Research Article 947 Microneme protein 8 – a new essential invasion factor in

Henning Kessler1, Angelika Herm-Götz1, Stephan Hegge1, Manuel Rauch1, Dominique Soldati-Favre2, Friedrich Frischknecht1 and Markus Meissner1,* 1Hygieneinstitute, Department of Parasitology, University Hospital Heidelberg, Im Neuenheimer Feld 324, D-69120 Heidelberg, Germany 2Department of and Molecular Medicine, University of Geneva, CMU, 1, rue Michel-Servet 1211, Geneva 4, Switzerland *Author for correspondence (e-mail: [email protected])

Accepted 7 January 2008 Journal of Cell Science 121, 947-956 Published by The Company of Biologists 2008 doi:10.1242/jcs.022350

Summary Apicomplexan parasites rely on sequential secretion of furthermore demonstrate that the cytosolic domain is crucial specialised secretory for the invasion of the host cell. for the function of MIC8 and can not be functionally First, micronemes release their content upon contact with the complemented by any other micronemal protein characterised host cell. Second, are discharged, leading to the so far, suggesting that MIC8 represents a novel, functionally formation of a tight interaction (moving junction) with the host distinct invasion factor in this apicomplexan parasite. cell, through which the parasite invades. The functional characterisation of several micronemal proteins in Toxoplasma gondii suggests the occurrence of a stepwise process. Here, we Supplementary material available online at show that the micronemal protein MIC8 of T. gondii is essential http://jcs.biologists.org/cgi/content/full/121/7/947/DC1 for the parasite to invade the host cell. When MIC8 is not present, a block in invasion is caused by the incapability of the Key words: , Invasion, Microneme, , Tet system, parasite to form a moving junction with the host cell. We Toxoplasma

Introduction secretion and subsequently to the formation of the MJ are not well The phylum Apicomplexa consists of several pathogens of immense understood. The functional linkage between micronemes and medical and veterinary importance, such as those causing malaria, rhoptries led us speculate that additional micronemal proteins might toxoplasmosis or coccidiosis. Because they are obligate intracellular be responsible for the establishment of the MJ. parasites, a key step during their life cycle is the invasion of the Micronemal protein 8 (MIC8) was previously identified by

Journal of Cell Science host cell. Apicomplexan parasites actively invade the host cell in sequence similarity to other micronemal proteins in the genome of a complex, organised, stepwise process (Carruthers and Boothroyd, Toxoplasma gondii. Analysis of MIC8 has demonstrated that it is, 2006) that involves specialised secretory organelles (the like other micronemal transmembrane proteins, secreted upon an micronemes, rhoptries and dense granules) and a unique form of increase in intracellular calcium concentration and processed close movement termed gliding motility (Soldati and Meissner, 2004). to its transmembrane domain (TMD). Furthermore, it has been Upon initial attachment to a surface, the parasite starts to glide in demonstrated that MIC8 interacts with the soluble micronemal a substrate-dependent manner. Gliding involves the discharge of protein MIC3 during the transit through the secretory pathway adhesive transmembrane proteins at the apical end by specialised (Meissner et al., 2002a). Here, employing a conditional gene- secretory organelles called micronemes. It has previously been expression system, we present a detailed functional characterisation demonstrated that micronemal proteins of the TRAP family of this essential protein (Meissner et al., 2002b). We show that (thrombospondin-related anonymous protein) are essential for invasion depends on the function of MIC8. Further detailed gliding motility and invasion (Huynh and Carruthers, 2006; Sultan characterisation of the resulting phenotype showed that this protein et al., 1997) by binding to host-cell receptors and bridging them to does not contribute to gliding motility. Instead it appears that MIC8 the cytoplasmic actin-myosin motor of the parasite via the glycolytic plays an essential role in a step before the formation of an MJ. We enzyme aldolase (Buscaglia et al., 2003; Jewett and Sibley, 2003; show, in a novel complementation strategy, that the cytosolic Starnes et al., 2006). The next step in invasion is the recognition domains (CTDs) of MIC2, PfTRAP or AMA1 cannot substitute the of the host cell followed by rhoptry secretion and intimate function of the MIC8 CTD, whereas the CTD of the homologous attachment to the host cell. A conditional-mutagenesis approach for protein MIC8-like1 restores function. This indicates that this the micronemal protein AMA1 has demonstrated a distinct role for transmembrane protein is involved in a novel invasion step leading a micronemal protein during invasion. In the absence of AMA1, to formation of the MJ. the parasite fails to form a moving junction (MJ) (Mital et al., 2005). The formation of the MJ depends on the secretion of rhoptry neck Results proteins (RONs) and their interaction with AMA1 (Alexander et Establishment of inducible conditional MIC8 knockout al., 2005; Lebrun et al., 2005). In the absence of AMA1, the RONs We employed a tetracycline-inducible system to generate a are still secreted and associate with the host cell in loose patches conditional knockout mutant for mic8. Parasites expressing the but do not form an MJ (Alexander et al., 2005; Boothroyd and tetracycline-dependent transactivator TATi1 (Meissner et al., 2002b) Dubremetz, 2008). Up until now, the events leading to rhoptry were transfected with a construct encoding a recombinant Ty-tagged 948 Journal of Cell Science 121 (7)

version of MIC8 (MIC8Ty) under control of the Tet-inducible ATc for the same amount of time, as judged from immunoblots promoters pT7S1 and pT7S4 (Meissner et al., 2001). We generated (Fig. 2A,B and see later). Unless otherwise indicated, parasites were three independent clones expressing mic8Ty at different levels when grown in the absence or presence of ATc for 48 hours before the grown in the absence of anhydrotetracycline (ATc) (Fig. 1A). The respective experiment was performed. We confirmed that the strongest-expressing clone (T7S4-5) showed a partial localisation observed downregulation in mic8 expression was specific, because of MIC8 in an apically localised compartment that corresponded the expression levels of other micronemal proteins (MIC2 and to early endosomes (Fig. 1B, Fig. 2D). By contrast, parasite strains MIC3) or of alpha-tubulin were not affected (Fig. 2B). expressing the transgene at a lower level showed a normal localisation of MIC8Ty in the micronemes. Interestingly, prolonged MIC8 is not required for trafficking of MIC3 to the micronemes culture in the absence of ATc resulted in loss of transgene expression, Previously, we found that MIC8 and MIC3 are capable of interacting indicating that overexpression of MIC8 is unfavourable for the with each other during the transit through the secretory pathway, parasite (data not shown). and we suggested that MIC8 could act as an escorter for MIC3 We removed endogenous mic8 in transgenic parasites (T7S1-4 (Meissner et al., 2002a), in analogy to other identified micronemal and T7S4-1) via homologous recombination with the construct escorter proteins, such as MIC6 (Reiss et al., 2001) and MIC2 MIC8KOCAT (see Materials and Methods) (Fig. 1C) and obtained (Rabenau et al., 2001). To whether MIC8 is required for correct two independent clonal parasite populations, as shown in analytical localisation of MIC3, we analysed localisation of MIC3 in parasites PCRs on genomic DNA (Fig. 1D). Both clones, MIC8KOi-T7S1- depleted for MIC8. Surprisingly, we did not observe a 4 and MIC8KOi-T7S4-1, showed the same behaviour in phenotypic mislocalisation or an altered expression level of MIC3 (Fig. 2B,C). assays (data not shown). We chose MIC8KOi-T7S1-4 for detailed Furthermore, we did not detect any accumulation of MIC3 in early phenotypic analysis and refer to this parasite strain as MIC8KOi. endosomes when parasites overexpressing MIC8 were analysed We analysed the expression of mic8Ty in MIC8KOi. Parasites (Fig. 2D). grown in the absence of ATc showed almost identical levels of Based on these results, we conclude that the interaction between expression of MIC8 (~75%, as judged from immunoblots) compared MIC8 and MIC3 is not essential for targeting of MIC3 to the to wild-type parasites (Fig. 2A,B). Incubation of parasites for 48 micronemes. To confirm that the observed accumulation of MIC8 hours in the presence of ATc led to almost complete ablation of in T7S4-5 corresponds to the early endosomes, we performed co- MIC8Ty (to ~2%) compared with parasites grown in absence of localisation analysis with the Golgi marker GRASP-RFP (Pfluger Journal of Cell Science

Fig. 1. Establishment of a conditional knockout for MIC8. (A) Regulation of mic8Ty expression by ATc in three independent parasite strains: T7S4-1, T7S4-5 and T7S1-4. Parasites were grown with (+) or without (–) ATc for 48 hours, fixed and stained with antibodies against Ty, or against MIC6 as a control. Pictures were taken under identical exposure conditions. (B) Immunofluorescence analysis of parasite strains T7S4-1 and T7S4-5 grown for 48 hours in the absence of ATc, as shown in A. Parasites were stained with the indicated antibodies. Strong overexpression of MIC8Ty results in partial accumulation of MIC8 within the secretory pathway (arrows). Other micronemal proteins, such as MIC6, appear not to be affected by overexpression of MIC8. (C) Homologous recombination of endogenous mic8 with a knockout construct in which a CAT-expression cassette is flanked by 1.6 kb (5′FS) and 2.5 kb (3′FS) results in a knockout locus that can be identified using analytical PCR with specific oligonucleotides, as indicated by the arrows. (D) Analytical PCR analysis to verify homologous recombination of the knockout construct in the recipient strains T7S1-4 and T7S4-1. Numbers on top of each lane indicate the respective oligonucleotide combination (see C) used in each PCR reaction. Scale bars: 25 μm (A); 5 μm (B). Role of MIC8 in T. gondii invasion 949

Phenotypic characterisation of MIC8KOi To analyse the function of MIC8 in detail, we inoculated human foreskin fibroblast (HFF) cells with parasites in the presence and absence of ATc, and followed the propagation of parasites. Whereas MIC8KOi parasites grown in the absence of ATc showed normal growth when compared to wild-type parasites, no growth was seen when the expression of mic8 was ablated (Fig. 3A). When we examined infected host cells within 4 days, we occasionally detected huge parasitophorous vacuoles containing large numbers of parasites, indicating that parasites were capable of replicating in the absence of MIC8 (see also supplementary material Fig. S1). Occasionally, we observed lysed vacuoles with freshly egressed motile parasites; these parasites appeared to be incapable of invading neighbouring host cells (Fig. 3A). Several days after the inoculation of cells with parasites, we failed to detect any intracellular parasites, indicating that initially infected host cells had been lysed and that released parasites were incapable of re- invading new host cells (Fig. 3B and data not shown).

Intracellular replication and host-cell egress are not affected by the absence of MIC8 To analyse the function of MIC8 during the asexual cycle of T. gondii, we generated MIC8KOi parasites constitutively expressing GFP. We inoculated HFF monolayers with equal numbers of freshly released MIC8KOi parasites that were grown in the absence of ATc, and subsequently compared growth of the parasites and lysis of initially infected host cells in presence or absence of ATc over a 72-hour time-course (Fig. 3B). We did not see any difference in non-induced host-cell egress: the number of initially infected host cells decreased irrespective of whether ATc was added to the medium or not. Whereas, at 24 hours post-infection, approximately 90% of initially infected host cells were intact, these numbers decreased to approximately 55% and 10% after 48 and 72 hours, respectively (Fig. 3B). However, under conditions of MIC8

Journal of Cell Science depletion, no subsequent infection of neighbouring host cells was detected. This indicates that this protein is specifically required Fig. 2. MIC8 is not required for trafficking of MIC3 to the micronemes. during host-cell invasion, whereas egress is not affected in parasites (A) Immunofluorescence analysis of MIC8KOi grown for 48 hours in the depleted of MIC8. presence (+) or absence (–) of ATc prior to fixation. Staining was performed with alpha-MIC8, or with alpha-MIC2 as a control (same exposure time for all Host-cell exit can be artificially triggered by the use of the micrographs). (B) Immunoblot analysis of MIC8KOi and RH parasites grown calcium-ionophore A23187 (Black et al., 2000). When we treated in the presence or absence of ATc for 48 hours and probed with the indicated intracellular MIC8KOi parasites grown in the presence or absence antibodies. No significant difference in the expression levels of MIC2, TUB1 of ATc with A23187 for 5 minutes, no significant reduction in egress or MIC3 could be detected in MIC8KOi cells under the two conditions. was detected (supplementary material Fig. S1). (C) Localisation of MIC3 in MIC8KOi parasites grown in the presence of ATc for 48 hours. Parasites were stained with the indicated antibodies. Merged images are shown on the right. Blue stain indicates DAPI (nuclei). (D) Co- Depletion of MIC8 does not affect parasite attachment or localisation studies in the parasite strain T7S4-5, which overexpresses microneme secretion MIC8Ty (see also Fig. 1B). Upper panels: co-localisation study of MIC8Ty Initial attachment is the first step in the invasion of the host cell, and GRASP-RFP (Pfluger et al., 2005) in the parasite strain T7S4-5 transiently transfected with GRASP-RFP demonstrates accumulation of MIC8 in a post- so we performed attachment assays on fixed host cells. MIC8KOi, Golgi compartment. Middle panels: co-localisation of MIC8Ty and the TATi1-expressing and wild-type (RH) parasites grown in the propeptide of M2AP (Harper et al., 2006) in early endosomes. Lower panels: presence or absence of ATc were allowed to attach to the host cell co-localisation study of MIC8Ty and MIC3 shows that MIC3 is not localised for 15 minutes. After removal of non-attached parasites, the in early endosomes because of overexpression of MIC8Ty. Right: enlargement number of parasites associated with the host cell was compared of boxed parasite(s) in the merged pictures. Scale bars, 10 μm. between these groups. We found a slight reduction in attachment when parasites expressing TATi1 were compared with wild-type parasites. Efficiency in attachment appeared to be further decreased in the case of MIC8KOi, irrespective of whether MIC8 et al., 2005) and found that the majority of retained MIC8 localised was expressed (Fig. 3C). The observed reduction in attachment apical to the Golgi (Fig. 2D). When we performed co-localisation could be due to the extensive genetic modifications of the parental analysis of MIC8 with an antibody against the propeptide of M2AP, RH parasites, which required three independent stable which localises to early endosomes (Harper et al., 2006), we detected transfections in order to generate MIC8KOi. However, no retained MIC8 within the same compartment (Fig. 2D). difference in initial host-cell attachment was observed in the case 950 Journal of Cell Science 121 (7)

Fig. 3. MIC8 is required for parasite survival and is essential during the invasion of the host cell. (A) Upper panels: growth of MIC8KOi parasites was compared to RHwt parasites using plaque assay. Parasites were inoculated on HFF cells in the presence or absence of ATc for 7 days prior to GIEMSA staining. Lower panels: same experiment as above. Parasites were analysed 72 hours post- infection. Note that in the case of MIC8KOi grown in the presence of ATc, freshly egressed parasites did not invade neighbouring host cells. (B) Analysis of non-induced egress by quantification of intact initially infected host cells. HFF cells were inoculated with MIC8KOi-GFP parasites in the presence or absence of ATc and the decrease in initially infected host cells was analysed over time. The depicted quantification is representative of the results derived from three independent experiments. (C) Attachment assay of freshly released parasites grown in the presence or absence of ATc to fixed cells. Equal numbers of parasites were allowed to attach on host cells at 37°C for 15 minutes before non-attached parasites were removed by consecutive washing steps with cold PBS. Data are mean values of three independent experiments ± s.d. (D) Invasion assay of freshly egressed parasites grown in the presence or absence of ATc as in C. After the indicated time, non-invaded parasites were removed by several washing steps. Intracellular parasites were further incubated for 12 hours and the number of parasitophorous vacuoles was compared. ON, over night. Data shown are mean values of three independent experiments ± s.d. (E) Re- expression of MIC8 in extracellular parasites restores invasion capability. Invasion assay was performed similar to as in D. After MIC8KOi grown in the presence or absence of ATc were allowed to attach for 10 minutes, non-attached parasites were removed. Attached parasites were further incubated in the presence or absence of ATc (–/–: parasites constantly kept in the absence of ATc; +/+: parasites constantly kept in the presence of ATc; +/–: parasites kept in the presence of ATc and further incubated in media without ATc). Data represent mean values of three independent experiments ± s.d.

of MIC8KOi parasites grown in the presence or absence of ATc. We were interested to know whether this protein is involved in

Journal of Cell Science Therefore, MIC8 appears to be not involved in this invasion step a signalling cascade during egress of the host cell that prepares the (Fig. 3C). parasite for the next round of invasion, because egress and invasion To exclude the possibility that secretion of micronemal proteins are tightly linked (Hoff and Carruthers, 2002). If this were the case, was affected, we compared the level of secreted MIC2. We found extracellular parasites expressing only nascent MIC8 would be no difference in secretion of MIC2 in parasites grown in the presence incapable of invasion, because no modification of MIC8 can occur or absence of ATc. Both constitutive microneme secretion as well during egress. In order to differentiate between a role of this protein as induced secretion showed no difference to wild type in the during egress and invasion, we modified the above experiment. HFF absence of MIC8 (supplementary material Fig. S1). Similar results cells were inoculated for 10 minutes with parasites grown in the were obtained with other micronemal proteins, such as MIC6 and presence of ATc. Non-attached parasites were then removed. MIC4 (data not shown). Subsequently, expression of mic8Ty was reactivated in attached extracellular parasites by the removal of ATc for 12 hours before MIC8 is required for invasion and not during egress the invasion of host cells was analysed. Reactivation of MIC8 Next, HFF cells were inoculated with MIC8KOi parasites in the expression in extracellular parasites restored the capability of presence or absence of ATc. After host-cell lysis, equal numbers of invasion (up to 20%) when compared with MIC8KOi parasites that freshly egressed parasites were analysed for their capability to were constantly kept in absence of ATc (Fig. 3E). These results invade host cells. Parasites were allowed to invade for different indicate that MIC8 is exclusively involved in invasion and is not durations before the removal of non-invaded parasites and were required or modified during egress. As soon as a critical level of further incubated in the presence or absence of ATc. We found that MIC8 was reached in extracellular parasites, invasion competency parasites depleted of MIC8 were completely deficient in invasion. was restored. The observed reduction in invasion, when compared Even a prolonged invasion time of more then 12 hours did not to parasites constantly grown in the absence of ATc (Fig. 3E), is increase invasion events significantly (Fig. 3D). A direct comparison probably due to reduced viability of parasites kept for a prolonged between MIC8KOi parasites grown in the presence or absence of time outside of the host cell. ATc showed that less then 0.1% of the parasites were invasive. To confirm that MIC8 is a general, essential factor for host-cell MIC8 is dispensable for gliding motility invasion, we performed analogous assays on different host-cell types Gliding motility is a prerequisite for active invasion of the host (green monkey kidney cells, HeLa cells and macrophages) and cell. Several types of movement can be detected in wild-type obtained identical results (data not shown). parasites (Fig. 4A) (Hakansson et al., 1999). Recently, it has been Role of MIC8 in T. gondii invasion 951

Fig. 4. MIC8 is not involved in gliding motility. (A) Three different forms of movement can be identified by quantitative image analysis: circular gliding (C), helical gliding (H) and twirling (T). These forms give distinct patterns when the acquired images are projected. The projection (proj.) of 100 images acquired at 3- second intervals represents an overall observation time of 5 minutes. All three forms of movements can be detected in the case of RHwt and MIC8KOi grown in the presence or absence of ATc. (B) Comparison of overall motility. Although differences between RH and MIC8KOi parasites can be observed (~30 vs 12% overall gliding motility, respectively), no significant reduction in gliding motility can be observed between MIC8KOi grown in the presence or absence of ATc. (C) Comparison of the relative frequency of each gliding type of motile parasites. A direct comparison between MIC8KOi grown in the presence or absence of ATc did not show any significant differences in the frequency of the three forms of gliding motility. Mean values of three independent experiments ± s.d. are shown.

shown that the micronemal protein MIC2 is essential for helical T. gondii expresses two highly conserved homologous of MIC8 and twirling motility but not for circular gliding (Huynh and Two highly conserved homologous proteins, MIC8-like1 (accession

Journal of Cell Science Carruthers, 2006). In the absence of MIC2, parasites show a number 76.m01679) and MIC8-like2 (accession number reduced ability to invade the host cell. To assess whether the 49m.03396) can be found in the available genome database of T. observed complete block in invasion upon depletion of MIC8 was gondii (www.toxoDB.org) (Cerede et al., 2005) (supplementary due to a reduction in gliding motility, we performed quantitative material Fig. S2). We confirmed expression of MIC8-like1 and live imaging. As described previously, three forms of gliding MIC8-like2 in T. gondii tachyzoites on the RNA level by reverse motility can be detected in this assay: circular gliding, upright transcriptase (RT)-PCR (supplementary material Fig. S2). The twirling and helical rotation (Fig. 4A) (Hakansson et al., 1999). function of a micronemal transmembrane protein appears to be We performed time-lapse microscopy on RH parasites and determined by the interaction of the C-terminal cytosolic tail MIC8KOi grown in the presence or absence of ATc. After image domain (CTD). In the case of members of the TRAP family, it has acquisition (one frame/3 seconds for 5 minutes), overlaying been demonstrated that the CTD interacts with the glycolytic projections of the acquired frames revealed the respective motility enzyme aldolase, which links this protein to the actin-myosin motor patterns. All three forms of gliding motility were identified in essential for gliding motility (Jewett and Sibley, 2003; Buscaglia parasites depleted of MIC8 (Fig. 4A). Although RH parasites et al., 2003). An alignment of the CTDs of different microneme showed a higher overall gliding motility than MIC8KOi (~30 vs proteins indicates that most sequences shown to be essential for 12% overall gliding motility), no significant reduction in gliding interaction with aldolase (Starnes et al., 2006) are not present in motility was observed between MIC8KOi grown in the presence the CTD of the MIC8 family members (supplementary material or absence of ATc (Fig. 4B). When the relative occurrences of the Fig. S2). This corresponds well with our phenotypic three different forms of gliding were compared, no significant characterisation of MIC8KOi that suggests a different function of difference between MIC8KOi grown in the presence or absence MIC8 during invasion. of ATc was observed (Fig. 4C). The general reduction of gliding motility in MIC8KOi parasites could be due to the extensive genetic Complementation analysis of MIC8KOi modifications of the parental RH parasites (similar to the effect We were interested to determine whether the CTD of different observed for attachment). However, because no difference in micronemal proteins can be exchanged for the CTD of MIC8. gliding motility was evident in the case of MIC8KOi parasites Therefore, we designed a complementation strategy, taking grown in the presence or absence of ATc, the total block in invasion advantage of the tight regulation and clear invasion phenotype of is unlikely to be due to reduced gliding motility in the absence of MIC8KOi, that allowed us to use ATc selection to obtain stable MIC8. transfected parasites. 952 Journal of Cell Science 121 (7)

Transfection of a construct leading to constitutive expression of the obtained parasite population was analysed for expression of a myc-tagged version of MIC8 served as a positive control. Stable either the myc- or Ty-tagged version of MIC8. We observed three selection was performed in the presence of ATc to deplete the different sub-populations that expressed MIC8Ty, MIC8myc or both parasite of the inducible Ty-tagged copy of MIC8. After 5-6 days, (Fig. 5A,B) in the presence of ATc, indicating that three independent Journal of Cell Science

Fig. 5. Complementation analysis of MIC8KOi using different chimeric proteins. (A) Schematic of the complementation strategy. Transfection of MIC8KOi with complementation constructs under ATc selection results in three different integration events: (1) homologous recombination leads to a promoter exchange resulting in constitutive expression of Ty-tagged MIC8 (stained in green); (2) a combination of both homologous and non-homologous recombination leads to constitutive expression of both MIC8Ty (stained in green) and the respective myc-tagged complementation construct (stained in red); (3) only non-homologous recombination leads to exclusive, constitutive expression of the myc-tagged complementation construct. This event indicates functional complementation. The immunofluorescence analysis shown was performed on a representative pool of MIC8KOi parasites complemented with pMIC8MIC8mycTMCTDMIC8 after 6 days under ATc selection. (B) Top: overview of the complementation constructs used in this study. The TMD and CTD of MIC8myc have been substituted as indicated. In the case of MIC2, two constructs were generated (see text). In the case of MIC8mycGPI (GPI), the TMD and CTD of MIC8 was exchanged for the GPI anchoring signal of the major surface antigen SAG1. Bottom: representative immunofluorescence analyses of MIC8KOi transfected with the indicated complementation construct. Stable transfection has been achieved by application of ATc selection as shown in A. Numbers in each image indicate the average percentage of parasites (± s.d.) in which random integration occurred in four independent experiments. Scale bars: 12.5 ␮m. Role of MIC8 in T. gondii invasion 953

integration events of the complementation construct occurred: the case of the construct MIC8myc and in the case of the chimera exchange of the promoter pT7S1 for pMIC8 (because of MIC8mycTMCTDMIC8-like1 (Fig. 5B). In all cases, we verified the homologous recombination of the complementation construct correct localisation of the respective chimeric protein in the pMIC8MIC8myc with construct T7S1MIC8Ty), random integration micronemes (Fig. 5B). Therefore, it appears that the correct targeting of the complementation construct, and a combination of random of the extracellular domains of MIC8 is not sufficient to restore the integration and homologous recombination (Fig. 5A). Only random function of MIC8. Furthermore, we conclude that the CTD of MIC8 integration, which resulted in exclusive expression of the myc- is functionally distinct and cannot be complemented by the CTD of tagged complementation construct, indicated functional microneme proteins belonging to the TRAP family or of AMA1. complementation for MIC8Ty (Fig. 5A). Interestingly, when we generated a mutant in which the highly We generated several complementation constructs in which the conserved C-terminal tryptophan of MIC8 was substituted for CTD and TMD of MIC8 were exchanged for the corresponding alanine, we were able to achieve complementation, indicating that domains of MIC2 (accession number AAB63303), AMA1 (accession this residue, although highly conserved in the MIC8 family (see number AAB65410), PfTRAP (accession number AAA29776), supplementary material Fig. S2), is not absolutely required for the MIC8-like1 or the GPI-anchoring signal of SAG1 (Fig. 5B, upper function of MIC8. panel). In the case of MIC2, we generated two different complementation constructs: one including sequences upstream of MIC8 is essential for formation of the moving junction the TMD that have previously been implicated in the correct When parasites are treated with cytochalasin D (CD), which leads processing of MIC2 (Brossier et al., 2003) (Mic2a), and one to an arrest in invasion, an MJ is formed and the content of the including only the TMD and CTD (Mic2b). Exclusive expression rhoptries is secreted into the cytosol of the host cell, leading to the of the respective complementation construct was only achieved in formation of evacuoles (Fig. 6A) (Hakansson et al., 2001). To Journal of Cell Science

Fig. 6. MIC8 is essential for the formation of the moving junction. (A) Formation of evacuoles is abolished in parasites depleted of MIC8. MIC8KOi, grown in the presence or absence of ATc, was incubated with host cells in the presence of CD. Formation of evacuoles (green) and an MJ (red) was analysed with monoclonal antibodies against Rop2-5 and polyclonal antibodies against RON4. No evacuoles were observed when MIC8KOi was grown in the presence of ATc. Blue staining indicates DAPI (nuclei). (B) Extracellular RON4 was detected by immunofluorescence analysis on non-permeabilised parasites expressing cytosolic GFP as a viability marker. MIC8KOi parasites were allowed to attach to host cells in the presence or absence of CD for 20 minutes. Red indicates an MJ. (C) Quantification of extracellular RON4, the presence of which indicates the formation of an MJ. Approximately 200 parasites were analysed in each assay (error bars indicate s.d. obtained from five independent assays). The units on the y-axis are percent of parasites with RON4 signal. (D) Depletion of MIC8 does not affect RON4 expression levels. Immunoblots from MIC8KOi and RH parasites, grown for 48 hours in the presence (+) or absence (–) of ATc, show efficient downregulation of MIC8 expression in the presence of ATc. No differences in RON4 expression levels could be detected. Scale bars: 15 ␮m (A); 20 ␮m (B). 954 Journal of Cell Science 121 (7)

determine whether parasites depleted of MIC8 were capable of that MIC8 and MIC3 are sorted to the micronemes by interacting forming an MJ and of secreting their rhoptries, MIC8KOi parasites with the same protein complex, that MIC3 contains sorting signals were grown in the presence or absence of ATc, treated with CD that ensure its transport to the micronemes independently of MIC8 and inoculated on HFF cells to determine evacuole formation. In or that the escorter function of MIC8 is redundant. Further the absence of MIC8, no formation of evacuoles was detected (Fig. experiments are required to clarify this complex issue. 6A). A similar phenotype has been reported for a conditional We examined crucial events during the asexual life cycle of the knockout of the micronemal protein AMA1. Parasites depleted of parasite and found that depletion of MIC8 explicitly affects the AMA1 still secrete RONs but fail to establish an MJ between the ability of the parasite to invade the host cell. We did not observe parasite and the host cell (Alexander et al., 2005; Mital et al., 2005). any effect on replication, egress or initial attachment to the host In this case, RONs associated with the MJ cannot be seen, but rather cell. Interestingly, reactivation of MIC8 expression in extracellular they localise in loose patches on the surface of the host cell. We parasites led to restoration of invasion potency. This shows that were interested to determine whether a similar phenotype could be MIC8 is not required during egress of the host cell and that the observed upon depletion of MIC8. invasion of parasites depends on a critical level of MIC8 expression. To analyse secretion and localisation of RONs, we used RON4 Furthermore, parasites depleted for this essential invasion factor as a marker. MIC8KOi parasites expressing GFP as a viability showed normal attachment and gliding motility, which is in sharp marker were grown in the presence or absence of ATc. Parasites contrast to the function determined for MIC2 (Huynh and were treated with or without CD before inoculation of HFF cells. Carruthers, 2006). Subsequently, analysis of RON4 secretion was performed in We noticed that parasites depleted of MIC8 did not form any immunofluorescence assays on non-permeabilised parasites (Fig. evacuoles, similar to a conditional mutant for AMA1 (Alexander et 6B). In the absence of MIC8, the occurrence of a RON4 signal was al., 2005; Mital et al., 2005). Parasites depleted for AMA1 are unable reduced by more then 90% (Fig. 6C) irrespective of whether the to form an MJ and it has been demonstrated that interaction of AMA1 parasites were treated with CD. Ablation of MIC8 had no influence with RONs is essential for this invasion step (Boothroyd and on the expression level of RON4 (Fig. 6D). Together, these results Dubremetz, 2008). However, in the absence of AMA1, RONs are demonstrate a crucial role of MIC8 during establishment of the MJ still released by the parasite and can be detected at the surface of the that is upstream of AMA1. host cell in loose patches (Alexander et al., 2005). In sharp contrast, we were not able to detect any signal for RON4 when intact non- Discussion permeabilised parasites were analysed for RON4 secretion. This Host-cell invasion by apicomplexan parasites is a multi-step process, indicates that MIC8 functions in an essential step during invasion requiring the sequential secretion of specialised secretory organelles that is downstream of MIC2 but upstream of AMA1 function. and the action of distinct sets of proteins (Carruthers and Boothroyd, The absence of RON4 on the surface of the host cell suggests 2006). Whereas secretion of micronemes has been demonstrated to that MIC8 is either required for the initial association of RONs with depend on a calcium-signalling cascade, the signals leading to the surface of the host cell or for rhoptry secretion. In the first case, secretion of the rhoptries and formation of the MJ have not been this role could be facilitated by function of the extracellular identified so far. Given the functional linkage of microneme and domains of MIC8. However, we demonstrated in a novel

Journal of Cell Science rhoptry secretion, we speculated that micronemal proteins might complementation approach that correct targeting of chimeric be implicated in this role. proteins consisting of the extracellular domains of MIC8 and the To shed light on the function of MIC8, we generated a conditional TMD and CTD of different micronemal proteins, such as MIC2, knockout for MIC8 by employing a tetracycline-inducible AMA1 and TRAP, did not restore the function conferred by wild- transactivator system (Meissner et al., 2007). For the generation of type MIC8. This demonstrates that correct targeting of the transgenic recipient strains with a second controllable copy of mic8, extracellular domains is not sufficient to restore the function of it was necessary to keep expression of the transgene turned off to MIC8. prevent any deleterious effects due to overexpression of mic8Ty. Having established that the CTDs of MIC8 and MIC8-like1 Interestingly, a strain overexpressing MIC8Ty showed a partial determine a novel functional role, we can now continue to identify accumulation of this protein in a post-Golgi compartment that the cytosolic factors interacting with this domain to gain more corresponded with early endosomes, as shown by co-localisation mechanistic insights into its function during the establishment of with the pro-peptide of M2AP (Harper et al., 2006). It is likely that, the MJ. Given the clear phenotype of MIC8KOi in invasion, it under conditions of MIC8 overexpression, a factor involved in appears that MIC8-like1 and MIC8-like2, which we show here to quality control or maturation of this protein is limiting, resulting in be expressed at the RNA level, cannot compensate for the loss of partial accumulation of MIC8 in early endosomes. A similar mic8 in tachyzoites. phenomenon was observed upon overexpression of MIC2 (Huynh Previously, it has been shown that MIC8 is capable of binding to and Carruthers, 2006). the host cell upon dimerisation (Cerede et al., 2002). It is therefore It has previously been demonstrated that MIC8 interacts with tempting to speculate that, upon dimerisation, MIC8 is involved in MIC3 during the transit through the secretory pathway (Meissner a signalling cascade that ultimately leads to rhoptry secretion. Similar et al., 2002a) and it was speculated that MIC8 might act as an mechanisms have been described in other eukaryotes for numerous escorter for MIC3, similar to other micronemal transmembrane non-tyrosine kinase receptors that trigger receptor-mediated proteins (Rabenau et al., 2001; Reiss et al., 2001). However, endocytosis and subsequently a signalling cascade (Li, 2005). In fact, depletion of MIC8 in the established MIC8KOi parasites did not we have recently demonstrated that Rab11A, a small G-protein result in any mislocalisation of MIC3, demonstrating that the essential for recycling endocytosis and associated with the rhoptries observed interaction between MIC8 and MIC3 is not essential for (Bradley et al., 2005), is required during invasion of the host cell the targeting of MIC3 to the micronemes. Currently, we can only (Herm-Gotz et al., 2007), indicating a link between receptor-mediated speculate on the mechanism involved in this targeting. It is possible endocytosis, receptor recycling and rhoptry secretion. Role of MIC8 in T. gondii invasion 955

In conclusion, here, we present evidence that the micronemal AGTA ATC CCCCTCGACC-3′); MIC2tail-a: MIC2sense1 (5′-GCGGCAATTGA - AGAGGGAGAACAGTCCAAGAAAG-3′) and MIC2as (5′-GCCTTAATTAAC - protein MIC8 of T. gondii plays an essential role during a distinct TACTCCATCCACATATCAC-3′); MIC2tail-b: MIC2sense2 (5′-GCGGCAATTG - step of host-cell invasion required for the early formation of the CGGGTGGCGTCATCGGCG-3′) and MIC2as (5′-GCCTTAATTAACTACTC- MJ. The characterisation of MIC8 as a key determinant for invasion CATCCACATATCAC-3′). See also alignment in Fig. 5B. demonstrates that micronemal proteins are involved in several Plaque assay independent steps during invasion that are likely to be conserved The plaque assay was performed as described previously (Roos et al., 1994). in apicomplexan parasites. The identification of each individual Monolayers of HFF, grown in six-well plates, were infected with 50-100 tachyzoites step will not only broaden our understanding of fundamental per well. After 1 week of incubation at normal growth conditions (37°C, 5% CO2), apicomplexan biology but might also lead to the identification of cells were fixed for 10 minutes with –20°C methanol 100%, dyed with Giemsa stain for 10 minutes and washed once with PBS. Documentation was performed with a new drug and vaccine candidates. The challenging issue now will Nikon SMZ 1500 binocular at 70-fold magnification. be to further elucidate the mechanistic action of MIC8, especially the role of the MIC8 CTD, which we show here to be essential for Invasion assay Monolayers of HFFs, grown in 24-well plates on glass slides, were infected with 107 its function. parasites/well and incubated for 5, 10, 20 or 60 minutes, or overnight, at 37°C. Subsequently, infected cells were washed four times with PBS and incubated for 12 Materials and Methods hours in fresh medium before fixation and indirect immunofluorescence. Data were compiled from more than three independent experiments each from counting six fields Parasite strains, transfection and culture ϫ T. gondii tachyzoites (RH hxgprt–) and TATi-1 (Meissner et al., 2002b) were grown of view per clone at 1000 total magnification with a Leica DMIL microscope in HFF cells and maintained in Dulbecco’s modified Eagle’s medium (DMEM), (Type:090-135-002 I/05). supplemented with 10% foetal calf serum (FCS), 2 mM glutamine and 25 μg/ml gentamycin. Selection based on chloramphenicol resistance or pyrimethamine Attachment assay resistance was performed as described previously (Kim et al., 1993; Donald and Roos, Attachment assays were performed as described previously (Huynh et al., 2003). 1993). Monolayers of HFFs, grown in 24-well plates on glass slides, were washed three times in PBS (containing 1 mM CaCl2 and 1 mM MgCl2) and fixed with PBS Generation of the inducible MIC8 knockout containing 2% glutaraldehyde for 5 minutes at 4°C. After three washing steps with PBS, cells were blocked with 0.16 M ethanolamine for 24 hours at 4°C. Subsequently To introduce a Ty-tag into mic8 and generate a vector allowing inducible expression 7 of this gene, the cDNA of mic8 was cloned into the vector p7TetOS1 and p7TetOS4 to another washing step with PBS, 10 parasites were loaded on each well and allowed downstream of the inducible promoter (Meissner et al., 2001). Next, inverse PCR to attach for 15 minutes at 37°C and rinsed three times with PBS before counting. ϫ using oligonucleotides MIC8Ty-as (antisense) (5′-CTTAATTAAGATGCATC- For each experiment, six fields of view at 400 total magnification were analysed GAGGGGGTCCTGGTTGGTGTGGACTTCTGGTAAACACTGCAACTTTGA - with a Leica DMIL microscope (Type: 090-135-002 I/05). CCC-3′) and MIC8Ty-s (sense) (5′-CCAATGCATCTACTGATCAGACAGATTTCG- 3′). The PCR fragment was subsequently digested with NsiI and re-ligated, resulting Analysis of microneme secretion 8 in the expression vectors T7S1MIC8Ty and T7S4MIC8Ty, respectively. In order to For preparation of excretory-secretory antigens (ESA), approximately 10 tachyzoites generate stable parasites with each construct, co-transfection of 60 μg T7S1MIC8Ty were washed and resuspended in 1 ml of HHE, stimulated to discharge micronemes or T7S4MIC8Ty with 30 μg pDHFR (Donald and Roos, 1993) was performed. by addition of ethanol to a final concentration of 1.0% and incubated at 37°C for 30 The knockout construct was based on the vector pT230CAT (Soldati and minutes, as described previously (Carruthers et al., 1999). Cells were removed by Boothroyd, 1995), in which the 5′UTR was amplified using oligonucleotides MIC8- centrifugation at 2000 g and the supernatant and pellet were analysed on immunoblot. 1 (5′-GGGGTACCACTCGTGGCGTCGCTG-3′) and MIC8-2 (5′-GGGGTACC - TGGTGGTGGGCGGAAAG-3′) and the 3′UTR with MIC8-3 (5′-GCGGATC - Analysis of moving-junction formation CTTAATTAAGCATAAGCTTTGTACCTGTGC-3′) and MIC8-4 (5′-TCCCCG - Analysis of evacuoles was performed as described previously (Hakansson et al., 2001). CGGCATTGAACAACATGTATGCG-3′) and placed within KpnI and BamHI/NotI Freshly released parasites were incubated in Endo buffer [44.7 mM K2SO4, 10 mM Journal of Cell Science restriction sites, respectively. 30 μg of the MIC8KOCAT construct was transfected Mg2SO4, 106 mM sucrose, 5 mM glucose, 20 mM Tris, 0.35% (wt/vol) BSA, pH and, subsequently, CAT selection was performed. MIC8KOi parasites were isolated 8.2] containing 1 μM CD for 10 minutes at room temperature, before adding them from the obtained stable pool of transfectants via limiting dilution and identified on confluent layers of HFFs. 3ϫ106 parasites were added per well and kept for 3 via genomic PCR using insert-, genome- and integration-specific oligonucleotide minutes at room temperature. After 20 minutes incubation at 37°C, medium was combinations (see Fig. 2A). MIC8KOi parasites were then verified in immunoblots replaced with pre-warmed Hanks/HEPES/1% FCS/1 μM CD media and incubated and immunofluorescence assays to confirm inducible expression of MIC8Ty. For a for 15 minutes at 37°C. After two washing steps with PBS, the cells were fixed with subset of studies, MIC8KOi was transfected with p5RT70GFP and GFP-expressing methanol at –20°C for 10 minutes before the immunofluorescence assay was parasites were isolated based on green fluorescence. performed. Analysis of RON4 secretion was performed as above with modifications, employing Generation of complementation constructs GFP-expressing parasites to distinguish intact parasites from disrupted parasites. After For the generation of complementation constructs encoding the extracellular domains incubation in the presence or absence of CD, parasites were fixed with 4% of MIC8 fused to the TMD and CTD of MIC2, AMA1, PfTRAP, MIC8/2, paraformaldehyde and immunofluorescence analysis was done without MIC8(W>A) and MIC8GPI, we exchanged the DNA sequence encoding the Ty-tag permeabilisation of parasites to distinguish between secreted and non-secreted in T7S1MIC8Ty for a myc-tag using complementary oligonucleotides MIC8myc-s RON4. Equal numbers of GFP-positive parasites (~200) were analysed for RON4 (5′-AATTCTTTCAAAGTTGCAGTGTTTACCTAGGGAGCAGAAGCTCATCTC - secretion per experiment. CGAGGAGGACCTAGATGCATCTTAAT-3′) and MIC8myc-as (5′-TAAGATGC - ATCTAGGTCCTCCTCGGAGATGAGCTTCTGCTCCCTAGGTAAACACTGC A - Live video microscopy gliding assay ACTTTGAAAG-3′). In a second step, the TMD and CTD of MIC8 were re- Glass slides in a 24-well plate were coated with PBS including 50% FCS for 24 introduced within NsiI and PacI restriction sites. Next, the promoter p7TetOS1 was hours at 4°C. The glass slides were washed once with PBS before adding 106 parasites exchanged for p5RT70 and pMic using KpnI/EcoRI, resulting in the constructs in a total volume of 300 μl pre-warmed Hank’s medium containing 20 mM HEPES p5RT70MIC8myc and pMIC8MIC8myc, respectively. Amplification of pMIC8 was and 1 mM EGTA. The parasites were then imaged (37°C/5% CO2) using an inverted achieved using oligonucleotides pMIC8-s (5′-CCCGGTACCTGGTGGTGGGCG- Zeiss microscope (Axiovert 200 M with incubator XL-3 and CO2 control) using 16ϫ 3′) and pMIC8-as (5′-GCCGAATTCGAAACGAAAGATACCGCGACCAGC-3′). magnification (objective 10ϫ, Optovar 1.6ϫ) and differential interference contrast Subsequently, the coding regions for different transmembrane CTDs were introduced (DIC). Videos were captured at one frame per 3 seconds using Axiovision AxioVs40 into MfeI/PacI restriction sites, resulting in different complementation constructs. V 4.6.3.0 software and the type of gliding/movement was counted by observing every The oligonucleotides used for amplification of the TMCTD-encoding DNA individual parasite. On average, 447 parasites/movie (101-1052 parasites/movie) were sequence were: MIC8CTD MIC8/2-tail: MIC8/2-sense2 (5′-GCGGCAAT - monitored and enumerated for all strains. Using ImageJ 1.34r software, projections TGGACTG GCA ACGGGAGCC-3′) and MIC8/2-as (5′-GCCTTAATTAATT - of the standard deviation and of the maximum of the fluorescent intensity between ATGACCACATTGA GCCTG-3 ′); MIC8CTD MIC8W>A tail: MIC8sense-MfeI frames revealed motility patterns over a 5-minute period. (5′-GCGGCAATTGCATTGGTGGTTGTGG-3 ′) and MIC8-W>Aas (5′-GCCT - TAATTAAGACG CG ATACCGCCCGAAGG-3′); PfTRAP-tail: PfTRAP-sense Immunofluorescence and immunoblot (5′-GCGGCAATTGC AGG TGGAATAGCTGGAGG-3′) and PfTRAP-as (5′-GCC - All manipulations were carried out at room temperature if not mentioned otherwise. TTAATTAATTCCACTCGTTTTCTTCAG-3 ′); AMA1-tail: AMA1-s (5′-GCGGCA- Tachyzoite-infected HFF cells on glass coverslips were fixed with 4% formaldehyde ATTGCTGGACT CG CCGTAGGAG-3′) and AMA1-as (5′-GCCTTAATTAACT - in PBS for 20 minutes. Fixed cells were permeabilised with 0.2% Triton X-100 in 956 Journal of Cell Science 121 (7)

PBS for 20 minutes and blocked in 3% bovine serum albumin in PBS for 20 minutes. Garcia-Reguet, N., Lebrun, M., Fourmaux, M. N., Mercereau-Puijalon, O., Mann, T., The cells were then stained with the primary antibodies, as indicated in the text, for Beckers, C. J., Samyn, B., Van Beeumen, J., Bout, D. and Dubremetz, J. F. (2000). 1 hour followed by Alexa-Fluor-594-conjugated goat anti-rabbit or Alexa-Fluor-488- The microneme protein MIC3 of Toxoplasma gondii is a secretory adhesin that binds conjugated goat anti-mouse antibodies (Molecular Probes). Images were taken using to both the surface of the host cells and the surface of the parasite. Cell. Microbiol. 2, Zeiss (Axiovert 200M) and processed using ImageJ 1.34r software. 353-364. SDS page was performed as described previously (Laemmli, 1970). Freshly Hakansson, S., Morisaki, H., Heuser, J. and Sibley, L. D. (1999). Time-lapse video egressed parasites were harvested by centrifugation at 300 g for 10 minutes and lysed microscopy of gliding motility in Toxoplasma gondii reveals a novel, biphasic mechanism in RIPA buffer (150 mM NaCl 1% Triton X-100, 0.5% DOC, 0.1% SDS, 50 mM of cell locomotion. Mol. Biol. Cell 10, 3539-3547. Hakansson, S., Charron, A. J. and Sibley, L. D. (2001). Toxoplasma evacuoles: a two- Tris, pH 8.0, 1 mM EDTA). Equal numbers of parasites were applied per line on a step process of secretion and fusion forms the . EMBO J. 20, 10% polyacrylamide gel. Transfer to nitrocellulose and immunoblot analysis was 3132-3144. performed as described previously (Reiss et al., 2001). Harper, J. M., Huynh, M. H., Coppens, I., Parussini, F., Moreno, S. and Carruthers, The following previously described primary antibodies have been used in these V. B. (2006). A cleavable propeptide influences Toxoplasma infection by facilitating the α α analysis: -MIC2 (6D10) (Wan et al., 1997), -MI3 (T82C10) (Garcia-Reguet et al., trafficking and secretion of the TgMIC2-M2AP invasion complex. Mol. Biol. Cell 17, 2000), α-Ron4 (t5 4H1) (Lebrun et al., 2005), α-RON4 (Alexander et al., 2006), α- 4551-4563. MIC6 (Reiss et al., 2001), α-MIC8Nt (Meissner et al., 2002a) and α-Rop2,3,4 (T3 Herm-Gotz, A., Agop-Nersesian, C., Munter, S., Grimley, J. S., Wandless, T. J., 4A7) (Sadak et al., 1988). Frischknecht, F. and Meissner, M. (2007). Rapid control of protein level in the apicomplexan Toxoplasma gondii. Nat. Methods 4, 1003-1005. Sequence analysis of micronemal proteins Hoff, E. F. and Carruthers, V. B. (2002). Is Toxoplasma egress the first step in invasion? DNA and predicted proteins homologous to MIC8 were identified using the BLAST Trends Parasitol. 18, 251-255. programme provided at ToxoDB (http:/ToxoDB.org). Multiple sequence alignments Huynh, M. H. and Carruthers, V. B. (2006). Toxoplasma MIC2 is a major determinant were performed with the help of ClustalW. Expression of MIC8, MIC8-like1 and of invasion and virulence. PLoS Pathog. 2, e84. MIC8-like2 was confirmed by RT-PCR (Titan one tube RT-PCR; Roche) on mRNA Huynh, M. H., Rabenau, K. E., Harper, J. M., Beatty, W. L., Sibley, L. D. and Carruthers, V. B. (2003). Rapid invasion of host cells by Toxoplasma requires secretion (isolated with mRNAeasy; Biochain) using oligonucleotide pairs MIC8-s (5′- of the MIC2-M2AP adhesive protein complex. EMBO J. 22, 2082-2090. GCGGCAATTGCATTGGTGGTTGTGG-3′), MIC8as (5′-GCCTTAATTAAGA - ′ ′ Jewett, T. J. and Sibley, C. H. (2003). Aldolase forms a bridge between cell surface adhesins CGCGATACCGCCCGAAGG-3 ), MIC8/2-s (5 -GCGGCAATTGGACTG GCA - Mol. Cell 11 ′ ′ and the actin cytoskeleton in apicomplexan parasites. , 885-894. ACGGGAGCC-3 ), MIC8/2-as (5 -GCCTTAATTAATTATGACCACATTGAG - Kim, K., Soldati, D. and Boothroyd, J. C. (1993). Gene replacement in Toxoplasma gondii ′ ′ ′ CCTG-3 ), MIC8/3-s (5 -GGCCAATTGCCGGCTGCGTTGTCGGC-3 ), MIC8/3- with chloramphenicol acetyltransferase as selectable marker. Science 262, 911-914. as (5′-GTGTTGTCGCCTTCTCGC-3′), Tub-s (5′-GGAGCTCTTCTGCCTG - Laemmli, U. K. (1970). Cleavage of structural proteins during the assembly of the head GAACATGGTATC-3′) and Tub-as (5′-AGAGTGAGTGGAAAGGACGGAGT - of bacterophage T4. Nature 227, 680-685. TGTACG-3′). Lebrun, M., Michelin, A., El Hajj, H., Poncet, J., Bradley, P. J., Vial, H. and Dubremetz, J. F. (2005). The rhoptry neck protein RON4 re-localizes at the moving junction during We are indebted to John Boothroyd, David Sibley, Vern Carruthers, Toxoplasma gondii invasion. Cell. Microbiol. 7, 1823-1833. Gary Ward, Jean-Francois Dubremetz and Maryse Lebrun for sharing Li, S. S. (2005). Specificity and versatility of SH3 and other proline-recognition domains: structural basis and implications for cellular signal transduction. Biochem. J. 390, 641- reagents and helpful discussions. We thank Petra Rohrbach, Vern 653. Carruthers and Kai Matuschewski for helpful discussions and critical Meissner, M., Brecht, S., Bujard, H. and Soldati, D. (2001). Modulation of myosin A comments on the manuscript. This work was funded by the SFB544 expression by a newly established tetracycline repressor-based inducible system in of the German research foundation (DFG) and the ‘Biofuture- Toxoplasma gondii. Nucleic Acids Res. 29, E115. Meissner, M., Reiss, M., Viebig, N., Carruthers, V. B., Toursel, C., Tomavo, S., Ajioka, Programm’ of the German Ministry of Science and Education (BMBF). J. W. and Soldati, D. (2002a). A family of transmembrane microneme proteins of Toxoplasma gondii contain EGF-like domains and function as escorters. J. Cell Sci. 115, References 563-574. Alexander, D. L., Mital, J., Ward, G. E., Bradley, P. and Boothroyd, J. C. (2005). Meissner, M., Schluter, D. and Soldati, D. (2002b). Role of Toxoplasma gondii myosin Identification of the moving junction complex of Toxoplasma gondii: a collaboration A in powering parasite gliding and host cell invasion. Science 298, 837-840. between distinct secretory organelles. PLoS Pathog. 1, e17. Meissner, M., Agop-Nersesian, C. and Sullivan, W. J., Jr (2007). Molecular tools for

Journal of Cell Science Alexander, D. L., Arastu-Kapur, S., Dubremetz, J. F. and Boothroyd, J. C. (2006). analysis of gene function in parasitic . Appl. Microbiol. Biotechnol. 75, falciparum AMA1 binds a rhoptry neck protein homologous to TgRON4, 963-975. a component of the moving junction in Toxoplasma gondii. Eukaryotic Cell 5, 1169- Mital, J., Meissner, M., Soldati, D. and Ward, G. E. (2005). Conditional expression of 1173. Toxoplasma gondii apical membrane antigen-1 (TgAMA1) demonstrates that TgAMA1 Black, M. W., Arrizabalaga, G. and Boothroyd, J. C. (2000). Ionophore-resistant mutants plays a critical role in host cell invasion. Mol. Biol. Cell 16, 4341-4349. of Toxoplasma gondii reveal host cell permeabilization as an early event in egress. Mol. Pfluger, S. L., Goodson, H. V., Moran, J. M., Ruggiero, C. J., Ye, X., Emmons, K. M. Cell. Biol. 20, 9399-9408. and Hager, K. M. (2005). Receptor for retrograde transport in the apicomplexan parasite Boothroyd, J. C. and Dubremetz, J. F. (2008). Kiss and spit: the dual roles of Toxoplasma Toxoplasma gondii. Eukaryotic Cell 4, 432-442. rhoptries. Nat. Rev. Microbiol. 6, 79-88. Rabenau, K. E., Sohrabi, A., Tripathy, A., Reitter, A., Ajioka, J. W., Tomley, F. M. Bradley, P. J., Ward, C., Cheng, S. J., Alexander, D. L., Coller, S., Coombs, G. H., and Carruthers, V. B. (2001). TgM2AP participates in Toxoplasma gondii invasion of Dunn, J. D., Ferguson, D. J., Sanderson, S. J., Wastling, J. M. et al. (2005). Proteomic host cells and is tightly associated with the adhesive protein TgMIC2. Mol. Microbiol. analysis of rhoptry organelles reveals many novel constituents for host-parasite 41, 1-12. interactions in Toxoplasma gondii. J. Biol. Chem. 280, 34245-34258. Reiss, M., Viebig, N., Brecht, S., Fourmaux, M. N., Soete, M., Di Cristina, M., Brossier, F., Jewett, T. J., Lovett, J. L. and Sibley, L. D. (2003). C-terminal processing Dubremetz, J. F. and Soldati, D. (2001). Identification and characterisation of an of the toxoplasma protein MIC2 is essential for invasion into host cells. J. Biol. Chem. Escorter for two secretory adhesins in Toxoplasma gondii. J. Cell Biol. 152, 563-578. 278, 6229-6234. Roos, D. S., Donald, R. G., Morrissette, N. S. and Moulton, A. L. (1994). Molecular Buscaglia, C. A., Coppens, I., Hol, W. G. and Nussenzweig, V. (2003). Sites of interaction tools for genetic dissection of the protozoan parasite Toxoplasma gondii. Methods Cell between aldolase and thrombospondin-related anonymous protein in plasmodium. Mol. Biol. 45, 27-63. Biol. Cell 14, 4947-4957. Sadak, A., Taghy, Z., Fortier, B. and Dubremetz, J. F. (1988). Characterization of a Carruthers, V. and Boothroyd, J. C. (2006). Pulling together: an integrated model of family of rhoptry proteins of Toxoplasma gondii. Mol. Biochem. Parasitol. 29, 203-211. Toxoplasma cell invasion. Curr. Opin. Microbiol. 10, 83-89. Soldati, D. and Boothroyd, J. C. (1995). A selector of transcription initiation in the Carruthers, V. B., Giddings, O. K. and Sibley, L. D. (1999). Secretion of micronemal protozoan parasite Toxoplasma gondii. Mol. Cell. Biol. 15, 87-93. proteins is associated with toxoplasma invasion of host cells. Cell. Microbiol. 1, 225- Soldati, D. and Meissner, M. (2004). Toxoplasma as a novel system for motility. Curr. 235. Opin. Cell Biol. 16, 32-40. Cerede, O., Dubremetz, J. F., Bout, D. and Lebrun, M. (2002). The Toxoplasma gondii Starnes, G. L., Jewett, T. J., Carruthers, V. B. and Sibley, L. D. (2006). Two separate, protein MIC3 requires pro-peptide cleavage and dimerization to function as adhesin. conserved acidic amino acid domains within the Toxoplasma gondii microneme protein EMBO J. 21, 2526-2536. 2 cytoplasmic tail are required for parasite survival. J. Biol. Chem. 281, 30745-30754. Cerede, O., Dubremetz, J. F., Soete, M., Deslee, D., Vial, H., Bout, D. and Lebrun, M. Sultan, A. A., Thathy, V., Frevert, U., Robson, K. J., Crisanti, A., Nussenzweig, V., (2005). Synergistic role of micronemal proteins in Toxoplasma gondii virulence. J. Exp. Nussenzweig, R. S. and Menard, R. (1997). TRAP is necessary for gliding motility Med. 201, 453-463. and infectivity of plasmodium sporozoites. Cell 90, 511-522. Donald, R. G. K. and Roos, D. S. (1993). Stable molecular transformation of Toxoplasma Wan, K. L., Carruthers, V. B., Sibley, L. D. and Ajioka, J. W. (1997). Molecular gondii: a selectable markerDHFR-TS marker based on drug resistance mutations in characterisation of an expressed sequence tag locus of Toxoplasma gondii encoding the malaria. Proc. Natl. Acad. Sci. USA 90, 11703-11707. micronemal protein MIC2. Mol. Biochem. Parasitol. 84, 203-214.