|Home | About | Journals | Submit | Contact Us | Français|
Accurate sorting of proteins to the three types of secretory granules in Toxoplasma gondii is crucial for successful cell invasion by this obligate intracellular parasite. As in other eukaryotic systems, propeptide sequences are a common yet poorly understood feature of proteins destined for regulated secretion, which for Toxoplasma occurs through two distinct invasion organelles, rhoptries and micronemes. Microneme discharged during parasite apical attachment plays a pivotal role in cell invasion by delivering adhesive proteins for host receptor engagement.
We show here that the small micronemal proprotein MIC5 undergoes proteolytic maturation at a site beyond the Golgi and only the processed form of MIC5 is secreted via the micronemes. Proper cleavage of the MIC5 propeptide relies on an arginine residue in the P1′ position, though P1′ mutants are still cleaved to a lesser extent at an alternative site downstream of the primary site. Nonetheless, this aberrantly cleaved species still correctly traffics to the micronemes, indicating that correct cleavage is not necessary for micronemal targeting. In contrast, a deletion mutant lacking the propeptide was retained within the secretory system, principally in the endoplasmic reticulum. The MIC5 propeptide also supported correct trafficking when exchanged for the M2AP propeptide, which was recently shown to also be required for micronemal trafficking of the TgMIC2-M2AP complex (Harper et al., Mol Biol Cell (2006) 4551–63).
Our results illuminate common and unique features of micronemal propeptides in their role as trafficking facilitators.
Many cell types use regulated secretion pathways to deliver biologically active molecules, such as hormones, neuropeptides, and enzymes, precisely where and when they are needed to execute function. Proteins destined for regulated secretion are often synthesized as preproproteins with an N-terminal signal peptide and a separate, cleavable prosequence. These preproproteins are cleaved by signal peptidase during translocation into the endoplasmic reticulum (ER), but the resulting proproteins vary both temporally and physically in maturational processing to remove the propeptide (reviewed in Burgess and Kelly, 1987, Tooze, 1991).
While classic examples of proproteins are inactive, zymogen versions of proteases such as pepsin and trypsin (Previero, et al., 1964, Ryle, 1965), propeptide sequences can also fulfill other roles, including protein trafficking, either to protect mature sequences from degradation in the ER, or as sorting signals to cellular compartments and regulated secretion pathways (Gullberg, et al., 1999, Conkright, et al., 2001). As many of these proteins are bioactive and potentially damaging to surrounding tissue if their release is uncontrolled, effective storage and proper sorting signals are of paramount importance.
Parasites in the phylum Apicomplexa, including the malaria parasite Plasmodium, Cryptosporidium, which is responsible for AIDS-related diarrhoea, and Toxoplasma gondii, the causative agent of toxoplasmosis, have the standard eukaryotic organelles and targeting pathways of higher organisms, as well as unique compartments such as the apicoplast, a vestigial chloroplast-like organelle (Kohler, et al., 1997, Waller, et al., 1998). Additionally, apicomplexans contain three different types of secretory vesicles (micronemes, rhoptries, and dense granules (DG)), suggesting that intricate trafficking pathways exist. Since each type of secretory vesicle has a separate and specific role in invasion and intracellular replication within host cells (Carruthers and Sibley, 1997, Dubremetz, 1998, Liendo and Joiner, 2000, Tomley and Soldati, 2001), accurate protein sorting is crucial to the survival of these obligate intracellular parasites.
Several distinct sorting signals have been identified in Toxoplasma, which, due to its accessibility to in vitro propagation and genetic manipulation, is one of the most well studied apicomplexans. A tyrosine-rich motif (YXX, where X is any amino acid and is a bulky hydrophobic residue) within the cytoplasmic domains of transmembrane proteins is implicated in micronemal targeting (Di Cristina, et al., 2000), whereas a separate cytoplasmic motif (LL) facilitates trafficking to rhoptries via an adapter protein-1 complex (Hoppe, et al., 2000). Although little is known about luminal forward targeting elements, previous studies have implicated propeptides as navigational facilitators for microneme or rhoptry proteins. For example, the propeptide of MIC2-associated protein (M2AP) is necessary but not sufficient for microneme trafficking as a propeptide deletion mutant is arrested in the post-Golgi region (Harper, et al., 2006), yet proM2AP still relies on its transmembrane partner MIC2 to access the micronemes (Huynh and Carruthers, 2006). Contrastingly, the propeptide of ROP1 is sufficient but not necessary as it supports rhoptry targeting when fused to a heterologous protein yet is dispensable for targeting of ROP1 (Soldati, et al., 1998, Bradley and Boothroyd, 2001). Proteins destined for the DG appear to lack forward targeting signals altogether and therefore traffic through this route by default (Karsten, et al., 1998).
Secretion from the three secretory organelles is differentially regulated. Of the three, DG most closely resemble constitutive secretory vesicles of higher eukaryotes in that secretion is essentially continuous, although DG are modestly upregulated within 20 min of invasion (Chaturvedi, et al., 1999, Liendo, et al., 2001). Micronemes show intermediate regulation, with a constant, basal level of release, but exhibiting a marked calcium-dependent burst of secretion upon host cell contact (Wan, et al., 1996, Carruthers and Sibley, 1997, Carruthers and Sibley, 1999). Finally, rhoptries are the most tightly regulated, secreting their contents in a bolus exclusively when the parasite begins to penetrate into the host cell (Carruthers and Sibley, 1997). Propeptide sequences are strictly associated with regulated secretion pathways, as they are common among rhoptry and micronemal proteins (Sadak, et al., 1988, Ossorio, et al., 1992, Soldati, et al., 2001), but not found on proteins destined for the DG.
We present here a detailed analysis of the role of the propeptide in trafficking of MIC5, a small, soluble micronemal protein necessary for correct trimming of invasion-related proteins on the parasite surface (Brydges, et al., 2000, Brydges, et al., 2006).
To determine where proMIC5 resides within the parasite and its approximate site of proteolytic maturation, we raised antibodies to a synthetic peptide encompassing the entire MIC5 propeptide. Immunoblotting of wild-type parasites showed that affinity purified αMIC5propeptide antibodies are highly specific for proMIC5 (Figure 1A). No signal was seen for ΔMIC5 parasites (Brydges, et al., 2006) by immunoblotting or immunofluorescence assay (IFA), confirming the antibody specificity (data not shown). When extracellular parasites were dual-stained with αMIC5propeptide and αMIC2 by IFA, some overlap was seen in the central anterior region (arrow in Figure 1Ba) but little co-staining was seen in the micronemes. αMIC5propeptide also uniquely stained small punctuate structures in both the anterior and posterior regions. As shown in Figure 1Bb, proMIC5 in the central anterior region partially overlapped with the Golgi marker GRASP52 (Pelletier, et al., 2002), but it also extended into immediately adjacent sites. Some costaining with GRA4 was also observed indicating that a portion of proMIC5 traverses this default pathway (Figure 1Bc). Most strikingly, however, the distribution of proMIC5 coincided closely with that of proM2AP (Figure 1Bd), which was recently shown to occupy early endosomes and associated structures prior to cleavage (Harper, et al., 2006).
Golgi/endosomal localization of proMIC5 is consistent with previous studies in which MIC5 propeptide processing was ablated by treatment with brefeldin A (Brydges, et al., 2000), implying that cleavage occurs in the trans Golgi or beyond, as found in mammalian cells (Lippincott-Schwartz, et al., 1989, Lippincott-Schwartz, et al., 1990). These results were further corroborated by immunoelectron microscopy (Figure 1C and 1C′), which detected substantial proMIC5 (10 nm gold particles) in the Golgi stack and within ~100 nm post-Golgi vesicles. These structures were also labelled with αMIC5 (5 nm gold) since this antibody recognizes both pro- and mature-MIC5. Consistent with the IFA results, occasional staining of the DG was also observed with αMIC5propeptide (Figure 1D). Sporadic labelling of the nucleus was also evident in some sections, but the significance of this observation is unknown. We conclude that, similar to proM2AP, proMIC5 traffics through the Golgi and post-Golgi vesicles, and that processing occurs prior to reaching the mature micronemes, possibly in the early endosome and associated tubulovesicular structures (Harper, et al., 2006).
Although proMIC5 is secreted (Brydges, et al., 2000), its non-micronemal distribution implies that it is discharged via a route independent of the micronemes. To confirm this, we analyzed the secretion profiles of proMIC5 and mature MIC5 with and without 1,2-bis(o-aminophenoxy)ethane-N,N,N′,N′-tetraacetic acid-acetoxymethlyester (B-AM) treatment, which selectively blocks microneme secretion by chelating intracellular calcium (Carruthers and Sibley, 1999). As shown in Figure 2, secretion of the dense granule protein GRA1 was only modestly affected by B-AM, as expected since dense granule secretion is calcium independent. In contrast, secretion of the micronemal protein, MIC2 and mature MIC5, was completely abrogated by B-AM treatment. However, treatment with B-AM had no effect on secretion of proMIC5, demonstrating that it is secreted via an alternative route than that used by microneme proteins, which at least in part involves DG, as indicated above.
To define the sequence requirements for propeptide cleavage, we performed alanine substitution mutagenesis in an attempt to create a cleavage-resistant mutant. Mutation of the P1 and/or P2 residues did not substantially block cleavage; however, mutation of the P1′ or P1′ and P1 residues resulted in marked reduction of mature MIC5 and accumulation of proMIC5 (Figure 3A and 3B). (Note that M5P1P1’myc has an inadvertent secondary mutation in P3 slightly retards its migration). Also, the residual mature MIC5 in these mutants migrated ~2 kDa below that of the wild-type protein and the other mutants. These aberrant products reacted with αmyc antibodies (data not shown), indicating that they retain the C-terminal myc tag and are therefore processed at the N-terminus. Since mutation of P1′ residue alone produced this cleavage phenotype, we performed further experiments on MIC5P1’-myc expressing parasites only.
To define the alternative cleavage site, we immuno-purified MIC5P1’-myc and subjected it to N-terminal sequencing by Edman degradation. Analysis of five cycles yielded the sequence N-D-V-E-S, which corresponds to a site six residues down from the wild-type processing site (Figure 3C). This finding explains the more rapid migration of MIC5P1’-myc and reveals the presence of an alternative cleavage site that is used, albeit at lower efficiency, upon disruption of the normal site.
In spite of the aberrant cleavage, MIC5P1’-myc expressing parasites showed extensive staining of apical micronemes (Figure 3D) and minimal staining of the DG (Figure 3E). These findings demonstrated that correct processing of proMIC5 is not a prerequisite for micronemal targeting, which is consistent with recent observations for a cleavage resistant mutant of M2AP (Harper, et al., 2006).
To determine whether the MIC5 propeptide contains targeting information, we engineered a propeptide deletion construct consisting of the signal sequence fused directly to the mature MIC5 sequence (MIC5Δpro-myc, Figure 4A) and stably transfected it into ΔMIC5 parasites. As expected, MIC5Δpro-myc migrated as a single band corresponding to the mature protein (Figure 4B). In extracellular parasites, MIC5Δpro-myc was scattered throughout the cytoplasm in contrast to the apical microneme pattern exhibited by MIC5-myc (Figure 4C). Although intracellular MIC5Δpro-myc parasites similarly showed diffuse cytoplasmic staining, they also prominently displayed perinuclear staining in a pattern consistent with the parasite ER. Secretion assays showed that MIC5Δpro-myc was secreted, but in a B-AM insensitive manner, corroborating the microscopy data suggesting MIC5Δpro-myc is not stored in the micronemes (Figure 4D). These results suggest that the MIC5 propeptide is necessary for efficient exit from the ER and that although some MIC5Δpro-myc escapes the ER and is secreted, this material is released through a non-micronemal route. Aberrant trafficking of MIC5Δpro-myc did not appear to overtly affect other microneme proteins since normal trafficking of AMA1 was seen (Figure 4C), along with correct processing of M2AP, MIC3, MIC6, and MIC11 (data not shown).
A previous study had shown that the 24 aa propeptide of M2AP is necessary for trafficking of the MIC2-M2AP protein complex to the micronemes, and that deletion of the M2AP propeptide (ΔproM2AP) resulted in accumulation of the complex in the post-Golgi region, partial degradation of MIC2, and inefficient secretion (Harper, et al., 2006). To test whether the MIC5 propeptide can rescue these defects, we fused the 14 aa MIC5 propeptide to the M2AP mature sequence and stably transfected the construct (M5proM2AP) into M2AP knockout (M2APKO) parasites. As shown in Figure 5A, M5proM2AP expressing parasites displayed a minor ~41 kDa band of the correct size for unprocessed M5proM2AP and a prominent 40 kDa band corresponding to mature M2AP, suggesting that the MIC5 propeptide was removed by proteolysis at or near the correct processing site. A minor ~38 kDa band (asterisk) was also visible that may be the result of further processing. Similar to 1C4 (a genetically complemented clone expressing M2AP in the knockout background), M5proM2AP expressing parasites showed minimal MIC2 degradation, unlike ΔproM2AP and M2APKO (Figure 5A and 5B). Also, the MIC2-M2AP complex is assembled correctly in M5proM2AP parasites, based on successful immunoprecipitation of both proteins with antibodies to either MIC2 or M2AP (data not shown). Moreover, induced secretion of MIC2 in M5proM2AP parasites was close to the normal level seen in 1C4, whereas ΔproM2AP and M2APKO parasites released substantially less MIC2 (Figure 5C). Finally, the apical distribution of M5proM2AP was indistinguishable from that of M2AP in 1C4, in contrast to ΔproM2AP, which localized prominently within a post-Golgi site (arrow in Figure 5D) and the parasitophorous vacuole (arrowhead). Together, these findings suggest that targeting information contained within the MIC5 propeptide is transferable to a heterologous microneme protein, indicating an exchangeable role of micronemal propeptides in protein trafficking.
Precisely how proteins are sorted for constitutive versus regulated secretion is not well understood in any biological system. Apicomplexan parasites possess at least four pathways through which proteins are delivered to the cell surface or beyond, including two constitutive pathways (DG and the pathway for glycosylinositol phosphate-anchored surface proteins i.e., SAGs in Toxoplasma) and two regulated pathways (rhoptries and micronemes). Proteolytic maturation of proproteins is a common feature of regulated pathways in such parasites (and in higher eukaryotes), with roles emerging for propeptides in sorting, folding, and/or regulation of enzymatic or other activities. Our findings reveal that the maturation and sorting of the small luminal microneme protein MIC5 shares many features with that of other micronemal proteins, but also several differences.
Using antibodies against their respective propeptides, we show that proMIC5 and proM2AP follow a similar path through the Golgi and vesicular post-Golgi structures. Also, both proteins appear to be processed as a similar location in the parasite based on their near perfect co-distribution in the proximal anterior region. This, together with the observation that MIC5 is cleaved more rapidly (T1/2 ~15 min) (Brydges, et al., 2000) than M2AP (T1/2 ~40 min) (Harper, et al., 2006), implies that MIC5 travels through the early secretory system to the processing compartment at a greater pace than M2AP. These differential transport kinetics may depend on how quickly the respective proteins assemble into protein complexes and on the transport properties of their partner proteins (MIC5 is presumed to couple with another MIC protein based on its association with the parasite surface during invasion). proMIC5 and proM2AP are also both released via a constitutive, non-micronemal pathway, implying that a fraction of proproteins either never enters the regulated pathway or is diverted to an alternative pathway along the way. Alternative trafficking and secretion of the lysosomal enzyme procathepsin L to small dense vesicles has been seen in transformed fibroblasts (Ahn, et al., 2002). This phenomenon may be due to saturation of the mannose 6-phosphate receptor-based lysosomal targeting system as a result of overexpression of procathepsin L. Diversion of proMIC5 and proM2AP to an alternative pathway might also be due to saturation of sorting receptors that interact directly with these proteins or indirectly with their partners.
The finding that propeptide deletion mutants of MIC5 and M2AP both fail to access the micronemes and are instead arrested in the post-Golgi region also argues for the existence of a cargo receptor(s) that facilitates trafficking to or packaging within the micronemes. Several cargo receptors have been identified in regulated pathways of animal cells. For example, muclin is a luminal protein that binds pancreatic zymogens in the Golgi through its sulfated O-linked oligosaccharides (Boulatnikov and De Lisle, 2004) and promotes formation of regulated secretory granules (De Lisle, et al., 2005). Carboxypeptidase E has been implicated as a sorting receptor for several prohormones in endocrine cells in a manner dependent on association with lipid rafts (Arnaoutova, et al., 2003). Similarly, prohormone convertase 3 associates with lipid rafts through its C-terminus, which is necessary and sufficient for sorting to the regulated pathway (Assadi, et al., 2004). However, direct involvement of cargo propeptides in these examples has not been demonstrated. The propeptide of the snail conotoxin TxVI binds to the cargo receptor sortilin and this interaction facilitates its exit from the ER of transfected COS7 cells (Conticello, et al., 2003). Sortilin is widely conserved in eukaryotes and BLAST queries (http://apidb.org/apidb/) suggest that it is also expressed in P. falciparum and T. gondii. Therefore it is possible that T. gondii sortilin plays a role in the propeptide dependent exit of proMIC5 from the ER similar to proTxVI. We cannot, however, eliminate the possibility that MIC5 is retained in the ER due to incorrect folding in the absence of its propeptide.
Consistent with a common role in trafficking to the micronemes, we demonstrated that the MIC5 propeptide could substitute for that of M2AP to promote correct delivery to micronemes, enhance the stability of MIC2, and support normal regulated secretion. Since there is no obvious sequence conservation among micronemal propeptides, the structural features necessary to facilitate micronemal trafficking remain unclear. There is, however, striking conservation of P1–P4 residues in the maturation cleavage site of several microneme proteins including M2AP, MIC3, MIC6, and AMA1 (Carruthers, 2004). Interestingly, mutation of these conserved residues in the M2AP cleavage site did not block trafficking but did diminish cleavage (Harper, et al., 2006), paralleling our findings for MIC5. A distinction, however, is that the MIC5 cleavage site does not resemble that of other micronemal proproteins. This, together with our finding that the P1′ residue is necessary for MIC5 processing, suggests that multiple maturases operate within the micronemal pathway. This contention is further supported by the differential effects of protease inhibitors in blocking M2AP and MIC5 maturation (F.P. and V.B. C., unpublished).
The cooperativity among micronemal proteins suggests that a highly complex and interdependent adhesion system exists in T. gondii. Identification of common mechanisms that underlie trafficking of micronemal proteins may lead to interventional strategies capable of disrupting the timely secretion of these key invasion proteins during parasite entry.
Parasites were cultured by passage in primary human foreskin fibroblast cells as described previously (Harper, et al., 2006). For transfection, MIC5 knockout parasites (ΔMIC5, (Brydges, et al., 2006)), were purified by membrane filtration, washed, and resuspended in cytomix buffer (2 mM EDTA, 120 mM KCl, 0.15 mM CaCl2, 10 mM K2HPO4/KH2PO4, 25 mM HEPES, 5 mM MgCl2.6H2O; pH 7.6). Parasites (2×107) were electroporated with 50 μg mutant construct and 10 μg pCAT (also known as pTUB5CATSAG1, D. Soldati, Imperial College, London, U.K.) respectively, using a Bio-Rad GenePulser II (settings 1.5kV, 25 μF, no resistance). After overnight growth, 20 μM chloramphenicol (Sigma) was added to select for pCAT-transfected parasites. Drug resistant populations were cloned by limiting dilution and examined for MIC5 expression by immunofluorescence and western blotting. Single clones were selected for further analysis.
To raise antiserum against the MIC5 propeptide, the complete propeptide sequence (NH2-LASHLRSRHMEAGR-OH) was synthesized and coupled to keyhole limpet hemocyanin carrier protein for injection into rabbits (service provided by Biosynthesis, Inc.) A primary injection of 200 μg conjugated peptide in Freund’s complete adjuvant was followed by four boosts at two-week intervals with 200 μg conjugated peptide in Freund’s incomplete adjuvant. Antibodies were affinity-purified with MIC5 propeptide using the Amino-Link kit from Pierce, according to the manufacturer’s instructions. Monoclonal antibodies 4G1. AH11 (αGRA4, (Labruyere, et al., 1999)), 6D10 (αMIC2, (Wan, et al., 1997)), Tg17-43 (αGRA1, (Cesbron-Delauw, et al., 1989)), rat αMIC5, and rabbit αM2AP (Rabenau, et al., 2001) were used for colocalization experiments and Western blotting.
The wild-type MIC5 ORF with a C-terminal myc tag was used to replace the M2AP ORF in construct pM2AP (Huynh, et al., 2003), creating construct MIC5-myc as follows. PCR primers TgMIC5.162.HindIII.kz.f (GACTAAGCTTCGCCACCATGCTGCGACCTACTGTTC) and TgMIC5.704.XbaI.r (GATCTCTAGATGCGAGTTTCACCTCGGAG), restriction sites in italics, were used to amplify the MIC5 ORF from construct pT-MIC5C1.1 (V. Carruthers, unpublished), minus the stop codon. The resulting PCR product was first subcloned into pGEM-T vector according to the manufacturer’s directions (Promega). pGEM clones were restricted with HindIII and XbaI and resulting inserts were cloned into pBUDCE4 (Invitrogen) vector in front of a myc tag to create pBUDCE4-TgMIC5WT-myc. The complete ORF with the C-terminal myc tag was then PCR-amplified using primers TgMIC5.162.NsiI.f (GATCATGCATCTGCGACCTACTGTTCG) and pBUDCE4.748.PacI.r (GATCTTAATTAAAGATCCTCTTCTGACATG) and cloned into pM2AP, cut with NsiI and PacI.
Propeptide cleavage site mutant constructs MIC5P1, MIC5P1’, MIC5P1P1’, MIC5P2, and MIC5P1P2 were created via PCR in two steps, using pBUDCE4-TgMIC5WT-myc as a template. The signal sequence and propeptide were amplified using TgMIC5.162.NsiI.f and the following reverse primers: TgMIC5.P1.r (ATCCATGGTTCGTGCTCCGGCTTCCATGTGTC), TgMIC5.P1’.r (GTTTTGGGTATCCATGGTTGCTCTTCCGGCTTCCATGTG), TgMIC5.P1P1’.r (GTTTTGGGTATCCATGGTTGCTGCTCCGGCTTCCATGTG), TgMIC5.P2.r (GGTATCCATGGTTCGTCTTGCGGCTTCCATGTG), and TgMIC5.P1P2.r (GGTATCCATGGTTCGTGCTGCGGCTTCCATGTG), respectively, mutations underlined). The mature sequence was amplified with the following forward primers: TgMIC5.P1.f (ATGGAAGCCGGAGCACGAACCATGGATACCC), TgMIC5.P1’.f (CACATGGAAGCCGGAAGAGCAACCATGGATACCCAAAAC), TgMIC5.P1P1’.f (CACATGGAAGCCGGAGCAGCAACCATGGATACCCAAAAC), TgMIC5.P2.f (AGACACATGGAAGCCGCAAGACGAACCATGGATACC), and TgMIC5.P1P2.f (AGACACATGGAAGCCGCAGCACGAACCATGGATACC), respectively, and pBUDCE4.748.PacI.r. In the second PCR step, one μl of each PCR in a pair were mixed together in one reaction and the full-length fusion PCR was amplified with TgMIC5.162.NsiI.f and pBUDCE4.748.PacI.r for cloning into pM2AP.
Construct MIC5Δpro, the complete MIC5 ORF without the propeptide, with a C-terminal myc tag, was created via fusion PCR with pBUDCE4-TgMIC5WT-myc template as follows. The signal sequence was amplified with primers TgMIC5.162.NsiI.f and TgMIC5Δpro.r (TTGGGTATCCATGGTTCGCGCATCAGCAGAGCCGG). The mature protein sequence was amplified with TgMIC5Δpro.f (CGAACCATGGATACCCAAAAC), regions of primer overlap in bold, and pBUDCE4.748.PacI.r. Cloning into pM2AP was performed as above.
All constructs were verified by restriction digests and sequencing. Large-scale DNA preparations of all constructs were made using Qiagen MAXI kits according to the manufacturer’s instructions.
Assays examining constitutive and ethanol-induced micronemal secretion into the excreted/secreted antigen fraction (ESA, i.e., culture supernatant) were performed as described previously (Huynh, et al., 2003). For assays examining the effect of the calcium chelator B-AM, parasites were first subjected to treatment with either 20 μM B-AM or DMSO for 10 min at 18°C before proceeding with a constitutive secretion assay as described.
Immunoblotting was performed as described previously (Wan, et al., 1997). Parasite lysates were made by resuspending parasites in 95°C reducing sample buffer (SDS sample buffer with 2% β-mercaptoethanol) and boiling for 3 min. Lysates were then vortexed, pelleted, and loaded on 12.5% SDS PAGE gels.
Immunofluorescence assays on extracellular parasites were performed as follows. Freshly-lysed tachyzoites were syringed twice each through 20 and 23 gauge needles, filtered and washed with room temperature invasion medium (Delbecco’s modified Eagle’s medium supplemented with 3% FBS and 10mM HEPES). Pellets were resuspended in 0°C IM and an equal volume of 2X fixative (6% formaldehyde, 0.05% glutaraldehyde, in PBS) and fixed for 30 min at 0°C. Fixed parasites were washed twice in 10 ml PBS and finally resuspended in 1 ml PBS. Parasites were allowed to settle for 30 min on chamber slides coated with 0.01% poly-L-lysine (ICN) for 30 min. Staining was performed as described previously (Huynh, et al., 2003).
Toxoplasma-infected fibroblasts were fixed in 4% paraformaldehyde (Electron Microscopy Sciences) in 0.25 M HEPES (pH 7.4) for 1 h at RT, then in 8% paraformaldehyde in the same buffer overnight at 4°C. Cell preparations were scraped, pelleted, and embedded in 10% bovine skingelatin in PBS. Pieces of the pellet were infiltrated overnightwith 2.3 M sucrose in PBS at 4°C, mounted on aluminium studs, and frozen in liquid nitrogen. Blocks were sectioned at −108°Cin an Ultracut cryo-ultramicrotome (Leica). 60-nm thick sectionswere collected using a 1:1 mixture of 2.3 M sucrose and 2% methylcellulose and transferred onto formvar- and carbon-coated nickelgrids. Ultrathin sections were incubated for 10 min with 0.1 M NH4Cl in PBS and for 20 min with 0.5% fish skin gelatin (FSG; Sigma-Aldrich)in PBS (PBS-FSG). To proceed for double immunolabelling, sections were incubated for 30 min at room temperature with rabbit αMIC5propeptide diluted 1/100 in PBS-FSG. After four washes in PBS, the sections were labelled with goat αrabbit antibodies, followed by 10 nm protein A-gold conjugate (Department of Cell Biology, Utrecht University, Netherlands) diluted in PBS-FSG for 30min. Sections were washed again in PBS and fixed with 1% glutaraldehydein PBS for 10 min (to maintain occupancy of protein A binding sites) before labelling with rat αMIC5 diluted 1/100 inPBS-FSG, followed by labelling with mouse αrat antibody and 5 nm protein A-gold. Thesections were then infiltrated with 0.5% uranyl acetate in 1.8% methylcellulose, air-dried, and examined in a Philips CM120 Electron Microscope (Eindhoven, the Netherlands) under 80 kV. Sections stained with each antibody individually showed similar labelling to that of the dual stained samples.
Metabolic labelling, immunoprecipitation, and autoradiography were performed essentially as described (Brydges, et al., 2000). Immunoprecipitation of MIC5 was achieved using 2.5 μl rabbit αMIC5.
To purify myc-tagged MIC5 bands from the TgMIC5P1P1’-myc-expressing parasites for N-terminal sequencing, mouse αmyc monoclonal antibody (K. Kim, Albert Einstein School of Medicine, NY, NY) was coupled to protein G beads (Zymed) using a modified protocol (Harlow and Lane, 1988). Briefly, whole αmyc ascites fluid was mixed with protein G sepharose beads overnight at 4°C, followed by extensive washing with 0.2 M sodium borate, pH 9.0. To cross-link the antibody to the beads, dimethylpimelimidate (Sigma) was added to antibody-coupled beads in sodium borate solution at a final concentration of 20 mM and mixed for 30 min at RT. Following cross-linking, beads were washed into 0.2 M ethanolamine (Sigma), pH 8.0 and mixed for 2 h to quench the reaction. Beads were then washed with PBS and stored at 4°C. Samples were taken before and after addition of DMP for SDS PAGE analysis to determine cross-linking efficiency.
A total of 2×1010 TgMIC5P1P1’-myc filter-purified parasites were solubilized in modified RIPA buffer (50 mM Tris-HCl, pH 7.5, 1% NP-40, 0.5% sodium deoxycholate, 0.1% SDS, 150 mM NaCl; supplemented with DNAse and RNAse) for 2 h on ice. Following removal of insoluble material, the supernatant was aliquoted into batches and mixed with the αmyc-coupled beads overnight at 4°C. The beads were then washed four times with modified RIPA, followed by a PBS wash to flush out remaining detergent. MIC5 bound to the column was eluted in four 1 ml fractions with 0.1M triethylamine (Sigma), pH 11.5. Fractions containing MIC5 were identified by western blot, pooled, and dried by centrifuge concentration. Following resuspension in reducing sample buffer, the entire pooled elution was run on SDS-PAGE and transferred to PVDF membrane (Millipore). Transfer, Coomassie-blue staining, and destaining were done according to the manufacturer’s directions. Protein bands were excised and subjected to Edman degradation as a service of Midwest Analytical, St. Louis, MO.
We thank Claudia Bordon and Tracey Schultz for providing technical assistance, Manami Nishi and David Roos for providing the GRASP parasites, and Gary Ward and Marie-France Cesbron-Delauw for providing antibodies. We also thank Bjorn F.C. Kafsack for help with revisions and Marc Pypaert in the Yale Center for Cell and Molecular Imaging for excellent assistance and scientific comments for electron microscopy. This work was supported by National Institutes of Health Grants AI-060767 (to I.C.) and AI-046675 (to V.B.C).