Molecular dissection of novel trafficking and processing of the T. gondii rhoptry metalloprotease Toxolysin-1

Abstract

Toxoplasma gondii utilizes specialized secretory organelles called rhoptries to invade and hijack its host cell. Many rhoptry proteins are proteolytically processed at a highly conserved SΦXE site to remove organellar targeting sequences that may also affect protein activity. We have studied the trafficking and biogenesis of a secreted rhoptry metalloprotease with homology to insulysin that we named Toxolysin-1 (TLN1). Through genetic ablation and molecular dissection of TLN1 we have identified the smallest rhoptry targeting domain yet reported and expanded the consensus sequence of the rhopty pro-domain cleavage site. In addition to removal of its pro-domain, Toxolysin-1 undergoes a C-terminal cleavage event that occurs at a processing site not previously seen in Toxoplasma rhoptry proteins. While pro-domain cleavage occurs in the nascent rhoptries, processing of the C-terminal region precedes commitment to rhoptry targeting, suggesting that it is mediated by a different maturase, and we have identified residues critical for proteolysis. We have additionally shown that both pieces of TLN1 associate in a detergent resistant complex, formation of which is necessary for trafficking of the C-terminal portion to the rhoptries. Together, these studies reveal novel processing and trafficking events that are present in the protein constituents of this unusual secretory organelle.

INTRODUCTION

Toxoplasma gondii is an obligate intracellular pathogen in the phylum Apicomplexa and is one of the most common parasitic infections of mammals worldwide (1). T. gondii infection can cause severe disease in immunosuppressed patients and is the leading cause of congenital neurological defects in infants (2). Other important human and veterinary pathogens in this phylum include Plasmodium falciparum, the causative agent of malaria, Cryptosporidium spp. a source of severe disease in the immunocompromised, and Neospora caninum, a major cause of abortion in cattle (3). One striking similarity among pathogenic Apicomplexa is their obligate intracellular lifestyle which is accomplished via a refined arsenal of proteins that facilitate invasion and intracellular survival (4).

Invasion and intracellular survival are mediated by proteins secreted from a set of specialized secretory organelles; the micronemes, rhoptries, and dense granules (5, 6). Proteases within the secretory pathway are important for the biogenesis and function of these organelles and their protein contents (7). Proteolytic maturation of micronemal adhesins is critical for their efficient assembly and trafficking (8). Once deployed on the parasite surface, proteolytic shedding of micronemal adhesins is essential for motility and invasion (9). Proteolysis also plays a role in the biogenesis of rhoptry proteins where the serine protease, SUB2, removes the N-terminal domains that are involved in rhoptry targeting (10-14). Finally, a family of dense granule cysteine proteases are essential for parasite growth and are thought to participate in nutrient acquisition in the parasitophorous vacuole (PV) (15). Together, these results demonstrate the important role proteases play in almost every aspect of T. gondii pathogenesis.

While there is an expanding body of research on the secreted serine and cysteine proteases, little is known about T. gondii metalloproteases. M16 proteases are present in all kingdoms except archea (16). The family is characterized by an inversion of the thermolysin zinc-binding motif, HXXEH (where X is any amino acid) (17), and can be divided into 3 subfamilies: A, B, and C (16). M16A and M16C proteases are large (>1000 aa) proteins that can roughly be divided into N- and C-terminal halves connected by a linker region (18, 19) while the two halves of M16B proteases are encoded on separate genes and associate to form heterodimers (20). These proteins form a clamshell-like structure, with each half containing a pair of homologous domains. The two halves of the clambshell open to bind a substrate then shut, enclosing it within the catalytic chamber (21).

The P. falciparum M16C, falcilysin, is the only protozoan M16C protease that has been characterized and it is unique among characterized insulysins in that it has dual localizations and functions within the parasite (22). The characterization of the micronemal M16A protease TLN4 revealed that it is unusual in that it is extensively processed (23), suggesting that its function within the parasite might deviate from uncleaved family members. Intriguingly, another M16A has been identified in the T. gondii rhoptry proteomic analyses (24) and provides an opportunity to further explore the role of secreted metalloproteases in host-parasite interactions.

Rhoptry proteins are deployed into the host cytosol during parasite invasion to facilitate entry and optimize infection (25-28) and thus rhoptry protein targeting is a vital process for this obligate intracellular pathogen. As with other complex secretory systems, T. gondii utilizes sorting signals to dictate differential localization of proteins within the complex secretory milieu. N-terminal pro-domains have been implicated in rhoptry protein sorting and consistent with this many ROPs are processed (11-14, 26, 29, 30). N-terminal rhoptry pro-domain processing is carried out by the rhoptry subtilisin SUB2 (31) which recognizes and cleaves after the consensus sequence SΦXE (in which Φ is hydrophobic and X is any amino acid), first characterized in ROP1 (12, 31). Although only three pro-domain cleavage sites have been confirmed experimentally (11, 31, 32), the presence of this “ROP1-like” site in a processed rhoptry protein has become a predictive tool and many putative cleavage sites have been reported by us and others without experimental confirmation (13, 26, 29, 30). Interestingly, failure to remove the pro-domain does not disrupt targeting (11, 32), suggesting a paradigm in which the pro-domain may serve a dual role within the secretory pathway, both directing rhoptry zymogens to the correct organelle and preventing premature activation of the effector where it might be detrimental to the parasite. Other rhoptry proteins do not appear to be processed and the mechanism by which they are targeted to the rhoptries is unknown. Regardless of targeting mechanism, most of the known rhoptry proteins are secreted into the host cell (4, 33) where they have been shown to modulate host cell functions (25, 34-37).

We show here that the T. gondii zinc metalloprotease, Toxolysin-1 (TLN1), is a soluble rhoptry protein that is secreted during invasion. We generated a Δtln1 strain and utilized the immunologically blank background provided by the knockout as a valuable tool for studying the trafficking and processing of this parasite protease. We show that TLN1 is targeted to the rhoptries via an N-terminal pro-domain which is processed at a ROP1-like cleavage site with a novel P1 residue. We also show that proteolytic maturation of TLN1 at the C-terminus removes a conserved domain and that this cleavage event occurs at a site not previously seen in T. gondii rhoptry proteins. In addition, TLN1 C-terminal processing occurs earlier in the secretory pathway than known processing events, together suggesting that cleavage occurs via a protease distinct from SUB2. We have determined that this cleavage event is not the result of autocatalysis and have identified specific residues that are important for maturation. Additionally, we demonstrate that that the C-terminal domain, once liberated, associates tightly with the N-terminal domain. Like other insulinases, TLN1 forms homo-oligomers, but unlike many insulinases TLN1 undergoes proteolytic processing. The insertion of a repeat containing region and the occurance of a cleavage event between two conserved domains involved in substrate recognition and binding sets TLN1 apart from other M16A family members. These features may be an evolutionary adaptation in order to alter the substrate repetoire of this T. gondii peptidase.

RESULTS

Toxolysin-1 is a soluble rhoptry zinc metalloprotease

Our proteomic analysis of a highly purified fraction of rhoptries from Toxoplasma identified thirty-eight novel proteins (24). Given the importance of proteases in Toxoplasma infections (38-40), we were particularly intrigued with one protein that has homology to insulysin, the most well known member of the M16 family of zinc metalloproteases. The complete cDNA of the metalloprotease predicts a protein with a mass of 181 kDa, an N-terminal signal peptide (41), the four conserved domains common to M16A proteases, and a series of 13 tandem repeats of the sequence YPDDLPTSSTP (Fig. 1A). Domain 1 contains the active site, characterized by the conserved zinc-coordinating motif HXXEHX₆₉EX₆E while domains 2 and 4 have homology to the M16A inactive catalytic domain of insulysin-like proteases as suggested by the NCBI Conserved Domain Database (42, 43). (Fig. 1A). The presence of all necessary structural domains in addition to the conserved zinc coordinating residues (44), suggests that TLN1 is a functional M16A protease. Due to its conserved domain structure and similarity to insulysin metalloproteases and following the nomenclature of the Plasmodium falciparum insulysin-like protease falcilysin, we named the protein Toxolysin-1 (for Toxoplasma Insulysin-like protein 1, TLN1).

Figure 1.

TLN1 is a rhoptry zinc metalloprotease that is soluble and secreted. (A) Diagram showing the features of TLN1 including a predicted signal peptide, the zinc-binding domain (HXXEH) necessary for catalysis, and the characteristic M16 active and inactive domains associated with M16A proteases. Colored lines demark the four domains that form the clamp-like structure seen in M16A proteases. In addition, TLN1 has 13 highly charged tandem repeats of 11 amino acids. (B) TLN1 colocalizes with ROP13 in the rhoptries and (C) is secreted during invasion as seen by its presence in or on the parasitophorous vacuole (PV) 3 hours post invasion. ROP13 staining in the rhoptries colocalizes with the internal TLN1 stain and both proteins can be seen in or on the vacuole. Scale bar = 5 μm. (D) Analysis of TLN1 coding sequence by several transmembrane prediction algorithms returned variable results. E) TX-114 extraction of parasite lysates confirms that TLN1 is not an integral membrane protein as the bulk of the protein fractionates with the known soluble rhoptry protein, ROP1, and not the GPI anchored protein SAG1. T = total lysate, M = membrane fraction, S = soluble fraction.

To determine the localization of TLN1, we generated specific antiserum against residues 687-1213 of the protein (Fig. 1A). Anti-TLN1 serum stains club-shaped structures in the apical end of the parasite. The signal overlaps well with that from the known rhoptry body marker, ROP13, demonstrating that this protein is indeed a rhoptry protein (Fig. 1B). We performed further in silico analysis of TLN1 to look for predicted transmembrane domains (TMs) and obtained different results depending on the algorithm used (Fig. 1D) (45-47). To clarify this we fractionated T. gondii lysates using TX-114 phase separation and probed the membrane and soluble fractions for TLN1. The majority of TLN1 fractionated with the known soluble protein ROP1 (Fig. 1E), indicating that TLN1 is a soluble rhoptry protein.

All secreted rhoptry proteins are believed to access the host cytoplasm where they likely execute effector functions for optimal hijacking of the host cell. To determine if TLN1 is secreted, we examined newly invaded parasites for the presence of TLN1 in the nascent parasitophorous vacuole. TLN1 was detected in the parasitophorous vacuole along with ROP13 (11), indicating that it is secreted from the rhoptries during host cell invasion (Fig. 1C) and is thus likely to access the host cell.

TLN1 is not essential for growth in vitro or virulence in vivo

To directly address the function of TLN1 in invasion, intracellular survival and virulence, we disrupted its gene using homologous recombination in T. gondii (Fig. S1). Δtln1 parasites display a mild growth defect in vitro (Fig. S1) but no other changes are apparent either in vitro or in in vivo virulence (data not shown). While an in vivo virulence phenotype may be masked by the hypervirulent Type I background (48) in which the knockout was generated, these results demonstrate that TLN1 is not essential for growth in rich media conditions in vitro and further validate the specificity of our TLN1 antibodies. In addition, Δtln1 parasites provide an immunologically blank background in which we can explore trafficking, processing, and oligomerization of TLN1 in the absence of the endogenous protein.

TLN1 has a large cleavage event and aberrant migration

TLN1 encodes a protein with a predicted molecular weight of 181 kDa. While Western blot analysis of T. gondii lysates detects a protein at ~180 kDa, a longer exposure shows an additional, less abundant species at ~250 kDa (Fig. 2A). This banding pattern is consistent with rhoptry proteins that undergo proteolytic maturation (13), with the larger species representing the uncleaved, pro-form of the protein. Migration of this predicted 181 kDa pro-protein at ~250 kDa could be due to the presence of the tandem repeats which have been shown in other rhoptry proteins to substantially retard migration in SDS-PAGE (12, 49). This is supported by exogenous expression of pro-TLN1-6xHis (lacking a signal peptide) in E. coli, which also migrates at ~250 kDa (Fig. 2B). We additionally performed pulse-chase labeling of TLN1 in parasite lysates (Fig. 2C) and confirmed that TLN1 is proteolytically processed during protein maturation to produce a substantial size shift between the pro- (TLN1₂₅₀) and mature (TLN1₁₈₁) forms of the protein.

Figure 2.

TLN1 migrates aberrantly and undergoes a large cleavage event. (A) Western blot of T. gondii lysates with TLN1 antiserum shows migration of the dominant TLN1 species at the predicted size of ~180 kDa (TLN1₁₈₁) while a longer exposure reveals a less abundant pro-form migrating at ~250 kDa (TLN1₂₅₀). (B) Exogenous expression of TLN1_35-1645-6xHIS in E. coli shows that the full length protein (minus the predicted signal peptide) migrates at ~250 kDa, similar to that seen for TLN1₂₅₀. (C) Pulse-chase of TLN1 confirms the presence of a large cleavage event. (D) Coomassie stained gel of separated products of TLN1 IP from T. gondii lysates. TLN1 antiserum precipitates both species of TLN1. (E) The ~180 kDa band was excised and used for tryptic MS/MS. Peptide hits covered the first three quarters of the protein including the active site (further detailed in Supplemental Fig. 3).

Toxolysin-1 is substantially processed at the C-terminus

Many rhoptry proteins are synthesized as pre-pro-proteins and their pro-domains are typically at the N-terminus (11-13, 26, 30, 50). As removal of ~50 kDa from either end could affect proteolytic activity, it is important to determine where TLN1 is cleaved. To this end, anti-TLN1 antiserum was used to immunoprecipitate (IP) TLN1 from parasite lysates. Both the pro (TLN1₂₅₀) and mature (TLN1₁₈₁) forms of TLN1 were detected in the eluted products separated by SDS-PAGE (Fig. 2D). TLN1₁₈₁ was excised from the gel and analyzed by tandem mass spectrometry (MS/MS). Forty-one unique tryptic peptides were returned spanning residues 84-1287 (Fig. 2E, Fig. S2) and were conspicuously absent from the C-terminal 358 residues. This analysis of TLN1₁₈₁ suggests that, in addition to a potential small processing event at the N-terminus, TLN1 undergoes a large processing event at the C-terminus.

TLN1 has an N-terminal pro-domain that is cleaved at a non-canonical cleavage site

To explore potential processing in the N-terminal region of TLN1, we engineered an expression construct containing the full-length TLN1 cDNA driven by its endogenous promoter. A FLAG tag was introduced just downstream of the predicted signal peptide cleavage site between residues 35 and 36 (41) and the construct was stably expressed in Δtln1 parasites (Fig. 3A). TLN1 staining in parasites expressing the FLAG-TLN1 construct reveals that TLN1 rhoptry localization is unaltered in the presence of the FLAG tag (Fig. 3B). Detection of the protein with anti-FLAG antibodies, however, shows that the FLAG-tagged portion of TLN1 localizes to punctate dots at the apical end of the parasite. As this pattern is consistent with staining the pro-domain of rhoptry proteins, we assessed colocalization of the FLAG signal with the pro-rhoptry marker pro-ROP4 (13) (Fig. 3C). Colocalization of FLAG staining with pro-ROP4 demonstrates that the extreme N-terminus of TLN1 contains a pro-domain that is cleaved during parasite maturation in addition to the large C-terminal cleavage event.

Figure 3.

TLN1 has an N-terminal pro-domain. (A) Schematic of TLN1 N-terminal tagging and mutagenesis. A FLAG tag was inserted two amino acids downstream of the predicted signal peptide cleavage site in the TLN1 expression vector and expressed in Δtln1 parasites. Diagramed is the sequence surrounding the inserted N-terminal FLAG tag and the pro-domain with putative cleavage site (SFVD). Amino acids are numbered and residues surrounding the cleavage site are labelled P4-P1 and P1’-P4’. Mutagenesis of the putative cleavage site is diagramed in magenta. (B) TLN1 staining of FLAG-TLN1 parasites confirms that pro-domain staining is specific for the FLAG tag and is not due to mislocalization of the entire protein. (C) FLAG staining of FLAG-TLN1 parasites colocalizes with the pro-rhoptry marker, pro-ROP4, indicating that the tag was inserted in an N-terminal pro-domain. (D) The putative cleavage site (SFVD) was mutated to disrupt the critical P1 residue in the context of FLAG-TLN1 coding sequence (FLAG-TLN1-D55A) and expressed in Δtln1 parasites. TLN1 staining is unchanged, indicating that the mutation did not disrupt targeting of the full length protein. (E) FLAG staining of FLAG-TLN1-D55A parasites colocalizes with the rhoptry marker ROP7. Disruption of processing results in detection of the pro-domain in the mature rhoptries as shown by colocalization with ROP7. (F) Schematic of TLN1_1-59-mCherry. The N-terminal pro-domain including the P4-1 and P1’-P4’ were fused in frame with mCherry and expressed in Δtln1 parasites. mCherry fluorescence colocalizes with the rhoptry marker, ROP7, indicating that TLN1_1-59 is sufficient to target a heterologous protein to the rhoptries. Scale bar = 5 μm.

Although we demonstrated the presence of an N-terminal pro-domain, there are no ROP1-like sites (SΦXE) in the N-terminal region of TLN1. There is, however, SΦXD (aa 52-55), 18 aa downstream of the predicted signal peptide cleavage site and upstream of the first tryptic peptide of mature TLN1 at residue 84 (Fig. 3A). While substitution of glutamic acid to aspartic acid is considered to be highly conserved, it has not been demonstrated in any known or predicted rhoptry protein cleavage sites. In addition, previous work has shown that mutation of the P1 residue of the ROP1 pro-domain cleavage site from glutamic acid to aspartic acid results in a significant inhibition of pro-domain cleavage (32). Thus, to test whether cleavage occurs at this site in TLN1, we mutated the P1 aspartic acid to an alanine (SΦXA) in the context of FLAG-TLN1 (FLAG-TLN1-D55A) and expressed the mutant protein in Δtln1 parasites (Fig. 3A). Staining of Δtln1:FLAG-TLN1-D55A parasites with anti-TLN1 antiserum confirms that the mutant protein traffics normally (Fig. 3D). Previous experiments of this kind have assessed disrupted pro-domain cleavage by a reduction in the mobility of the mutated protein by SDS-PAGE (11, 32), but a shift of 22 amino acids out of 180 kDa is not visible by Western blot. We were able to exploit the fact that uncleaved pro-domains can be detected in the mature rhoptries along with the unprocessed protein (11, 32). Thus if the TLN1 N-terminal pro-domain is cleaved at SΦXD and we disrupt processing, the FLAG-tagged pro-domain will remain attached and be detectable in the mature rhoptries. Confirming our hypothesis, staining of Δtln1:FLAG-TLN1-D55A with FLAG antiserum shows that the FLAG epitope is now present in the mature rhoptries (Fig. 3E), where in the native context it was confined to the pro-rhoptries. The presence of the FLAG epitope in the mature rhoptries along with the mature protein indicates that processing of the TLN1 N-terminal pro-domain is blocked by the D55A mutation and the uncleaved protein is targeted properly, as seen for ROP1 (32). These data demonstrate that TLN1 utilizes a P1 aspartic acid in its ROP1-like cleavage site.

TLN1 N-terminal pro-domain is a rhoptry sorting signal

The ROP1 pro-domain is sufficient to traffic a heterologous protein to the rhoptries (10, 51). To determine if the TLN1 N-terminal pro-domain also functions as a targeting signal, we fused amino acids 1-59 (the signal peptide, N-terminal pro-domain, and the P1-P4 and P1’-P4’ residues of the N-terminal cleavage site) to the reporter protein mCherry (TLN1_1-59-mCherry). Fluorescence in Δtln1 parasites transfected with TLN1_1-59-mCherry colocalizes with ROP7 in the rhoptries (Fig. 3F), demonstrating that the TLN1 N-terminal pro-domain is sufficient to target a heterologous protein to the rhoptries. At 22 amino acids in length the TLN1 N-terminal pro-domain is substantially smaller than the ~40 residues seen in other rhoptry pro-domains. It also bears little resemblance to the other identified rhoptry pro-domains in either its primary sequence or predicted secondary structure (Fig. S3).

TLN1 is cleaved into N- and C-terminal fragments during protein maturation

Mass spectrometry of TLN1₁₈₁ suggests a large C-terminal processing event. Alignment of TLN1 to human and rat IDE shows that the conserved domain 4 aligns with the C-terminal region of TLN1 removed during protein maturation (21). Removal of this domain from other insulysins is detrimental to their activity and such a loss would be expected to impact the ability of TLN1 to function as a protease (19). To explore the fate of the processed TLN1 C-terminal domain (TLN1-C), we generated polyclonal antiserum against the C-terminal region (residues 687-1213, anti-TLN1-C). Western blot analysis of T. gondii lysates with anti-TLN1-C antiserum shows that the C-terminal portion of TLN1 is intact and migrates at ~50 kDa (Fig. 4A) consistent with the size shift observed between the TLN1₁₈₁ and TLN1₂₅₀. IFA of intracellular parasites with anti-TLN1-C sera shows colocalization with the rhoptry marker ROP13, demonstrating that the C-terminal portion of TLN1 also localizes to the rhoptries (Fig. 4B). These data demonstrate that, while TLN1-C is cleaved from the larger precursor protein during maturation, both pieces are retained and trafficked to the rhoptries.

Fig 4.

The C-terminal domain of TLN1 (TLN1-C) is intact and localizes to the rhoptries. (A,B) TLN1-C is detected by Western blot migrating at ~50 kDa in WT but not Δtln1 lysates (ROP7 is a loading control) and colocalizes with ROP13 in the parasite rhoptries, demonstrating that it is maintained within the parasite after cleavage from TLN1-N. Scale bar = 5 μm. (C) The last peptide from the TLN1-N MS/MS and the first peptide from the TLN1-C MS/MS are highlighted on the relevant portion of TLN1 sequence. A potential ROP1-like cleavage site is indicated with a closed arrowhead and the putative RAMA1-like cleavage site is indicated with an open arrowhead. (D) Mutation of the P1 residue in the putative ROP1-like site between TLN1-N and TLN1-C tryptic peptide hits does not abolish cleavage. The arrow indicates the ~250 kDa pro-form of TLN1 in WT and Δtln1:TLN1-Q1306A lysates. (E) Cleavage of TLN1-C is not an autocatalytic event as mutations abolishing zinc coordination (TLN1-E129A,H130D) do not prevent maturation of TLN1. TLN1-E129A,H130D was overexpessed in the Δtln1 background. (F) Mutation of Ser1303 and Leu1305 significantly reduce cleavage of TLN1-C. An increase in the abundance of TLN1₂₅₀ and near disappearance of TLN1₁₈₀ in the TLN1-S1303A,L1305A double mutant compared to the wild-type demonstrate the importance of these residues in site recognition or cleavage.

The TLN1 C-terminal pro-domain is processed at a novel cleavage site

To more accurately define TLN1-C, the TLN1 expression construct was modified with a tandem affinity purification tag at the extreme C-terminus. The tag consists of two Strep tags, a FLAG tag, and a 6xHIS tag (TLN1-SFH) (Fig. S4A). TLN1-SFH is cleaved normally (Fig. S4B) and both portions traffic to the rhoptries (Fig S4C, S4D). We purified TLN1-C-SFH from parasite lysates using nickel agarose chromatography (Fig. S4E) and analyzed this fragment by MS/MS. The N-terminal most tryptic peptide detected begins at residue 1308 (Fig. 4C) and peptides were found throughout TLN1-C (S4A). There are only 21 amino acids between the last TLN1-N peptide and the first TLN1-C peptide and thus the cleavage site must lie within those residues (Fig. 4C). While there are no canonical ROP1-like cleavage sites in this region, a potential site with a non-canonical P1 residue, SΦXQ, is present (Fig. 4C). As we have already shown that TLN1 tolerates a non-canonical P1 residue in its N-terminal ROP1-like site, we mutated the putative P1 glutamine to alanine in the context of the TLN1 expression vector (TLN1-Q1306A) and expressed the construct in Δtln1 parasites. Δtln1:TLN1-Q1306A lysates probed with TLN1 antiserum displayed normal cleavage of TLN1-C (Fig. 4D). These data indicate that the maturation of TLN1-C takes place at a novel cleavage site.

Mutation of specific residues reduces cleavage of TLN1-C

Because TLN1-C processing does not take place at a ROP1-like site, we inspected the sequence between the known portions of TLN1-N and -C for cleavage sites identified in other systems. In Plasmodium falciparum, the rhoptry protein RAMA1 undergoes proteolytic maturation at the cleavage site SFL↓Q (52). Though the PfRAMA1 site resembles a ROP1-like site, the difference in the chemical nature of the P1 residues makes it unlikely that the rhoptry subtilisin is capable of cleaving both of these substrates. While P1 residues are frequently important for cleavage by a protease, nothing is known about the residues that are important for cleavage at a RAMA1-like site. Alignment of RAMA1 sequences across Plasmodium species suggest that the P3 and P1 are conserved (SXL↓) (52, 53) and so we explored their importance for cleavage in the context of TLN1-C. We mutated the TLN1-SFH serine at position 1303 (S1303) and leucine at 1305 (L1305) to alanines (TLN1-S1303A,L1305A, Fig. 4F) and transfected the mutant construct into Δtln1 parasites. Western blot analysis of Δtln1:TLN1-S1303A,L1305A lysates demonstrate that, unlike the TLN1-Q1306A mutation, the TLN1-S1303A,L1305A mutations substantially reduced processing of TLN1-C (Fig. 4F) as observed by a dramatic increase in the relative amount of unprocessed TLN1. This demonstrates that TLN1 S1303 and L1305, but not Q1306, are important residues for cleavage of TLN1-C. These data provide the first evidence for Toxoplasma rhoptry protein processing at a non ROP1-like site as well as demonstrate the importance of cleavage site-proximal residues for proteolysis.

TLN1 processing is not autocatalytic

Many proteases are known to participate directly in their own maturation (15, 31, 54, 55) and since TLN1-C is cleaved at a novel site, we explored the possibility that TLN1 is catalyzing its own proteolytic maturation. We generated a TLN1 mutant in which the zinc coordinating amino acids of the active site were altered to abolish catalytic activity (TLN1-E129A,H130D). Western blot analysis shows that this catalytically inactive TLN1 mutant is processed normally (Fig. 4E), demonstrating that cleavage of the C-terminal domain is not carried out by TLN1. Together with the identification of residues important for cleavage, these data strongly suggest the presence of an additional rhoptry protein maturase in the secretory pathway.

TLN1-C cleavage occurs prior to rhoptry commitment

SUB2 cleavage of N-terminal pro-domains is thought to take place within the immature rhoptries of nascent daughter parasites and consistent with this, SUB2 is a rhoptry localized protease. If TLN1-C cleavage is also the result of a rhoptry protease, maturation should occur in the immature rhoptries. To examine the timing of TLN1-C cleavage, we fused TLN1_1-59-mCherry to TLN1-C (creating TLN1_1-59-mCherry-C_1287-1645) (Fig. 5A) and transfected the construct into Δtln1 parasites. mCherry is sorted to the rhoptries as is TLN1_1-59-mCherry (Fig. 5B) and Western blot analysis demonstrates that TLN1-C cleavage from TLN1_1-59-mCherry occurs and results in a TLN1-C reactive product of the same apparent molecular weight as wild-type parasites (Fig. 5C). Surprisingly, TLN1-C staining is not seen in the rhoptries of TLN1_1-59-mCherry-C_1287-1645 parasites. Instead, TLN1-C localizes to the parasitophorous vacuole (Fig. 5D), even though it is synthesized on the same peptide chain as mCherry, which is sorted to the rhoptries. These data indicate that TLN1-C is cleaved prior to rhoptry commitment and does not contain its own rhoptry targeting information.

Fig 5.

TLN1-C is cleaved before commitment to rhoptry sorting. (A) Schematic of TLN1_1-59-mCherry-C_1287-1645. TLN1-C was fused in frame to the rhoptry-targeted mCherry construct described in Fig 4F. (B) TLN1_1-59-mCherry-C_1287-1645 parasites display mCherry fluorescence colocalizing with ROP7 in the rhoptries, indicating the TLN1-C does not alter rhoptry sorting. (C) TLN1_1-59-mCherry-C_1287-1645 is processed similarly to WT TLN1 with TLN1-C migrating at the same apparent size as that from WT parasites. (D) TLN1-C staining in Δtln1:TLN1_1-59-mCherry-C_1287-1645 parasites does not colocalize with mCherry in the rhoptries. Instead, it is present in the PV, indicating that after cleavage from the rhoptry targeting construct, TLN1-C is secreted via the dense granules. Scale bar = 5 μm.

TLN1-N and TLN1-C associate in a detergent-resistant complex and oligomerize

Since both TLN1-N and TLN1-C are intact and present in the rhoptries of wild-type parasites, we hypothesized that an association between the two pieces could be responsible for rhoptry sorting of TLN1-C. Consistent with this hypothesis, when we affinity purify TLN1-C-SFH we find that TLN1-N co-precipitates with the target. This interaction is specific as the soluble rhoptry protein ROP7 does not co-precipitate (Fig. 6A). While strong, this association is not due to disulphide bonding as a non-reducing Western blot of TLN1-N and TLN1-C shows both the N- and C-terminal fragments migrating similarly in reducing conditions, and not at the combined apparent mass of ~250 kDa (Fig. 6B). Even though cleavage of TLN1-C removes a critical structural domain from the protein, the tight association of TLN1-N with TLN1-C indicates that all four domains are present in the mature enzyme, allowing TLN1 to retain all the structural components necessary for proteolytic function in other insulysins.

Fig 6.

TLN1-N and TLN1-C associate in a detergent resistant complex and form multimers. (A) IP of TLN1-SFH in which TLN1 is tagged with a Strep-Flag-His tandem affinity tag on the extreme C-terminus and precipitated with nickel agarose in RIPA lysates. Eluted products were probed for TLN1-C with FLAG antiserum and with TLN1. Both portions of TLN1 were present in eluted material, indicating that they form a detergent-resistant complex. ROP7 is not detected in the eluted material, demonstrating the specificity of the co-IP. (B) The TLN1-N/TLN1-C interaction is not the product of a disulfide bond as neither species demonstrate a major migratory shift under non-reducing conditions. We observe a slight reduction in the apparent mass of TLN1-C in non-reducing conditions consistent with the presence of multiple cysteins within the domain. (C) TLN1-SFH lysates were mixed with untagged lysates and TLN1-C was immunoprecipitated with nickel agarose in RIPA. Both the native and slower migrating TLN1-C-SFH were present in eluted materials, along with TLN1-N. SAG1 was not present in the eluted products, indicating that the IP was specific.

Mammalian insulysins are known to form higher order structures with each other (19), an interaction that is mediated by domain 4 and is an important regulator of catalytic function among these enzymes (56). To ascertain if TLN1 is also able to form higher order structures, and thus subject to similar regulatory mechanisms as mammalian insulysins, we expressed TLN1-SFH in wild-type parasites and immunoprecipitated TLN1-C-SFH. Western blot analysis of the eluted material revealed not only the presence of the target ~55 kDa TLN1-C-SFH band, but also the smaller native TLN1-C band (Fig. 6C). These data indicate that TLN1 forms homo-oligomers comprised of at least two copies of TLN1-N and TLN1-C and thus may also utilize allosteric activation by binding partners as a regulator of catalytic activity.

DISCUSSION

Toxolysin-1 is the first rhoptry metalloprotease to be identified among the Apicomplexa (24), the study of which has revealed new cell biology of rhoptry protein maturation and may have implications for insulysin function within Toxoplasma. Like ROP1, Toxolysin-1 has an N-terminal pro-domain that is sufficient for rhoptry trafficking. Unlike ROP1, however, the TLN1 pro-domain is cleaved after aspartic acid. The predictive value of the SΦXE consensus sequence is such that N-terminal pro-domain cleavage sites are being suggested without experimental confirmation (13, 14, 26, 28, 29, 57). Although aspartic acid is poorly tolerated in the context of the ROP1 cleavage site (32), we show here that it is a valid P1 residue in other proteins such as TLN1. These data expand the repertoire of residues that are permitted in the P1 position of ROP1-like cleavage sites, allowing us to redefine the ROP1-like motif as SΦX(E/D). In light of these data, the predicted N-termini of processed rhoptry proteins should be reexamined for ROP1-like sites with aspartic acid as a P1 residue that are consistent with the pro-mature size shift observed by Western blot. For example, the pro-domain of ROP18, an important virulence factor, has an SΦXD immediately upstream of the predicted SΦXE cleavage site (26) (Fig. S3) and the size difference between cleavage at these sites would not be visible by SDS-PAGE. Similarly, ROP17 contains a candidate SΦXD and its pro-domain has not been experimentally determined (30).

The identification of the N-terminal TLN1 pro-domain also contributes to our understanding of the components necessary for their function as rhoptry sorting signals. Analysis of rhoptry pro-domains has been complicated as most rhoptry proteins studied are the large family of ROP kinases that likely arose from gene duplication events (28), limiting their utility in searching for commonalities within pro-domains. The addition of the TLN1 N-terminal pro-domain to those of ROP1, ROP13, and ROP2 (the founding ROP kinase member) has the potential to reveal commonalities within the sorting signals without the bias of common ancestry. However, lack of apparent similarities at the primary sequence level (Fig. S3) suggests that the recognition of these peptides as sorting signals likely involves tertiary structural elements. The small size of the TLN1 N-terminal pro-domain will be useful for delimiting regions necessary and sufficient for rhoptry targeting.

The separation of TLN1 into N- and C-terminal domains is an interesting phenomenon in the biologies of both the parasite and insulysins. We have shown here that TLN1-C cleavage does not occur at SΦXE↓ and may occur at SXL↓Q instead. Additionally, mutation of the P3 serine and P1 leucine of the putative cleavage site are sufficient to inhibit the majority of TLN1-C cleavage. This data represents a novel rhoptry protein maturation event in Toxoplasma and the only analysis of residues that are important for this cleavage.

The SXL↓Q cleavage event was first described in P. falciparum (52) and its presence in a Toxoplasma protein suggests that the maturase responsible is conserved across the phylum. While Plasmodium also has a rhoptry subtilisin (PfSUB1), recent work suggests that its consensus sequence is highly variable (58). The promiscuity of the PfSUB1 consensus sequence allows for the possibility that SUB2 is capable of cleaving TLN1-C, but the earlier timing of the cleavage event and strong conservation of the SΦXE/D consensus among other ROPs does not support this hypothesis. Instead, cleavage of TLN1-C prior to rhoptry commitment in conjunction with proteolysis at a novel site suggests the presence of an additional rhoptry maturase in the secretory pathway.

Cleavage of TLN1-C removes a conserved functional domain (19) and we show here that this domain is intact and trafficked to the rhoptries. We demonstrate that TLN1-N and TLN1-C associate in a detergent resistant complex and that, lacking TLN1-N, TLN1-C is not trafficked to the rhoptries. Together, these data indicate that TLN1-C is reliant on its association with TLN1-N for rhoptry targeting.

The tight association between TLN1-N and TLN1-C also has functional implications. Toxoplasma gondii is the only known species in which the functional domains of an insulysin are separated by proteolysis. Recent work on Toxolysin-4 demonstrates that this micronemal insulysin undergoes extensive proteolysis and associates in a complex (23). In combination with TLN1, these reports suggest that Toxoplasma insulysin biogenesis diverges from that of other family members, a change that may result in novel functionalities among the otherwise highly conserved M16As. In the case of TLN1, the tight association between the two portions of the protein indicate that all the structural components necessary for activity are present in the mature enzyme. Furthermore, we demonstrate that differently tagged species of TLN1 co-precipitate, indicating the formation of homo-multimers. In mammalian insulysins, formation of these complexes is mediated by domain 4 and is an important regulator of allosteric activation for the enzyme (56). The existence of TLN1 multimers supports the hypothesis that the Toxoplasma enzyme is functional despite the proteolytic processing of its C-terminal domain. As these data suggest the existence of a complete enzyme, we speculate that the cleavage of domain 4 from the rest of the protein may serve a novel functional role such as allowing the enzyme to bind larger or more diverse substrates.

Structural studies of several M16A and C enzymes have indicated that these proteases are limited to substrates of a certain size because they must enclose the peptide within the catalytic chamber in order to cleave the scissile bond (21). M16B protease complexes, on the other hand, are unable to close completely due to the presence of a glycine-rich loop (18). This G-loop provides steric hindrance at the mouth of the clamp, preventing it from closing and allowing larger substrates to protrude from the catalytic chamber. While TLN1 lacks a G-loop, the separation of Domain 4 could provide additional flexibility upon substrate binding, allowing the enzyme to cleave larger target proteins. It is also possible that the tandem repeats in TLN1-N may contribute to substrate binding or they might function as a G-loop themselves. Identifying targets of TLN1 will be the best way to reveal the purpose of the TLN1-C cleavage event.

While we cannot rule out the possibility that TLN1 cleaves one or more parasite substrates, it is likely that its target is a host protein. Rhoptry proteins are increasingly being recognized as important effectors in the parasite’s arsenal against the host cell. Although our attempt at exploring its function by genetic ablation of TLN1 did not yield an observable phenotype, pathogens frequently have multiple levels of redundancy built in to important processes (59). While there are two other predicted M16A proteases with signal peptides in the T. gondii genome, their transcriptional profiles (60) do not suggest that they are rhoptry proteins. It is also possible that TLN1 plays an important role in a portion of the parasite’s life cycle that was not examined here or that it confers a minor advantage that is not easily identified but would be sufficient to assure its propagation in the parasite over time. In any of these cases, identification of the TLN1 substrate within the host or the parasite will play an important role in elucidating its function.

MATERIALS AND METHODS

Host cell and parasite cultures

T. gondii tachyzoites were maintained by serial passage in confluent HFF monolayers as previously described (61).

Antibodies

Primary antibodies used in Western blot analysis and immunofluorescence assays (IFAs): mouse anti-TLN1, mouse anti-TLN1-C, anti-pro-ROP4 UVT70 (13), anti-ROP7 1B10 (62), rabbit anti-SAG1 (63), rabbit anti-ROP13 (24), anti-ROP2/3/4 4A7 (29), anti-ROP1 TG 49 (64), rat anti-FLAG (65), mouse anti-HIS (Sigma).

In silico analyses

The TLN1 coding sequence was analysed for conserved domains by the NCBI Conserved Domain Database (42, 43), a signal peptide by SignalP (41), predicted transmembrane domains by TMPred (45), DAS (46), and TMHMM (47) algorithms. Alignment of TLN1 with human and rat IDE utilized the COBALT (66) multiple alignment algorithm.

Generation of polyclonal antisera against recombinant portions of TLN1

Anti-TLN1 and anti-TLN1-C polyclonal antisera were generated against residues 687-1213 (Primers P1 and P2, Table 1) and 1303-1645 (P3 and P4), respectively. All constructs were amplified from cDNA and cloned into pET161-GW-D-TOPO (TLN1) or pET101-D-TOPO (TLN1-C) (Invitrogen). Recombinant proteins were purified under denaturing conditions according to the manufacturer’s guidelines (Qiagen). Purified recombinant proteins were dialyzed against PBS and concentrated. Antisera were generated in BALB/c mice as previously described (62).

Table 1. Oligonucleotide primers utilized in this study.

All primers are in the 5’-3’ orientation.

Name	Primer Sequence
P1	CACCGTGCTGACACTCTCTGCGGCA
P2	AACTGGCCGTAGAGCCTGCGC
P3	CACCATGTCATTTCTGCAACGCCAAGGC
P4	ATCAGATTCCATATCAGGGACAG
P5	GGCCGTTTAAACCACTGACATTGGACAGCAGCC
P6	TACTGCTGCATCCGCAAGTCTAGATCTGGCC
P7	TAGCTAGCTCACGCAACACG
P8	TAGGACGGCGGTCAAAAAGG
P9	GGCCAGATCTATGAAGCAGGGAACGACCCGC
P10	GGCGCGCCTTAATCGGATTCCATATCAGGGACAGTGATGTTCGACAG
P11	CATGGCGCGCCATCAGATTCCATATCAGGGACAG
P12	ACTAGGCGCGCCATGGTCTCATCCTCAGTTTGAAAAA
P13	GCATGGCGCGCCTTAGTGATGATGGTGGTGATGACTACCGAATCCTTTATCGTCGTCGTCTTTGTAATC
P14	TCTTCAGGCAGTCGGGATTACAAGGATGACGATGATAAAGTCCTTTTTGACCGC
P15	GCGGTCAAAAAGGACTTTATCATCGTCATCCTTGTAATCCCGACTGCCTGAAGA
P16	CTGGCCCATTTTACAGCTGACATGCTGTTTCAGGGA
P17	TCCCTGAAACAGCATGTCAGCTGTAAAATGGGCCAG
P18	GCAACTTCCTTTGTCGCCTCCACCAGCGGTCAG
P19	CTGACCGCTGGTGGAGGCGACAAAGGAAGTTGC
P20	ATTGAACAAGGTCCCGCATTTGCGCAACGCCAAGGCGA
P21	TCGCCTTGGCGTTGCGCAAATGCGGGACCTTGTTCAAT
P22	GGTCCCTCATTTCTGGCGCGCCAAGGCGAGTAT
P23	ATACTCGCCTTGGCGCGCCAGAAATGAGGGACC
P24	GGTAGCTAGCTCACGCAACACGCGGGAGTTG
P25	CATCGGACCGCTCCGACTGCCTGAAGAACAATC
P26	CCATCGGACCGCGGTAGTTTCTACTGTCG
P27	GCCGGTCCGATGGTGAGCAAGGGCGAGG
P28	TTAATTAACTTGTACAGCTCGTCCATGC
P29	CTGGGCGCGCCTTACTTGTACAGCTCGTCCATGC
P30	AAGTTAATTAAGCGTCTTGCCCCAATGCCACG
P31	GCTGGGCGCGCCTTAATCAGAT
P32	GGCGACTAGTCTGGACTTCTCGGCTATGCAA
P33	CGGGACTAGTGTCTACCACCGAACTTGCGAC
P34	CGCCAAGCTTTTTGTGTCTGTTGTATCCGGTTC
P35	GCCGGGTACCCGCCACCTGCTTTTTGTTCTA

Construction of TLN1 expression vector

The TLN1 coding sequence was cloned into pYFPYFP (67) in which the CAT selectable marker was replaced with the selectable marker HPT (pYFPYFP-HPT). 2kB of the TLN1 5’ UTR were amplified from RHΔhpt genomic DNA (P5 and P6) and cloned into pYFPYFP-HPT (pTLN1-YFPYFP). To create a cDNA driven by its native promoter in the native context, 917 nucleotides of the 5’ UTR in addition to 124 nucleotides of the first exon were amplified from genomic DNA (TLN1(nt −917-115))(Primers P7 and P8). The full length TLN1 coding sequence (nt 1 - 4938) was amplified from cDNA (P9 and P10) and TLN1(nt −917-115) TLN1(nt 1-4938) and pTLN1-YFPYFP were cloned by triple ligation to create pTLN1. The Strep II - Flag tandem affinity tag (68) was amplified with an additional 6xHIS tag and cloned in frame with TLN1 (P12 and P13) to generate TLN1-SFH. Sequencing was used to confirm all junctions.

Construction of mutated plasmids

QuikChange™ (Clontech) mutagenesis was utilized to generate all mutated plasmids. The FLAG epitope (DYKDDDDK) (69) was inserted between residues 35 and 36 (P14 and P15) to generate FLAG-TLN1. The catalytically inactive TLN1 expression vector (TLN1-E129A/H130D) was generated by changing two of the conserved zinc coordinating residues to those known to disrupt the catalytic function of other M16 proteases (70) (P16 and P17).

The FLAG-TLN1-D55A and TLN1-S1303A/L1305A, and Q1306A mutations were generated with mutagenesis primers as follows. D55A (P18 and P19), S1303A/L1305A (P20 and P21), Q1306A (P22 and P23). FLAG-TLN1-D55A and TLN1-Q1306A were generated in the context of wt TLN1. TLN1-S1303A/L1305A was generated in the context of TLN1-SFH. QuikChange™ reactions were conducted according to manufacturer’s guidelines and mutations were confirmed by sequencing.

Construction of TLN1-mCherry fusions

Constructs fusing the signal peptide, N-terminal pro-domain, and C-terminal pro-domain (TLN1_1-59-mCherry-C_1287-1645), signal peptide and N-terminal pro-domain (TLN1_1-59-mCherry), and signal peptide alone (TLN1_1-35-mCherry) to mCherry (71) were generated to assess trafficking properties of TLN1 pro-domains. The signal peptide alone (aa 1-35, P24 and P25) or signal peptide plus N-terminal pro-domain (aa 1-59 P24 and P26) were fused in frame with mCherry (P27 and P29) or mCherry fused to TLN1-C (aa 1287-1645) (P27 & P28, P30 & P31).

Western blot and Immunofluorescence Assays

Parasite lysates were separated by SDS-PAGE, transferred to nitrocellulose, and probed with primary antibodies (see above). For all secondary antibody incubations, horseradish peroxidase (HRP)-conjugated goat anti-mouse, goat anti-rat or donkey anti-rabbit antibodies (Chemicon Laboratories) were used at a 1:2000 dilution and detected by ECL (Amersham Biosciences).

Intracellular immunofluorescence assays (IFAs) of Toxoplasma infected coverslips were performed as previously described (62). For detection of TLN1 in the newly formed vacuole, the following modifications were employed: the coverslips were fixed in 4% paraformaldehyde (PFA) (Alfa Aesir), then permeabilized and probed in PBS/3% BSA/0.05% saponin.

Generation of Δtln1 parasites

The TLN1 knockout plasmid was generated by cloning 3.5 kB and 4.5 kB of 5’ (P32 and P33) and 3’ (P34 and P35) flanking sequence, respectively, around the HPT minigene in pminiGFP (72). TLN1 knockouts were generated in the RHΔhpt strain (73) and underwent a second round of homologous recombination and selection to remove the TgDHFR-HPT minigene as previously described (11). These parasites are referred to as Δtln1. Complemented parasites were generated by transfection with 25 μg to 75 μg of linearized expression constructs.

Immunoaffinity purification of TLN1 from T. gondii lysates

For immunoaffinity purification of TLN1, polyclonal anti-TLN1 antiserum was dimethylpimelimidate cross-linked to protein G sepharose beads (74). Mature TLN1 was purified from extracellular tachyzoites lysed in modified RIPA buffer plus protease inhibitors (75, 76). Eluted products were separated by SDS-PAGE and the mature protein was excised from the gel and sequenced by tryptic digest followed by MS/MS as previously described (24).

TLN1-C-SFH was purified from parasite lysates using NiNTA chromatography. Lysates in high SDS-RIPA (1% NP40, 1% DOC, 0.15M NaCl, 50 mM Tris•Cl pH 7.5, 2% SDS, 2.5 mM DTT) were boiled for 5 min at 100°C, cooled to room temperature and diluted to 0.5% SDS RIPA buffer (1% NP40, 1% DOC, 0.15M NaCl, 50 mM Tris-HCl pH 7.5, 0.5% SDS, 2.5 mM DTT). Lysates were spun at 10,000 xg to remove insoluble materal before incubating with NiNTA beads for 2 hours at room temperature. The beads were washed in 0.5% RIPA + 25mM imidazole and protein eluted in 0.5% SDS RIPA + 500 mM imidazole. Eluted products were separated by SDS-PAGE, the target band excised and subject to nano-liquid chromatography with tandem mass spectrometry as described (77).

Triton X-114 phase partitioning

To assess whether TLN1 partitions with membrane or soluble fractions, 4 × 10⁷ parasites were subjected to Triton X-114 phase partitioning as described (34).

Supplementary Material

Supp Figure S1-S4

Fig S1. Disruption of TLN1. We replaced tln1 with the selectable marker hxgprt and employed a second round of homologous recombination using negative selection to remove hxgprt and produce a strain that merely lacks the TLN1 coding region which we designated Δtnl1. (A) IFA of Δtln1 parasites stained using TLN1 and ROP13 antiserum. Scale bar = 5μm. (B) Western blot of Δtln1 parasite lysates with TLN1 antiserum using ROP1 as a loading control. Absence of TLN1 signal by IFA and Western blot confirms ablation of TLN1. (C) Parasites deficient in TLN1 have a modest growth defect. Δtln1 and WT parasites were mixed and cultured together over sequential passages through host monolayers. At each lysis a sample of the culture was used to infect host cells on coverslips. Infected coverslips were fixed and stained for TLN1 and the control protein, ROP13 and the percentage of parasites positive for TLN1 were counted. Results are representative of multiple experiments. Comparison of Δtln1 parasites to Δtln1:TLN1 complemented parasites show similar kinetics to those of the knockout versus wild-type, demonstrating that the growth defect is specific for the presence of TLN1.

Fig. S2. MS/MS of TLN1₁₈₁ covers the N-terminal 2/3 of the protein. The protein sequence of TLN1 is detailed and the 41 tryptic peptides returned from MS/MS of TLN1₁₈₁ are boxed and shaded yellow. The signal peptide is blue and underlined and zinc coordinating residues in the characteristic HXXEH₆₉EX₆E motif of M16 metalloproteases are colored purple and underlined. The repeats are green and underlined.

Fig S3. Pro-domains of rhoptry proteins do not have any obvious similarities. The three experimentally confirmed pro-domains are listed, beginning downstream of the predicted signal peptide cleavage site and ending after the ROP1-like consensus sequence. SUB2 has been omitted due to its size. Representative of ROP2-family pro-domains, which are predicted but not experimentally confirmed, are marked with an asterix (*). The putative SΦXD in ROP18 has been underlined to illustrate the potential validity of the site.

Fig S4. (A) Cartoon of tandem tagged (Strep-Strep-FLAG-HIS, SFH) TLN1 (TLN1-SFH). Green curly bracket denotes region covered by peptides recovered from MS/MS of TLN1-C-SFH (part E). (B) TLN1-SFH is processed normally. Western blot of WT and Δtln1:TLN1-SFH parasites. Lysates were probed for TLN1-N or TLN1-C. While TLN1-SFH is more highly expressed, the relative abundance of the TLN1₁₈₁ to TLN1₂₅₀ in anti-TLN1-N blots are equivalent. Cleaved TLN1-C is also present in normal amounts and the upshift of TLN1-C-SFH due to the presence of the tag is visible. (C) TLN1-C-SFH localizes to the rhoptries normally. Δtln1:TLN1-SFH parasites stained for TLN1-N and (D) FLAG show colocalization of the TLN1 signals with ROP13 in the rhotpries. Scale bar = 5μm. (E) Coomassie staining of eluted material from TLN1-C-SFH IP. The band indicated by the arrow was excised and submitted for tryptic MS/MS. Coverage of recovered peptides are diagramed by a green curly-bracket in part A.