NEW ROLES FOR PROTEASES IN THE INFECTION BIOLOGY OF COCCIDIAN PARASITES

Fabiola Parussini1, Chin Fen Teo1, Jill M. Harper1, Emily Binder2, Kami Kim2, Vern B. Carruthers1
1W. Harry Feinstone Department of Molecular Microbiology and Immunology, Johns Hopkins Bloomberg School of Public Health, 615 N. Wolfe Street, Baltimore, MD 21205 USA
e-mail: vcarruth@jhsph.edu
2 Departments of Medicine and of Microbiology and Immunology, Albert Einstein College of Medicine, 1300 Morris Park Ave, Bronx, NY 10461 USA

Summary
Coccidian parasites including Toxoplasma gondii cause widespread infections and disease in humans and animals. The need for new therapeutic agents along with the fascinating biology of these parasites has fueled a keen interest in understanding how key steps in the life cycle are regulated. Proteolysis is intimately associated with cell and tissue invasion by these obligate intracellular parasites and recent studies have begun to identify the proteases involved in these processes. Based on clues from inhibitor experiments and cleave site mapping studies, we and others are using emerging genome information and molecular genetics to identify and validate proteases that regulate secretory organelle biogenesis and invasion protein activity. These studies are revealing roles for a variety of proteases including cathepsins, subtilases, and rhomboids in cell and tissue invasion. The identification of highly selective inhibitors for these proteases has the potential to not only further dissect their role in infection but also to potentially prevent or ameliorate disease.

Introduction
Coccidian parasites are responsible for several important animal and human diseases. Neospora caninum is a leading cause of spontaneous abortion in bovines, leading to significant economic losses in the dairy industry (1). Eimeria infection of commercial livestock remains an important source of production losses (2), particularly in the poultry industry where high density facilities provide an ideal environment for rampant spread of disease. Toxoplasma gondii commonly infects animals and humans, causing birth defects, ocular disease, and neurological complications, particularly in immunocompromised individuals (3). Collectively these parasitic infections have multi-billion dollar yearly economic impact due to production losses and the medical care and institutionalization of afflicted people.
Although a variety of antibiotics are currently used to treat coccidian infections, these anti-coccidial compounds remain hampered by limitations. Eimeria drug resistance or tolerance is an ongoing problem in poultry production facilities. Although drug resistance in Toxoplasma infections is uncommon, front line treatments such as antifolate combination therapy (e.g., pyrimethamine and sulfadiazine) are often poorly tolerated or cause severe allergic reactions in patients. These limitations, combined with the fascinating properties of coccidian parasites, have accelerated interest in investigating mechanisms governing the infection biology of these pathogens.
Most of the attention has focused on unique aspects of coccidian parasites that are distinct from their host or other eukaryotes because these exceptional features and properties represent both potential targets for therapeutic intervention and opportunities to expand the boundaries of biological insight. Cell invasion is a unique and essential property of coccidian parasites that has received particularly intense attention because of its dramatic features. Typically complete within 10-20 seconds, coccidian zoites power their way into susceptible host cells by using a unique form of locomotion, gliding motility. Molecular interrogations of gliding motility and host cell invasion have recently revealed that proteases play central roles several distinct aspects of these processes. This article is intended to highlight new and emerging insight into the function of motility and invasion proteases in coccidian parasites.

Invasion organelles and proteins
Because of its experimental tractability, Toxoplasma has been most extensively investigated coccidian parasite for invasion studies. Like other coccidians, Toxoplasma actively invades host cells in a process that is largely parasite driven and distinct from the facilitated endocytosis mechanisms used by many intracellular bacteria and viruses. The parasite uses gliding motility to approach a target host cell and slide along its surface. Gliding motility is driven by the parasite’s intrapellicular actin-myosin motor system (4), which connects both with the inner membrane complex (5) and transmembrane adhesive proteins secreted from micronemes (6). Micronemes are small cigar-shaped secretory organelles that discharge their contents by fusion at the extreme apical end of the parasite (7). Basal secretion of micronemes occurs continuously as the parasite glides over surfaces by treadmilling microneme adhesive proteins from the apical tip to posterior pole. Upon reaching a suitable site for invasion, the parasite further relies of sequential discharge of secretory organelles to mediate entry (8). In the first event, microneme contents accumulate on the parasite apical surface where they bind host receptors during a step that coincides with the formation of the moving junction, a tight apposition of the parasite and host plasma membranes. Immediately thereafter, the contents of the club-shaped rhoptries are extruded through an apical duct where they translocate into the host cell. Underneath the site of invasion, rhoptry proteins and lipids reconfigure into small vesicles (evacuoles) that fuse with the parasitophorous vacuole (PV) as the parasite plunges into the cell (9). Similar to gliding motility, entry is powered by the actin-myosin depend translocation of microneme adhesins backward across the parasite surface to the posterior end where they accumulate in a “cap-like” structure before being shed during the last few seconds of penetration (10). Although the PV is created by pushing in the host plasma membrane, the moving junction excludes most of the host surface components and this is thought to be a key survival strategy for avoiding acidification and fusion with the endocytic or lysosomal systems. Within a few minutes of entry, a third class of secretory organelles called dense granules (DG) are released by fusion with the apical membrane. DG proteins extensively modify the PV and are thought to participate in nutrient acquisition. Although DGs resemble regulated dense core secretory granules of higher eukaryotes, their secretion is not as tightly controlled as micronemes and rhoptries. This, along with evidence that DGs are a “default” pathway through which proteins traffic in the absence of specific forward targeting signals (11), suggests that DGs are not classic regulated secretory organelles. Interestingly, among the three classes of secretory proteins associated with invasion, DG proteins are the only group that is not proteolytically processed. Thus, proteolysis is tightly associated with micronemes and rhoptries, the two pathways that are most intimately coupled with parasite invasion.

Understanding the role of proteases in cell and tissue invasion: from proteolytic events and inhibitors to candidate enzymes
Two main tacks have been used to uncover evidence of proteases involvement in coccidian invasion. One approach has been to identify proteolytically processed invasion proteins and map their cleavage sites. Although usually not definitive, cleavage site information can often provide clues about the type of protease involved based on the known recognition specificities of different classes of proteases. A second approach is the use of class specific protease inhibitors. For example, the serine protease inhibitors 3,4 dichloroisocumarin (3,4 DCI) and 4-(2-Aminoethyl) benzenesulphonyl fluoride (AEBSF) have been reported to block Toxoplasma invasion of human fibroblast cells (12). Other serine protease inhibitors similarly block Eimeria invasion (13,14). Cysteine proteases have also been implicated in invasion based on findings that cysteine protease inhibitor (PRT2253F) with activity against cathepsin B inhibits Toxoplasma cell entry (15). Although, limited by the possibility of non-specific “off target” effects, these studies at least provided initial evidence for the role of proteolysis in coccidian invasion. The identification and analysis of proteolytic substrates and the use of protease inhibitors have revealed that proteases function in several distinct steps of invasion (Fig. 1). This section will discuss how these approaches, along with the recent completion of the Toxoplasma genome, have recently led to the identification of candidate enzymes involved in each step.

Figure 1. Sites of invasion protein proteolysis
Microneme and rhoptry invasion proteins are subjected to a series of proteolytic processing steps including: (1) Proteolytic maturation en route to secretory organelles; (2) Mobilization to the parasite surface during parasite attachment; (3) Primary processing that trims microneme proteins while they are expressed on the parasite surface; and (4) Shedding by which the microneme products are eventually released from the parasite surface after they translocate toward the posterior end. Protease inhibitors that block each of these steps are listed.

Proteolytic maturation of invasion proteins
While performing pulse-chase metabolic labeling experiments, Achbarou (16) and Soldati (17) noted that nascent microneme (TgMIC3) and rhoptry (TgROP1) proteins are proteolytically processed within minutes of their initial translation. It is now appreciated that proteolytic maturation is a widespread phenomenon associated with most invasion proteins destined for secretion via the micronemes or rhoptries (Fig. 2). Such proteins are initially synthesized as preproproteins. The “pre” segment is the signal peptide, which is removed cotranslationally by signal peptidase during import into the endoplasmic reticulum. The “pro” peptide is subsequently removed as the protein transits through the secretory pathway. Rhoptry proteins appear to undergo processing in the nascent rhoptries during parasite cell division by endodyogeny (17,18). To investigate the sub-cellular site of microneme processing, we generated antibodies to the propeptides of TgMIC5 and TgM2AP. Immunofluorescent staining of extracellular or intracellular parasites revealed that these precursors occupy the trans-Golgi network and an early endosome compartment defined by co-staining with Rab51. No staining of mature micronemes was seen, however, some parasites showed partial localization of TgMIC5 within the DGs suggesting this may be an alternative route for secretion of immature microneme proteins. Collectively, our findings imply that microneme protein maturation occurs within or just beyond the early endosome, and not within in mature micronemes.
What roles do propeptides play in the biogenesis of secretory organelles? Although this remains an incompletely resolved question, recent evidence suggests that propeptides assist in the trafficking and regulation of their cognate proteins. For example, the TgROP1 propeptide and a segment containing the TgROP4 propeptide both supported the trafficking of heterologous proteins to the rhoptries (19,20). It should be noted, however, that TgROP1 can also use alternative sorting signals since a propeptide deletion mutant was still correctly targeted to the rhoptries (17). Also, based on analysis of a non-cleavable site directed mutant, processing of TgROP1 is not necessary for its trafficking or for rhoptry biogenesis (21). However, since ROP1 is not an essential protein (22) and only one of many that are trafficked to the rhoptries, it may not be ideal for assessing a general role of propeptides in organellar biogenesis. On the other hand, protease inhibitors are capable of interfering with the processing of multiple substrates cleaved by a single protease and therefore are expected to have more profound effects. Consistent with this notion, Shaw et al. (23) showed that subtilisin inhibitor III and cathepsin inhibitor III block parasite replication and cause marked abnormalities of secretory compartments including the rhoptries. Based in part on this study, Miller and coworkers (24) investigated a role for TgSUB2 in propeptide processing. TgSUB2 is a member of the subtilase family of serine proteases. Although Toxoplasma harbors at least 12 subtilase genes, most of these have not been investigated in any detail. TgSUB2 undergoes autocatalytic processing en route to the rhoptries and mutation of the natural autocleavage site revealed a key role for and acidic residue at the P1 position (cleavage site residues are designated P4-P3-P2-P1 / P1’-P2’-P3’-P4’, where / indicates the scissile site). Interestingly, an acidic P1 residue is also required for ROP1 processing and TgSUB2 co-immunoprecipitates with ROP1. Also, ROP2, 4, and 8 have similar putative cleavage sites, suggesting they are candidate substrates for TgSUB2. However, a cysteine protease of the cathepsin family, TgCPB (also called TgCP1 or Toxopain 1), has also been implicated in rhoptry protein processing based on its trafficking to the rhoptries and on the observations that a cathepsin B inhibitor partially blocks ROP2,3,4 processing, disrupts rhoptry biogenesis, and interferes with parasite invasion and infection (15,25). Thus, it appears that Toxoplasma expresses at least two distinct rhoptry maturases that participate in the processing of rhoptry proteins and in rhoptry biogenesis.
Propeptides appear to also play a central role in the trafficking and regulation of microneme proteins. TgMIC3 uses a short N-terminal propeptide to mask the carbohydrate binding activity of its lectin-like domain (26). This presumably prevents inappropriate binding to parasite glycoproteins within the secretory pathway. A propeptide deletion mutant (Dpro) of TgMIC3 also fails to reach the micronemes and is instead retained within the secretory pathway along with its partner protein TgMIC8 (M. Lebrun, personal communication). Similarly, we have shown that DproTgMIC5 and DproTgM2AP are retained within or near the early endosome, in addition to other sites. These findings indicate that microneme propeptides function in the trafficking of their cognate proteins, possibly by binding to cargo receptors. Interestingly, TgM2AP’s partner protein TgMIC2 also contains sorting signals in its C-terminal cytosolic domain (27) suggesting that multiple forward targeting elements are required for correct sorting to the micronemes. These signals may work at distinct sites, however, since deletion of TgMIC2 does not cause retention of TgM2AP but instead results in the misdirection of TgM2AP to the parasitophorous vacuole. Such findings are not universal though because deletion of the TgMIC6 propeptide has no effect on sorting of TgMIC6 or its partner proteins (TgMIC1 and TgMIC4) to the micronemes (28). To determine the type of protease involved in the proteolytic maturation of microneme proteins, we tested a series of protease inhibitors using the processing of nascent microneme proteins as an indicator of activity. Whereas serine or aspartyl protease inhibitors had no effect, cysteine protease inhibitors delayed the processing of TgM2AP and TgMIC3. Because cathepsin L inhibitor II was among the most effective inhibitors, we hypothesized that microneme proteins are processed by a cathepsin L-like enzyme. Analysis of the Toxoplasma genome database revealed that this parasite likely only has a single cathepsin L-like gene, termed TgCPL. TgCPL features a 30 kDa catalytic domain preceded by a proregion that includes a transmembrane anchor, which is a unique feature of apicomplexan cathepsin L proteases including the falcipain family in Plasmodium (29). After purification, recombinant proTgCPL undergoes autoactivation in vitro under mildly acidic conditions. Although definitive evidence is still forthcoming, several findings suggest a link between TgCPL and microneme protein maturation. First, the results of immunofluorescence and immunoelectron microscopy experiments suggest that TgCPL occupies a novel apical compartment intermediately positioned between the trans-Golgi network and the mature micronemes. Second, recombinant TgCPL can cleave recombinant proTgM2AP at or very near the correct cleavage site. Third, based on microarray screening of synthetic peptide substrate libraries, recombinant TgCPL shows a preference for leucine and structurally similar residues in the P2 position, which is consistent with the cleavage sites of TgM2AP, TgMIC3, TgMIC6, and TgAMA1, a microneme protein recently shown to be necessary for Toxoplasma invasion (30). Moreover, mutation of the TgM2AP P2 leucine residue to aspartic acid results in markedly less efficient cleavage and partial use of an alternative cleavage site 2 amino acids upstream of the normal cut site. Additional studies to further test the role of TgCPL in microneme protein maturation are underway. A parallel may also exist in Sarcocystis muris since this parasite expresses both a cathepsin L like enzyme (SmTP1; (31)) and a microneme protein (16/17 kDa antigen; (32)) that undergoes proteolytic maturation at a cleavage site with a P2 leucine.


Figure2. Selective proteolysis of Toxoplasma invasion proteins.
Microneme and rhoptry proteins are often multiply processed both within the secretory pathway and on the parasite surface whereas DG proteins are not subjected to postranslational processing.



Mobilization of invasion proteins
While screening a small library of cysteine protease inhibitors in an invasion assay, we noted that two compounds substantially impaired invasion while simultaneously disrupting microneme protein release. These compounds LHVS and ZL3VS both feature a vinyl sulfone warhead for electrophilic attack of the active site cysteine of a thiolprotease. LHVS and ZL3VS showed low micromolar dose-dependent inhibition of parasite entry in two different invasion assays, and had similar potency for blocking microneme protein release. We focused on LHVS for subsequent experiments because it was slightly more potent than ZL3VS. Using fluorescent differential 2-dimensional gel electrophoresis (2D-DIGE), we showed that LHVS impaired the release of a variety of microneme proteins but did not affect secretion of DG proteins or proteins released from other internal sites. LHVS also disrupted gliding motility since treated parasites fewer and shorter trails compared to solvent or control compound treated parasites. We initially reasoned that these compounds may block gliding and invasion by preventing the proteolytic shedding of microneme proteins from the parasite surface. However, the recent evidence that serine proteases of the rhomboid family are responsible for microneme protein shedding (see below) is inconsistent with this notion. Also, unlike the serine protease inhibitor 3,4 DCI, LHVS treatment did not result in the accumulation of microneme proteins such as TgMIC2 and TgM2AP on the parasite surface after stimulating secretion with a calcium agonist. Based on these findings, we conclude that a cysteine protease is required at an earlier step, perhaps facilitating the mobilization of microneme contents to the parasite surface. Interestingly, a fluorescent derivative of LHVS (Bodipy-LHVS) covalently labels a single 30 kDa band that immunoprecipitates with TgCPL antibodies. Labeling was block by pretreatment with LHVS, suggesting that the recognition is highly specific. Moreover, bodypy-LHVS illuminates the same apical compartment occupied by TgCPL. Whether TgCPL’s role in the mobilization of microneme proteins is related to its putative function as a maturase remains under investigation.

Surface trimming (primary processing)
After being discharged from the micronemes, several invasion proteins are subjected to proteolytic cleavages that do not affect their association with the parasite surface. These events are termed primary processing or “trimming” because they occur before microneme proteins are proteolytically liberated from the surface (see below). Amino or carboxy terminal peptides of low structural complexity are often the targets of trimming, which often occurs at multiple sites in the same substrate. For example, an amino-terminal peptide extending from the globular A-domain of TgMIC2 is removed with with at least three endoproteolytic cleavages by microneme protein protease 2 (MPP2) activity (33). This processing was recently shown to activate the TgMIC2 A-domain for binding to ICAM1, which the parasite uses in part to traverse cell barriers to reach deep tissues where it replicates (34). Therefore, disrupting MPP2 activity is predicted to interfere with the pathogenesis of infection. MPP2 also cleaves TgMIC4 near its carboxy terminus and it trims off the carboxy terminal “coiled” domain from TgM2AP, along with another putative proteolytic activity MPP3 (35). Although all of the MPP2 cleavage sites for TgMIC2 and TgM2AP have been defined, this information was not particularly revealing, apart from suggesting that MPP2 prefers small to medium sized uncharged amino acids in the P1-P4 sites. Whereas MPP3 activity is resistant to all of the inhibitors tested, MPP2 activity is blocked by the tripeptide aldehyde compounds ALLN and ALLM and by the serine protease inhibitors chymostatin and PMSF (35). Although it was recently proposed (36) that surface trimming is responsible for activating adhesive proteins for tight binding to host receptors, little evidence exists to support this idea. Critical to testing this hypothesis is the identification and characterization of the protease(s) responsible for MPP2 and MPP3 activities. Intriguing new findings from collaborative studies in between our labs (Carruthers and Kim) suggest a potential breakthrough in this quest.
While investigating phenotypic changes stemming from the targeted deletion of the TgSUB1 gene, we noted from western blots that the trimming of TgMIC2, TgM2AP, and TgMIC4 was markedly diminished in the TgSUB1 knockout (KO) parasites compared to a control parasite line. This microneme derived subtilase uses a glycosylphosphatidyl inositol (GPI) anchor to transiently occupy the parasite surface before being proteolytically shed into the culture supernatant where it becomes a major component of the excreted/secreted antigen (ESA) fraction (37). To more widely examine changes in the processing of microneme proteins, we performed 2D-DIGE on ESA fractions collected from control parasites and TgSUB1KO. This analysis vividly showed the near complete absence of TgMIC2, TgM2AP, and TgMIC4 processing products, with the corresponding accumulation of precursor species. Since this pattern closely resembles that of ESAs collected from ALLN treated parasites (35), we tentatively conclude that TgSUB1 is MPP2. Although ALLN was initially reported to be a selective inhibitor of calpains (calcium dependent cysteine proteases), this compound also has activity against some serine proteases including the proteosome. Also, chymostatin and PMSF inhibition of MPP2 is also consistent with it being TgSUB1. Since subtilases are often activated by high calcium concentrations, this may be how TgSUB1 is regulated upon reaching the parasite surface and the extracellular environment. Intriguingly, preliminary mouse infection experiments suggest that TgSUB1KO parasites are moderately attenuated in virulence, with some mice surviving a normally lethal infectious dose. Further studies will be necessary to determine whether this is due to an effect on cell entry, tissue invasion, or both. Regardless, these studies may have wider implications since orthologous subtilases are expressed a variety of apicomplexans including Neospora (38,39) and Plasmodium (40).

Surface shedding (secondary processing)
In contrast to surface antigens (SAGs) which continuously occupy the parasite surface, invasion proteins derived from the microneme are only transiently associated with the parasite plasma membrane. Consequently, steady state levels of microneme proteins on extracellular tachyzoites are generally low. Microneme proteins are most readily detected on the parasite surface during invasion, where they can be seen accumulating on the extracellular portion of the parasite as they translocate backwards driven by the actin-myosin motor system. Our early studies on TgMIC2 indicated that this protein was shed into the culture supernatant as a smaller species that was devoid of its carboxy-terminal cytosolic domain and transmembrane anchor. This finding coupled with the observation that TgMIC2 is not seen on intracellular parasites suggested that microneme proteins are proteolytically released from the parasite surface during invasion. MPP1, the hypothetical protease responsible for shedding is unaffected by a wide range of protease inhibitors (33), suggesting it may be an unusual enzyme. This notion was subsequently corroborated when Dominique Soldati’s group demonstrated that TgMIC6 is cleaved within its transmembrane anchor near the extracellular interface. Although this suggested the existence of an intramembranous protease, David Sibley’s group simultaneously reported that mutation of two lysine residues outside the transmembrane anchor abolished shedding and disrupted invasion (41). To address this apparent discrepancy, Zhou et al. (35) used mass spectroscopy to determine that TgMIC2 is also cleaved intramembranously at a site precisely corresponding to that of TgMIC6. Consolidating these findings, it appears likely that while cleavage occurs within the transmembrane anchor, sequences outside of the anchor are required for protease recognition or for supporting a favorable structural configuration for cleavage.
Intramembrane proteolysis is a recently described phenomenon performed by integral membrane proteases that usually do not cleave until another protease has processed the substrate at another site (42). However, inhibition of MPP2 processing of TgMIC2 had no effect on its shedding, implying that MPP1 does not require a primary processing event. This clue helped focus attention on the rhomboid family of intramembrane serine proteases, which can cleave their substrates without prior processing. Also, rhomboids typically cleave near the extracellular portion of the transmembrane anchor in a region populated by small, helix breaking amino acids such as alanine and glycine (43), properties that are consistent with the TgMIC2 and TgMIC6 cleavage sites. Similar highly conserved putative cleavage sites are seen a multitude of transmembrane microneme proteins expressed by Neospora, Eimeria, and Plasmodium, among other apicomplexans (44). Rhomboids are widely present throughout the phylum (44). Heterologous expression of fusion proteins with microneme transmembrane anchors showed susceptibility to cleavage by human and drosophila rhomboids (43). Analysis of the Toxoplasma genome revealed the presence of six rhomboid-like genes (ROM1, 2, 3, 4, 5, and 6) (44-46). Since ROM6 is highly homologous to a mitochondrial rhomboid involved in organellar fusion, this was eliminated as a candidate for MPP1. Also, ROM3 is not expressed in tachyzoites, reducing the likelihood that it encodes MPP1. Localization studies of the remaining candidates revealed that ROM1 is expressed in the micronemes, ROM2 is in the Golgi, ROM4 occupies the entire parasite surface, while ROM5 is most abundant on the posterior surface (45,46). Since MPP1 is constitutively active on the parasite surface (33,47), ROM4 and ROM5 are currently the best candidates. While ROM5 is the only ROM capable of cleaving a full length fusion protein of TgMIC2 in a heterologous expression system (45), this protease is not as well conserved among the Apicomplexa as is ROM4 (46). Determining precisely which ROM is MPP1 will probably require conditional expression experiments.

Conclusions
Building on data from protease inhibitor and cleavage site mapping, we and others are using the effectively complete genome sequences of Toxoplasma and other related parasites to begin matching proteolytic events with the associated enzymes. Emerging insight from these developments suggest that proteases play distinct roles in the trafficking, mobilization, and regulation of invasion proteins. Cysteine proteases (TgCPB and TgCPL) along with the subtilase TgSUB2 appear to participate in the proteolytic maturation of microneme and rhoptry substrates during the biogenesis of these organelles. Additionally, TgSUB1 and integral membrane protease of the rhomboid family function on the parasite surface where they regulate proteins involved in cell entry and tissue invasion. Future challenges not only include uncovering a deeper understanding of the biological roles of coccidian proteases but also the identification of selective inhibitors designed to interfere with their function for therapeutic gain.


Literature Cited
1. Waldner, C. L., Janzen, E. D., and Ribble, C. S. (1998) J Am Vet Med Assoc 213, 685-690

2. Ruff, M. D. (1999) Vet Parasitol 84, 337-347

3. Montoya, J. G., and Liesenfeld, O. (2004) Lancet 363, 1965-1976

4. Dobrowolski, J. M., and Sibley, L. D. (1996) Cell 84, 933-939

5. Gaskins, E., Gilk, S., DeVore, N., Mann, T., Ward, G., and Beckers, C. (2004) J Cell Biol 165, 383-393

6. Jewett, T. J., and Sibley, L. D. (2003) Mol Cell 11, 885-894

7. Carruthers, V. B., and Sibley, L. D. (1999) Mol Microbiol 31, 421-428

8. Carruthers, V. B., and Sibley, L. D. (1997) Eur J Cell Biol 73, 114-123

9. Hakansson, S., Charron, A. J., and Sibley, L. D. (2001) Embo J 20, 3132-3144

10. Carruthers, V. B., Giddings, O. K., and Sibley, L. D. (1999) Cell Microbiol 1, 225-235

11. Karsten, V., Qi, H., Beckers, C. J., Reddy, A., Dubremetz, J. F., Webster, P., and Joiner, K. A. (1998) J Cell Biol 141, 1323-1333

12. Conseil, V., Soete, M., and Dubremetz, J. F. (1999) Antimicrob Agents Chemother 43, 1358-1361

13. Adams, J. H., and Bushell, G. R. (1988) Int J Parasitol 18, 683-685

14. Fuller, A. L., and McDougald, L. R. (1990) J Parasitol 76, 464-467

15. Que, X., Ngo, H., Lawton, J., Gray, M., Liu, Q., Engel, J., Brinen, L., Ghosh, P., Joiner, K. A., and Reed, S. L. (2002) J Biol Chem 277, 25791-25797

16. Achbarou, A., Mercereau-Puijalon, O., Autheman, J. M., Fortier, B., Camus, D., and Dubremetz, J. F. (1991) Mol Biochem Parasitol 47, 223-233

17. Soldati, D., Lassen, A., Dubremetz, J. F., and Boothroyd, J. C. (1998) Mol Biochem Parasitol 96, 37-48

18. Carey, K. L., Jongco, A. M., Kim, K., and Ward, G. E. (2004) Eukaryot Cell 3, 1320-1330

19. Bradley, P. J., and Boothroyd, J. C. (2001) Int J Parasitol 31, 1177-1186

20. Bradley, P. J., Li, N., and Boothroyd, J. C. (2004) Mol Biochem Parasitol 137, 111-120

21. Bradley, P. J., Hsieh, C. L., and Boothroyd, J. C. (2002) Mol Biochem Parasitol 125, 189-193

22. Kim, K., Soldati, D., and Boothroyd, J. C. (1993) Science 262, 911-914

23. Shaw, M. K., Roos, D. S., and Tilney, L. G. (2002) Microbes Infect 4, 119-132

24. Miller, S. A., Thathy, V., Ajioka, J. W., Blackman, M. J., and Kim, K. (2003) Mol Microbiol 49, 883-894

25. Que, X., Wunderlich, A., Joiner, K. A., and Reed, S. L. (2004) Infect Immun 72, 2915-2921

26. Cerede, O., Dubremetz, J. F., Bout, D., and Lebrun, M. (2002) Embo J 21, 2526-2536

27. Di Cristina, M., Spaccapelo, R., Soldati, D., Bistoni, F., and Crisanti, A. (2000) Mol Cell Biol 20, 7332-7341

28. Reiss, M., Viebig, N., Brecht, S., Fourmaux, M. N., Soete, M., Di Cristina, M., Dubremetz, J. F., and Soldati, D. (2001) J Cell Biol 152, 563-578

29. Rosenthal, P. J. (2004) Int J Parasitol 34, 1489-1499

30. Mital, J., Meissner, M., Soldati, D., and Ward, G. E. (2005) Mol Biol Cell

31. Hansner, T., Freyer, B., Mehlhorn, H., and Ruger, W. (1998) Parasitol Res 84, 578-582

32. Klein, H., Mehlhorn, H., and Ruger, W. (1996) Parasitol Res 82, 468-474

33. Carruthers, V. B., Sherman, G. D., and Sibley, L. D. (2000) J Biol Chem 275, 14346-14353

34. Barragan, A., Brossier, F., and Sibley, L. D. (2005) Cell Microbiol 7, 561-568

35. Zhou, X. W., Blackman, M. J., Howell, S. A., and Carruthers, V. B. (2004) Mol Cell Proteomics 3, 565-576

36. Carruthers, V. B., and Blackman, M. J. (2005) Mol Microbiol 55, 1617-1630

37. Miller, S. A., Binder, E. M., Blackman, M. J., Carruthers, V. B., and Kim, K. (2001) J Biol Chem 276, 45341-45348

38. Louie, K., Nordhausen, R., Robinson, T. W., Barr, B. C., and Conrad, P. A. (2002) J Parasitol 88, 1113-1119

39. Louie, K., and Conrad, P. A. (1999) Mol Biochem Parasitol 103, 211-223

40. Blackman, M. J., Fujioka, H., Stafford, W. H., Sajid, M., Clough, B., Fleck, S. L., Aikawa, M., Grainger, M., and Hackett, F. (1998) J Biol Chem 273, 23398-23409

41. Brossier, F., Jewett, T. J., Lovett, J. L., and Sibley, L. D. (2003) J Biol Chem 278, 6229-6234

42. Brown, M. S., Ye, J., Rawson, R. B., and Goldstein, J. L. (2000) Cell 100, 391-398

43. Urban, S., and Freeman, M. (2003) Mol Cell 11, 1425-1434

44. Dowse, T. J., and Soldati, D. (2005) Trends Parasitol 21, 254-258

45. Brossier, F., Jewett, T. J., Sibley, L. D., and Urban, S. (2005) Proc Natl Acad Sci U S A 102, 4146-4151

46. Dowse, T. J., Pascall, J. C., Brown, K. D., and Soldati, D. (2005) Int J Parasitol 35, 747-756

47. Opitz, C., Di Cristina, M., Reiss, M., Ruppert, T., Crisanti, A., and Soldati, D. (2002) Embo J 21, 1577-1585