNOWLEDGEMENTS We thank H-G. W. Meyer, S. Bhakdi and A. Hibbs for the generous gift of SLO. We thank M. Duraisingh, B. Crabb, M. Rug, and A. Maier for helpful discussions. We thank the Red Cross Blood Service (Melbourne, Australia) for supply of red cells and serum. A.C. is supported by a Howard Hughes International Research Fellowship from the Howard Hughes Medical Institute. This work was supported by a grant from the National Institutes of Health USA (RO1 AI44008) and the National Health and Medical Research Council of Australia. M.W. is supported by an Australian Postgraduate Research Award.

19 19 REFERENCES 1. Delplace, P., Bhatia, A., Cagnard, M., Camus, D., and Colombet, G. (1988) Biol. Cell. 64, Lyon, J. A., and Haynes, J. D. (1986) J Immunol 136, Hadley, T., Aikawa, M., and Miller, L. H. (1983) Exp. Parasitol. 55, Banyal, H. S., Misra, G. C., Gupta, C. M., and Dutta, G. P. (1981) J. Parasitol. 67, Salmon, B. L., Oksman, A., and Goldberg, D. E. (2001) Proc Natl Acad Sci U S A 98, Greenbaum, D. C., Baruch, A., Grainger, M., Bozdech, Z., Medzihradszky, K. F., Engel, J., DeRisi, J., Holder, A. A., and Bogyo, M. (2002) Science 298, Le Bonniec, S., Deregnaucourt, C., Redeker, V., Banerjee, R., Grellier, P., Goldberg, D. E., and Schrevel, J. (1999) J. Biol. Chem. 274, Dua, M., Raphael, P., Sijwali, P. S., Rosenthal, P. J., and Hanspal, M. (2001) Mol. Biochem. Parasitol. 116, Raphael, P., Takakuwa, Y., Manno, S., Liu, S. C., Chishti, A. H., and Hanspal, M. (2000) Mol. Biochem. Parasitol. 110, Shenai, B. R., Sijwali, P. S., Singh, A., and Rosenthal, P. J. (2000) J. Biol. Chem. 275, Hanspal, M., Dua, M., Takakuwa, Y., Chishti, A. H., and Mizuno, A. (2002) Blood 100, Miller, S. K., Good, R. T., Drew, D. R., Delorenzi, M., Sanders, P. R., Hodder, A. N., Speed, T. P., Cowman, A. F., de Koning-Ward, T. F., and Crabb, B. S. (2002) J Biol Chem 277, Delplace, P., Fortier, B., Tronchin, G., Dubremetz, J. F., and Vernes, A. (1987) Mol. Biochem. Parasitol. 23, Knapp, B., Hundt, E., Nau, U., and K$upper, H. A. (1989) Mol Biochem Parasitol 32, Li, J., H., M., Mitamuraa, T., and Horii, T. (2002) Mol. Biochem. Parasitol. 120, Banerjee, R., Liu, J., Beatty, W., Pelosof, L., Klemba, M., and Goldberg, D. E. (2002) Proc Natl Acad Sci U S A 99, Langreth, S. G., Jensen, J. B., Reese, R. T., and Trager, W. (1978) J. Protozool. 25, Aikawa, M. (1971) Exp. Parasitol. 30, Waller, R. F., Reed, M. B., Cowman, A. F., and McFadden, G. I. (2000) EMBO J. 19, Reed, M. B., Saliba, K. J., Caruana, S. R., Kirk, K., and Cowman, A. F. (2000) Nature 403, Wickham, M. E., Rug, M., Ralph, S. A., Klonis, N., McFadden, G. I., Tilley, L., and Cowman, A. F. (2001) EMBO J. 20, Fidock, D. A., and Wellems, T. E. (1997) Proc. Natl. Acad. Sci. USA 94, Ansorge, I., Benting, J., Bhakdi, S., and Lingelbach, K. (1996) Biochem. J. 315, Rug, M., Wickham, M. E., Foley, M., Cowman, A. F., and Tilley, L. (2003) Submitted 25. Coppel, R. L., Cowman, A. F., Lingelbach, K. R., Brown, G. V., Saint, R. B., Kemp, D. J., and Anders, R. F. (1983) Nature 306,

20 Cowman, A. F., Saint, R. B., Coppel, R. L., Brown, G. V., Favaloro, J., Crewther, P. E., Culvenor, J. G., Bianco, A. E., Stahl, H. D., Mitchell, G. F., Kemp, D. J., and Anders, R. F. (1985) in Modern Approaches to Vaccines of Molecular and Chemical Basis of Resistance to Viral, Bacterial and Parasitic Diseases - Vaccines 85 (Lerner, R., Chanock, R., and Brown, F., eds), pp , Cold Spring Harbor Laboratory, Cold Spring Harbor 27. Stenzel, D. J., and Kara, U. A. (1989) Eur. J. Cell. Biol. 49, Lauer, S., VanWye, J., Harrison, T., McManus, H., Samuel, B. U., Hiller, N. L., Mohandas, N., and Haldar, K. (2000) EMBO J. 19, Ward, G. E., Miller, L. H., and Dvorak, J. A. (1993) J. Cell Sci. 106, Delplace, P., Bhatia, A., Cagnard, M., Camus, D., Colombet, G., Debrabant, A., Dubremetz, J. F., Dubreuil, N., Prensier, G., Fortier, B., and et al. (1988) Biol. Cell 64, Deguercy, A., Hommel, M., and Schrevel, J. (1990) Mol. Biochem. Parasitol. 38, Dhawan, S., Dua, M., Chishti, A. H., and Hanspal, M. (2003) J Biol Chem 33. Nwagwu, M., Haynes, J. D., Orlandi, P. A., and Chulay, J. D. (1992) Exp. Parasitol. 75, Chulay, J. D., Lyon, J. A., Haynes, J. D., Meierovics, A. I., Atkinson, C. T., and Aikawa, M. (1987) J. Immunol. 139, Francis, S. E., Gluzman, I. Y., Oksman, A., Knickerbocker, A., Mueller, R., Bryant, M. L., Sherman, D. R., Russell, D. G., and Goldberg, D. E. (1994) EMBO J. 13, Moon, R. P., Tyas, L., Certa, U., Rupp, K., Bur, D., Jacquet, C., Matile, H., Loetscher, H., Grueninger-Leitch, F., Kay, J., Dunn, B. M., Berry, C., and Ridley, R. G. (1997) Eur. J. Biochem 244, Silva, A. M., Lee, A. Y., Gulnik, S. V., Maier, P., Collins, J., Bhat, T. N., Collins, P. J., Cachau, R. E., Luker, K. E., Gluzman, I. Y., Francis, S. E., Oksman, A., Goldberg, D. E., and Erickson, J. W. (1996) Proc. Natl. Acad. Sci. U S A 93, Ring, C. S., Sun, E., McKerrow, J. H., Lee, G. K., Rosenthal, P. J., Kuntz, I. D., and Cohen, F. E. (1993) Proc Natl Acad Sci U S A 90, Rosenthal, P. J., Wollish, W. S., Palmer, J. T., and Rasnick, D. (1991) J. Clin. Invest. 88, Rosenthal, P. J., Olson, J. E., Lee, G. K., Palmer, J. T., Klaus, J. L., and Rasnick, D. (1996) Antimicrob. Agents Chemother. 40, Rosenthal, P. J., Lee, G. K., and Smith, R. E. (1993) J. Clin. Invest. 91, Dominguez, J. N., Lopez, S., Charris, J., Iarruso, L., Lobo, G., Semenov, A., Olson, J. E., and Rosenthal, P. J. (1997) J Med Chem 40,

21 21 Figure 1: Models of merozoite escape from the host erythrocyte. (A) Primary rupture of the erythrocyte plasma membrane late in schizogony results in parasitophorous vacuole membrane enclosed merozoite structures (PEMS). Following a secondary extraerythrocytic rupture of the parasitophorous vacuole membrane, invasive merozoites are released. (B) Primary rupture of the parasitophorous vacuole membrane late in schizogony results in erythrocyte plasma membrane enclosed merozoite structures, and mixing of vacuolar and erythrocyte cytosolic contents. Following a secondary rupture of the erythrocyte plasma membrane, invasive merozoites are released. Figure 2: P.falciparum ruptures the parasitophorous vacuole within the host erythrocyte during asexual division. (A) Early 3D7-His schizonts traffic GFP (green) to the parasitophorous vacuole where it surrounds the fully formed daughter merozoites. Early 3D7+His schizonts traffic GFP into the host erythrocyte, and it is not observed in the vacuole surrounding the daughter merozoites. Parasite nuclei are stained with DAPI (blue). (C) Maturation of the schizont involves lysis of the parasitophorous vacuole; following vacuolar lysis, late 3D7-His schizonts display GFP fluorescence immediately surrounding the merozoites in addition to the erythrocyte cytosol (white arrow). Late 3D7+His schizonts traffic GFP beyond the vacuole into the host erythrocyte, and GFP fluorescence is observed immediately surrounding the daughter merozoites (black arrow). This localisation of GFP during early and late schizogony is represented schematically in B and D, respectively. (E) Examination of intraerythrocytic rupture of the parasitophorous vacuole at a population level. Parasites that

22 22 traffic GFP to the vacuole were fractionated before M phase (T, Trophozoites) and during schizogony (S). Upon saponin fractionation (Sap), GFP is detected in the pellet (P) and predominantly in the supernatant (Sn), indicating that GFP is exported from the parasite in both stages. Upon Streptolysin O fractionation (SLO) GFP is detected in the pellet in both stages, indicating that the destination of export is the vacuole. The presence of GFP in the SLO supernatant in schizonts indicates the presence of GFP in the erythrocyte cytosol during schizogony. Figure 3: Intraerythrocytic rupture of the parasitophorous vacuole membrane during schizogony of untransfected P.falciparum. Detection of S-Antigen by immunogold labelling with anti-s-antigen antibodies on ultra-thin sections of wild-type P.falciparum parasites. (A) In untreated early schizonts S-Antigen localizes to the parasitophorous vacuole. (B) In untreated late schizonts S-Antigen localizes throughout the erythrocyte cytosol. (C) In leupeptin and chymostatin treated early schizonts S-Antigen localizes to the parasitophorous vacuole. (D) In leupeptin and chymostatin treated late schizonts, S-Antigen localizes throughout the erythrocyte cytosol. Figure 4: Selective inhibition of the process of P.falciparum exit from the host erythrocyte. (A) Treatment of GFP expressing parasites with E-64 inhibits vacuolar, but not erythrocyte membrane, lysis resulting in PVM-enclosed merozoite structures (PEMS). 3D7-His, but not 3D7+His PEMS display GFP fluorescence (green). Parasite nuclei are stained with DAPI (blue). The merge of the 3 channels is shown on the right. These structures are represented schematically in B. (C)

23 23 Treatment of GFP expressing parasites with leupeptin and antipain inhibits erythrocyte, but not vacuolar membrane lysis. Both 3D7-His and 3D7+His clusters of merozoites display GFP fluorescence. This localisation of GFP is represented schematically in D. (E) Identification of limiting membrane in protease inhibitor treated P.falciparum parasites. Indirect immunofluorescence assay with anti-kahrp antibodies to determine the origin of the limiting membrane in E-64, and leupeptin and antipain treated parasites. Fixed blood smears were reacted with anti-kahrp antibodies (green). Parasite nuclei are stained with DAPI (blue).

24 Late Schizont A Primary rupture of erythrocyte Wickham et al. Figure 1 Merozoite Release Secondary rupture of parasitophorous vacuole membrane B Primary rupture of parasitophorous vacuole membrane Secondary rupture of erythrocyte plasma

25

26

27

28 Selective inhibition of a two-step egress of Malaria parasites from the host erythrocyte Mark E. Wickham, Janetta G. Culvenor and Alan F. Cowman J. Biol. Chem. published online July 11, 2003 Access the most updated version of this article at doi: /jbc.M Alerts: When this article is cited When a correction for this article is posted Click here to choose from all of JBC's alerts