A Human Proteome Array Approach to Identifying Key Host Proteins Targeted by Toxoplasma Kinase ROP18*

Toxoplasma kinase ROP18 is a key molecule responsible for the virulence of Toxoplasma gondii; however, the mechanisms by which ROP18 exerts parasite virulence via interaction with host proteins remain limited to a small number of identified substrates. To identify a broader array of ROP18 substrates, we successfully purified bioactive mature ROP18 and used it to probe a human proteome array. Sixty eight new putative host targets were identified. Functional annotation analysis suggested that these proteins have a variety of functions, including metabolic process, kinase activity and phosphorylation, cell growth, apoptosis and cell death, and immunity, indicating a pleiotropic role of ROP18 kinase. Among these proteins, four candidates, p53, p38, UBE2N, and Smad1, were further validated. We demonstrated that ROP18 targets p53, p38, UBE2N, and Smad1 for degradation. Importantly, we demonstrated that ROP18 phosphorylates Smad1 Ser-187 to trigger its proteasome-dependent degradation. Further functional characterization of the substrates of ROP18 may enhance understanding of the pathogenesis of Toxoplasma infection and provide new therapeutic targets. Similar strategies could be used to identify novel host targets for other microbial kinases functioning at the pathogen-host interface.

Toxoplasma kinase ROP18 is a key molecule responsible for the virulence of Toxoplasma gondii; however, the mechanisms by which ROP18 exerts parasite virulence via interaction with host proteins remain limited to a small number of identified substrates. To identify a broader array of ROP18 substrates, we successfully purified bioactive mature ROP18 and used it to probe a human proteome array. Sixty eight new putative host targets were identified. Functional annotation analysis suggested that these proteins have a variety of functions, including metabolic process, kinase activity and phosphorylation, cell growth, apoptosis and cell death, and immunity, indicating a pleiotropic role of ROP18 kinase. Among these proteins, four candidates, p53, p38, UBE2N, and Smad1, were further validated. We demonstrated that ROP18 targets p53, p38, UBE2N, and Smad1 for degradation. Importantly, we demonstrated that ROP18 phosphorylates Smad1 Ser-187 to trigger its proteasome-dependent degradation. Toxoplasma gondii is a widespread pathogen parasite that infects a broad range of warm-blooded animals, including humans, causing severe toxoplasmosis in congenitally infected newborns and immunocompromised individuals (1,2). Despite progress in the clinical management of this disease, important limitations remain, such as the lack of a commercial vaccine and imperfect drugs for toxoplasmic encephalitis (3,4). Therefore, the discovery of novel targets and effective therapeutic strategies against toxoplasmosis is of acute importance.
During infection, T. gondii delivers numerous effectors into the parasitophorous vacuole and the host cytoplasm to coopt the host cell for survival and replication (5). Among these effectors, ROP18 is suggested to be a key determinant responsible for the virulence of T. gondii (6,7). ROP18 is a member of the ROP2 subfamily of protein kinase, which is secreted from apical organelle rhoptries into the host cell and localizes to the parasitophorous vacuole membranes (PVMs) 1 during invasion (8). ROP18 is a Ser/Thr protein kinase, which mediates parasite virulence through targeting and phosphorylating host proteins (9).
In recent years, several substrates have been identified for ROP18 (10 -13). ROP18 has been reported to phosphorylate the immunity-related GTPases, Irga6 and Irgb6, preventing their accumulation at PVMs and the clearance of parasites at an early stage after infection (10,11). Given the absence of an immunity-related GTPase system in humans, it is likely that additional host defense systems may be targeted by ROP18. Using a yeast two-hybrid screening, Yamamoto et al. (12) have identified another target of ROP18, the host endoplasmic reticulum-bound transcription factor ATF6␤. ROP18 phosphorylates ATF6␤ and triggers its proteasomal degradation, and thus it impairs the ATF6␤-dependent immune responses at a later stage after infection (12). More recently, a study showed that via phosphorylation, ROP18 mediates ubiquitin-dependent degradation of p65, leading to suppression of the NF-B pathway to limit expression of proinflammatory cytokines (13). Although Cheng et al. (14) have identified several ROP18-interacting human proteins using a yeast two-hybrid technology, it remains unclear whether these proteins are substrates of ROP18. Overall, these studies have provided important insights into the basis of ROP18-mediated parasite virulence; however, the full scope of ROP18's contribution to virulence remains poorly understood. Identification of other ROP18 substrates is needed to more completely understand the mechanisms of ROP18-mediated augmentation of the disease.
Functional proteome array technology has been established as a powerful tool for proteome-wide identification of substrates for a variety of post-translational modification enzymes, such as protein kinases, acetyltransferases, and ubiquitin E3 ligases (15)(16)(17). For example, functional human protein microarrays have been used to successfully identify substrates of Ͼ370 human kinases (18 -24). In this study, we utilized a human proteome array (i.e. HuProt), composed of Ͼ16,000 human proteins, to globally identify host cell proteins that could be phosphorylated by ROP18 kinase, and we found multiple potential host cellular substrates involved in a wide range of functions. Four of these, p53, p38, UBE2N, and Smad1, were selected for further in vitro and in vivo characterization.

MATERIALS AND METHODS
Cells and Parasites-All parasites were maintained by serial passage in human foreskin fibroblasts (HFFs) as described previously (25). Tachyzoites of the RH strain and ROP18 knockout (ROP18-KO) RH strain of T. gondii were kindly provided by Professor Vern B. Carruthers (University of Michigan School of Medicine). Tachyzoites of the T. gondii ROP18-Ty1 (overexpression of Ty1-tagged ROP18) RH strain were kindly provided by Professor Jian Du (Anhui Medical University, Anhui, China). The HFF, HeLa, and 293T cell lines were purchased from the American Type Culture Collection (ATCC, Manassas, VA) and cultured in Dulbecco's modified Eagle's medium (DMEM, high glucose, and pyruvate from Gibco) containing 100 units/ml penicillin and 100 g/ml streptomycin (Gibco) and 10% fetal bovine serum (Gibco).
Dot Blot Kinase Assay-A dot blot kinase assay was performed as described previously with slight modifications (30). Briefly, 10 l of GST-ROP18 ⌬83 or GST-ROP18 ⌬83 -KD protein solution was added to a 96-well plate with 10 l of kinase assay buffer containing 2 Ci of [␥-32 P]ATP and incubated for 30 min at 30°C. Then 5 l of mixture was dotted onto the nitrocellulose membrane. After air-drying, the membrane was washed three times with TBST buffer (100 mM Tris-HCl, pH 7.5, 0.9% NaCl, and 0.1% Tween 20). After air-drying again, the membrane was exposed to the X-ray film.
Protein Array Phosphorylation Assays-The HuProt array was fabricated by spotting down 16,368 unique full-length human recombinant proteins in duplicate along with several control proteins provided by CDI Laboratories, Inc. (31,32). The protocol for phosphorylation assays on HuProt arrays was modified from that of previous publications (23,24). Briefly, HuProt arrays were preincubated in Superblock (Pierce) containing 0.2% BSA and 1% Triton X-100 for 1 h at room temperature. Purified GST-ROP18 ⌬83 and GST-ROP18 ⌬83 -KD recombinant proteins were diluted with kinase buffer (50 mM Tris-HCl, pH 7.4, 20 mM HEPES, pH 7.5, 100 mM NaCl, 10 mM MgCl 2 , 1 mM DTT, 1 mM EGTA, and 20 mM cold ATP) and [␥-32 P]ATP (33.3 nM final concentration). Duplicate HuProt arrays were incubated with the kinase mixture in a humidity chamber for 30 min at 30°C. Slides were subjected to three 10-min washes with TBST buffer (25 mM Tris-HCl, pH 7.4, 3 mM KCl, 150 mM NaCl, 0.05% Tween 20) and three 10-min washes with 10 mM Tris-HCl, pH 7.4, and 0.5% SDS, rinsed briefly with doubly distilled water, and spun dry. Control slides were incubated with kinase buffer without kinase and processed in parallel. The slides were exposed to X-ray films that were scanned at 1,800 pixels/ inch and analyzed using GenePix 3.0 (Molecular Devices). The experiments were performed in duplicate, and only those that were reproducible were scored as positives.
Protein Microarray Data Analysis-Following chip scanning, background correction was performed to obtain the signal intensity value for each spot using the ratio of foreground to background signals. Then, between-array normalization was performed. To identify positive hits on the microarrays, we estimated the standard deviation from the signal intensity distribution, as described in detail previously (33). Cutoff values of 2.0 and 2.5 standard deviations above the mean were chosen in this study. For proteins printed in duplicate, the ones with both spots meeting our defined cutoff were defined as positive hits. Finally, because it is impossible to distinguish signals arising from auto-phosphorylation events and trans-phosphorylation events, we conducted negative control experiments during each set of assays to identify proteins that underwent auto-phosphorylation on the protein microarrays. The proteins that exhibited auto-phosphorylation activity in the negative control experiments were excluded from the substrate list for further analysis. These proteins were obtained using cutoff values of 2.0 standard deviations above the mean. Background correction and between-array normalization were conducted using limma package (34).
Gene Ontology and Functional Protein Analysis-We calculated the occurrence frequency of each GO term or KEGG pathway within foreground (e.g. ROP18 substrates) and background gene groups (e.g. the whole human genome), respectively. Then we used a hypergeometric model to evaluate the statistical significance of comparison between foreground and background, followed by multipletest Bonferroni correction (for GO) or false discovery rate correction (for KEGG pathway). A GO term or KEGG pathway was defined as an enriched one if its frequency in the foreground was larger than that in the background and its corrected p value was less than 0.05.
Because many detailed GO terms were found enriched for ROP18 substrates, we merged them into five general groups as follows: metabolic process, kinase activity/phosphorylation, response to growth factor, apoptotic process/cell death, and immune system process. We merged associated GO terms into these five categories and then re-calculated the occurrence frequencies for the foreground and background gene groups, respectively, and we obtained corresponding p values based on hypergeometric distribution.
Similarity Comparison with ROP18 -We used the database, Phos-phoNetworks, to investigate phosphorylation substrates of 289 kinases. The hypergeometric probability of finding a certain number of common targets between kinase i and ROP18 is given in Equation 1, where S 0 is the total number of unique substrates in the database; K i is the number of substrates of kinase i; n is the number of ROP18 substrates detected in the database; and m is the number of common targets between ROP18 and kinase i. In Vitro Infection-Cells were infected by tachyzoites of RH strain, as described previously (35) with slight modifications. Briefly, cells were plated in 12-well plates 24 h before infection. Fresh tachyzoites of the RH strain were added to plates at a density of 5.0 ϫ 10 5 per well. After 5 h of infection, uninvaded parasites were washed with pre-warmed PBS and refilled with fresh medium. At 36 h post-infection, total cell lysates were harvested for further experiments.
Transfection-Transfections were carried out using LipoFiter TM reagent (HANBio, Shanghai, China) as instructed by the manufacturer. 293T cells (2 ϫ 10 5 per well) were seeded into 12-well plates coated with poly-L-lysine (Sangon Biotech, Shanghai, China). 24 h later, cells were incubated with plasmids mixed with LipoFiter TM reagent in DMEM basic medium for 6 h. Then the DMEM basic medium was replaced with complete DMEM. Cells were harvested at the indicated time points and used for further experiments.
Immunofluorescence Staining-Immunofluorescence staining was performed as described previously with slight modifications (25,28). Briefly, HeLa monolayers grown on coverslips were transfected with V5-ROP18 ⌬83 alone or together with FLAG-p53. At 24 h post-transfection, cells were infected with tachyzoites of T. gondii. 24 h after infection, cells were fixed with 4% paraformaldehyde for 10 min and permeabilized with 0.02% digitonin in PBS for 10 min (26). After blocking with 3% BSA in PBS for 30 min, cells were incubated with primary anti-V5 and anti-FLAG (or anti-p38 MAPK) antibodies at 4°C overnight followed by incubation with Alexa Fluorா488 goat antimouse IgG and Alexa Fluorா555 goat anti-rabbit IgG. Coverslips were mounted with ProLongா Gold antifade reagent containing DAPI (Life Technologies, Inc.) that was used for DNA visualization. Slides were viewed under an inverted fluorescence microscope (OLYMPUS 1X71) with appropriate filters.
GST Pulldown Assay-GST pulldown was performed as described previously (36). GST-ROP18 ⌬83 and GST tag proteins were purified and conjugated to glutathione-Sepharose beads. 293T cells were harvested and lysed in PBS buffer containing 1.0% Nonidet P-40, 0.5% deoxycholic acid sodium, and protease inhibitor mixtures. After incubation on ice for 25 min, the lysate was centrifuged at 13,000 ϫ g for 25 min at 4°C, and the supernatant was collected. The supernatant was buffer exchanged with PBS, pH 7.4, containing protease inhibitor mixtures using the Amiconா Ultra 10K device (Millipore) and then incubated with GST-ROP18 ⌬83 -conjugated or GST-conjugated beads for 4 h at 4°C. After washing three times with PBS containing 0.1% Triton X-100, beads were eluted in elution buffer (100 mM Tris-HCl, pH 8.0, 100 mM NaCl, 10 mM MgCl 2 , 20 mM glutathione, and 20% glycerol). Eluates were boiled for 5 min with Laemmli buffer and analyzed by Western blotting.
Coimmunoprecipitation-Immunoprecipitation assay was performed using the coimmunoprecipitation kit (Thermo Scientific Pierce) according to the manufacturer's instructions. Briefly, cells were lysed in IP lysis buffer (50 mM Tris-HCl, pH 7.4, 150 mM NaCl, 1 mM EDTA, 1% Nonidet P-40, and 5% glycerol) on ice for 25 min and centrifuged at 13,000 ϫ g for 25 min at 4°C. Corresponding primary antibodies were covalently coupled onto the AminoLink Plus Coupling Resin. Cell lysates were incubated with the coupled resin for 4 h at room temperature. The resin was washed once with IP lysis buffer and two times with conditioning buffer and eluted with elution buffer provided in the kit. Eluates were analyzed by Western blotting with the indicated antibodies.
Western Blotting-Cells were harvested and lysed in lysis buffer (20 mM Tris-HCl, pH 7.5, 150 mM NaCl, 1% Triton X-100, 2.5 mM sodium pyrophosphate, 1 mM EDTA, and 1% Na 3 VO 4 ) supplemented with phosphatase inhibitor mixture on ice for 20 min. Total cell lysates were resolved by SDS-PAGE and transferred to PVDF membranes (Millipore). The membranes were blocked for 1 h in 5% non-fat dry milk in TBST buffer (100 mM Tris-HCl, pH 7.5, 0.9% NaCl, and 0.1% Tween 20) and incubated with primary antibodies diluted in 3% BSA at 4°C overnight. Then the membranes were incubated with HRPconjugated secondary antibodies and detected with ECL Plus (Millipore).
In Vitro Phosphorylation Assay-The His-tagged full-length Smad1 was expressed in Escherichia coli BL21 (DE3) strain and purified by using Ni 2ϩ -Sepharose beads (Qiagen) as described previously (37). The GST-tagged ROP18 ⌬83 and ROP18 ⌬83 -KD recombinant proteins were purified as described above. For the in vitro phosphorylation assay, GST-ROP18 ⌬83 or GST-ROP18 ⌬83 -KD recombinant proteins were incubated with purified recombinant His-Smad1 in kinase buffer (50 mM Tris-HCl, pH 7.4, 20 mM HEPES, pH 7.5, 100 mM NaCl, 10 mM MgCl 2 , 1 mM DTT, 1 mM EGTA, and 20 M ATP) with 0.5 Ci of [␥-32 P]ATP. The reaction mixtures were incubated at 30°C for 30 min and terminated by adding SDS-PAGE sample buffer and heating immediately at 95°C for 5 min. Proteins were analyzed by SDS-PAGE, and the gel was exposed to X-ray films after dried.

Expression and Purification of ROP18 Recombinant
Proteins-To ensure purification of sufficient amounts of active ROP18 kinase proteins to carry out phosphorylation reactions on the HuProt arrays, we employed the budding yeast as an expression and purification system, as it has been successfully used to purify Ͼ400 active human and viral kinases in our previous studies (18,22). The GST-tagged ROP18 ⌬83 and the kinase-dead ROP18 mutant (D394A) (ROP18 ⌬83 -KD), a negative control, were expressed in yeast cells and purified (Fig.  1A). As judged by Coomassie staining, both GST-tagged ROP18 ⌬83 and ROP18 ⌬83 -KD proteins showed a single band at the expected molecular weights, indicating that both proteins were successfully purified with high purity (Fig. 1B). To examine whether the purified ROP18 ⌬83 retained its kinase activity, we employed auto-phosphorylation reactions using [␥-32 P]ATP as a labeling reagent. Commercially available EGFR kinase proteins were used as a positive control. Both ROP18 ⌬83 and EGFR kinases showed auto-phosphorylation activity, whereas ROP18 ⌬83 -KD did not, as demonstrated by the dot blot kinase assay (Fig. 1C). Together, these results confirmed that the purified ROP18 kinase was active and validated the kinase-dead mutant as an inactive negative control.
Identification of ROP18 Substrates in Vitro-To better understand the contribution of ROP18 to parasite virulence via its kinase activity, we set out to systematically identify the FIG. 1. Purification of active ROP18 kinase proteins. A, schematic illustration of the GST-tagged ROP18 constructs GST-ROP18 ⌬83 and GST-ROP18 ⌬83 -KD. ROP18 ⌬83 lacks the signal peptide and the prodomain, and the kinase-dead mutant ROP18 ⌬83 -KD is a D394A-mutated version of ROP18 ⌬83 . B, GST-ROP18 ⌬83 and GST-ROP18 ⌬83 -KD proteins were expressed in yeast and purified using glutathione beads. The purified proteins were separated by SDS-PAGE and stained with Coomassie Brilliant Blue. C, purified GST-ROP18 ⌬83 , GST-ROP18 ⌬83 -KD, and EGFR (positive control) were incubated with kinase assay buffer containing [␥-32 P]ATP. Proteins were then dotted on nitrocellulose membrane, followed by autoradiography.
ROP18 substrates using the HuProt arrays (version II) that contain 16,368 unique, full-length human proteins that were individually purified (Fig. 2, A and B) (31,32). Purified GSTtagged ROP18 ⌬83 and ROP18 ⌬83 -KD protein were separately incubated with a pre-blocked HuProt array in a generic kinase reaction buffer containing [␥-32 P]ATP. After the kinase reactions, the HuProt arrays were washed under denaturing conditions to ensure removal of the added kinases and unincorporated [␥-32 P]ATP. The phosphorylation signals on the HuProt arrays were detected by exposure to X-ray film (Fig.  2C). Phosphorylated proteins on each array were identified using an algorithm that measures the relative signal intensity of each spot at cutoff values of 2.0 and 2.5 standard deviations (S.D.) above the background signals (18). Proteins that showed phosphorylation signals on the array incubated with ROP18 ⌬83 -KD were omitted from the ROP18 ⌬83 hit list, because the signals were likely produced by auto-phosphorylation reactions. This approach has resulted in the identification of 172 and 68 potential substrates of ROP18 kinase at the cutoff values of 2.0 and 2.5 S.D., respectively (Table I and  supplemental Tables S1 and S2).
Bioinformatics Analyses of ROP18-Substrate Relationships-We were well aware that the 68 substrates identified using the above in vitro kinase reactions might not reflect in vivo kinasesubstrate relationships. To improve the likelihood of revealing ROP18's in vivo function, we decided to take advantage of the previous proteomics studies by integrating those existing large datasets with ours.
Because a common means for a pathogen to hijack the cellular pathways is by targeting multiple components of the same host signaling pathways (20), we first performed the STRING FIG. 2. Identification of ROP18 kinase substrates using HuProt arrays. A, overall scheme to identify the ROP18 kinase substrates. Purified GST-ROP18 ⌬83 and GST-ROP18 ⌬83 -KD proteins were assayed on HuProt arrays in kinase buffer containing [␥-32 P]ATP. After the kinase reaction, protein microarrays were washed to remove added proteins and unincorporated [␥-32 P]ATP followed by exposure to X-ray films. The films were then scanned, and bioinformatics analysis and partial validation were performed. B, representative image of the HuProt array. Individually purified human proteins tagged with GST were spotted in duplicate on a single slide, and the immobilized human proteins were visualized by probing with anti-GST antibody. C, representative photographs of X-ray films. After the kinase reactions on the HuProt arrays, the phosphorylation signals were detected by exposure of the slides to a piece of X-ray film. Developed X-ray film was then scanned, and the enlarged images of p53, MAPK11 (i.e. p38), UBE2N, and Smad1 protein dots are shown.  C4orf19  MAPK9  RFX4  CLK2  MOSPD3  RNF25  CNBP  NADK  SFRS12  CSNK1G3  NAP1L1  SMAD1  CSRP3  NME2  SMYD2  CYB5B  NPM2  SRPX2  DAPK2  OSBPL6  SSPN  DIXDC1  PACSIN2  SUPT6H  FAM5B  PARN  TBC1D16  FLT1  PAX9  THUMPD3  FN3K  PCDHGC3  TP53  GSK3B  PDPK1  TSSC4  HARS  PELI3  UBE2N  HDGF2  PLEK2  WDR45  HNRPK  PRAC  XRCC4  IRF1  PRKAR1A  ZBTB22  KIF3C  PRKCA  ZMYM3  KLF11  PRKRA  ZNF16  KPNA2  PSMD7  ZNF641  MAGEB4  RAB43 analysis (STRING.org), which integrates known protein-protein interaction (PPI) information, against the 68 in vitro substrates of ROP18, to determine whether there is any significant PPI network among these substrates (Fig. 3A) (38). As expected, 32 of the 68 substrates of ROP18 formed a tightly connected PPI network with 54 edges (i.e. known PPI relationships). Interestingly, p53 (i.e. TP53) is identified as the central hub located in the middle of the PPI network with 20 edges. Overall, 31 of the 68 substrates were also identified as either the kinases (Fig. 3A, circled in red) or kinase substrates , are significantly enriched in some biological functions, e.g. metabolic process, kinase activity/phosphorylation, response to growth factor, apoptotic process/cell death, and immune system process. C, ROP18 targets are over-represented in some signaling pathways. Examples of the insulin signaling and MAPK signaling pathways are shown. Identified ROP18 substrates are indicated by circles with GO terms annotated in different colors using the same code as A. D, using the M3 algorithm, ROP18 was found to specifically recognize an acidic phospho-motif, which is most similar to CSNK2A1. E, comparison of identified substrates using the protein microarray approach revealed 14 human kinases that share a significant portion of targets as ROP18. p values of the statistical significance of each human kinase are shown in the bottom panel. (Fig. 3A, with asterisk) in our previous study, in which an activity-based kinase-substrate relationship (KSR) network was established (18,39). Interestingly, 8 of the 31 ROP18 substrates were protein kinase themselves, whereas only 105 of the total 1,967 substrates also phosphorylated others in the KSR networks. This phenomenon is more pronounced in the PPI network formed by the 32 ROP18 substrates, 11 are known protein kinases and 21 are found in the KSR networks. Note that the majority of the KSR was determined by performing phosphorylation reactions on an early version of the human protein microarrays, each composed of ϳ4,200 unique proteins (18,39). Taken together, these analyses indicate a tendency of ROP18 to act as a master kinase that regulates many host kinases and signaling components (p ϭ 1.2E-4).
Second, we performed Gene Ontology (GO) analysis to determine possible roles of ROP18 in pathogen-host interactions. We found that the 68 ROP18 substrates were significantly enriched in following biological functions: metabolic process (p ϭ 1.2E-5); kinase activity or phosphorylation (p ϭ 5.6E-6); response to growth factor (p ϭ 5.0E-5); apoptotic process or cell death (p ϭ 6.1E-4), and immune system process (p ϭ 2.9E-4) (Fig. 3B). When focusing on the 32 substrates that form a highly connected PPI network, the p values of the corresponding GO terms were further improved by at least 1.7 logs (right panel, Fig. 3B). This analysis suggests that ROP18 is likely to play a pleiotropic role as a master kinase that is capable of regulating multiple signaling pathways in parallel.
To better understand how ROP18 recognizes its substrates, we next employed a previously reported M3 algorithm, which integrates the phosphorylation sites identified with mass spectrometry approaches with ROP18's substrates, and we determined the phosphorylation motif logo of ROP18 as (S/ T)XXXXXDE (Fig. 3D) (18,39). After searching the KSR database, we found that the acidic phospho-motif recognized by ROP18 is most similar to that of the catalytic ␣ subunit of casein kinase 2 (i.e. CSNK2A1), which is well known to be involved in many cellular processes, including apoptosis. Like ROP18, CSNK2A1 was also identified as a master kinase (p ϭ 0.025), which tends to phosphorylate many other kinases.
Because a common theme utilized by pathogenic kinases is to imitate the host counterparts, we directly compared the substrate similarity of ROP18 with the 289 human kinases in the PhosphoNetworks database. We observed that 14 human kinases showed a significant similarity (p Ͻ 0.01) with ROP18 in terms of substrate specificity (Fig. 3E). Moreover, four of these 14 kinases, including NIM1, PRKX, PRKCH, and GSK3␤, have a strong tendency to phosphorylate other kinases. This is significant (p Ͻ 0.05) because only 57 of the 289 tested kinases showed a similar behavior. This observation suggests again that ROP18 is likely to mimic human master kinases rather than general kinases. Of the 14 kinases, STK3 showed the lowest p value, at least 10-fold lower than the second best one. STK3 (also known as MST2) is a proapoptotic serine/threonine kinase and plays a critical role in the Hippo-YAP signaling pathway (40). In response to proapoptotic molecules, STK3 is activated by caspase cleavage of the auto-inhibitory domain, and the cleaved N-terminal portion of STK3 goes to the nucleus and forms the active kinase dimer. A recent study by Wu et al. (41) suggested that ROP18 might manipulate the host mitochondrial apoptosis pathways. They observed that overexpression of T. gondii ROP18 significantly suppressed 293T cell apoptosis induced by actinomycin D and maintained mitochondrial membrane potential and integrity. Interestingly, STK3 recognizes a very different phosphomotif (i.e. HXRXX(S/T), where X is any amino acid) than ROP18, suggesting that ROP18 phosphorylates different Ser/ Thr residues of these shared substrates, which might lead to different physiological outcomes of the substrates. Two other kinases, RPS6KA6 and PRKX, were found to recognize a very similar acidic phospho-motif as ROP18. Although not much is known about PRKX functions in cells, RPS6KA6 (i.e. RSK4) is well characterized for its role in the neurotrophin signaling (42). RSK4 acts as a downstream kinase of ERK1/2 that activates cell differentiation and cell survival processes. Therefore, it is plausible that ROP18 hijacks RSK4's role by phosphorylating the same downstream substrates of RSK4.
ROP18-mediated p53 Degradation via Phosphorylation-On the basis of the study by Wu et al. (41) and the above bioinformatics analyses, we suspected that ROP18 might play an important role in anti-apoptosis. Intriguingly, p53 was identified as a novel substrate of ROP18 in our in vitro kinase assays and is located at the central hub position in the PPI network among the 32 ROP18 substrates (Fig. 3A). p53 is a master transcription factor and is known to respond to diverse cellular stresses to regulate expression of target genes, thereby inducing cell cycle arrest, apoptosis, senescence, DNA repair, or changes in metabolism (43). p53 is also a well known tumor suppressor involved in immunity (44). It has been reported that infection from T. gondii induces endogenous p53 degradation; however, the exact molecular mechanism remains unknown (45,46). Therefore, we decided to validate the relationship between ROP18 and p53 and determine the impact of ROP18 phosphorylation on p53.
To do so, immunofluorescence staining was performed. The results showed that p53 was colocalized with ROP18 at the PVMs in host cells (Fig. 4A), indicating that ROP18 and p53 might form a protein complex. To access whether ROP18 interacts with p53, we performed a GST pulldown assay with recombinant GST-ROP18 ⌬83 protein. The results showed that GST-ROP18 ⌬83 , but not the GST tag, was able to pull down p53 (Fig. 4B). Moreover, we performed coimmunoprecipitation assay and detected an interaction between ROP18 and p53 (Fig. 4C). We then investigated the influence of ROP18 on p53 expression. We observed that overexpression of ROP18 ⌬83 but not ROP18 ⌬83 -KD significantly decreased the protein level of p53 (Fig. 4D), indicating that the kinase activity is essential for ROP18-mediated p53 degradation. In addition, ROP18 decreased the protein levels of p53 in a dose-dependent manner (Fig. 4E). To further demonstrate the specificity of ROP18-dependent p53 degradation, we cotransfected the FLAG-tagged p53 construct with different amounts of the V5-tagged ROP18 ⌬83 constructs to 293T cells. We found that overexpression of ROP18 ⌬83 caused dose-dependent degradation of the exogenous p53 protein (Fig. 4F). To investigate the impact of ROP18 on p53 expression in vivo, we infected 293T cells with wild-type (WT) RH or ROP18-Ty1 RH strain. Consistent with the previous studies (45,46), we found that in 293T cells infection with WT RH strain led to a significant degradation of p53 protein. Additionally, infection of 293T cells with ROP18-Ty1 RH strain further extended the degradation (Fig. 4G). Taken together, these results demonstrated that ROP18 relies on its kinase activity to inhibit apoptosis in T. gondii-infected cells via targeting p53 for degradation.
ROP18-mediated p38 Degradation via Phosphoryla tion-MAPK signaling has been reported to be involved in the infection of T. gondii (47,48). Intriguingly, our bioinformatics analyses suggested that ROP18 might regulate the MAPK pathways via targeting p38 (e.g. MAPK11). The p38 MAPKs are known to be activated by cytokines, LPS, and environmental stress and to be implicated in several processes, including cell survival, apoptosis, and immune function (49).

FIG. 4. ROP18 interacts with p53 and promotes its degradation.
A, HeLa cells were transfected with V5-ROP18 ⌬83 and FLAG-p53 followed by infection with tachyzoites of RH strain for 24 h, and then immunofluorescence staining was performed with mouse monoclonal anti-V5 (green) and rabbit polyclonal anti-FLAG (red) antibodies. DAPI (blue) was used to visualize cell nuclei. Phase, bright field; bar, 5 m. B, GST pulldown assay was performed using GST-ROP18 ⌬83 protein purified from yeast and lysates of 293T cells overexpressing FLAG-p53. IB, immunoblot. C, 293T cells were transfected with ROP18 ⌬27 -GFP expression plasmid, and coimmunoprecipitation was performed with anti-GFP, followed by immunoblotting with the indicated antibodies. D, 293T cells were transfected with V5-ROP18 ⌬83 or V5-ROP18 ⌬83 -KD plasmids for 36 h, and the total cell lysates were subjected to Western blotting with the indicated antibodies. E, lysates of 293T cells transiently transfected with the indicated amounts of V5-ROP18 ⌬83 plasmids were detected by Western blotting with the indicated antibodies. F, 293T cells were cotransfected with the indicated amounts of V5-ROP18 ⌬83 and 1 g of FLAG-p53 plasmids. 36 h after transfection, total cell lysates were analyzed by Western blotting. G, lysates of 293T cells infected with wild-type (WT) RH and ROP18-Ty1 RH strains were detected by Western blotting with the indicated antibodies. SAG1 and GAPDH were used as loading controls.
The p38 MAPK pathway has been related to T. gondii invasion (50,51). To confirm the association between ROP18 and p38, immunofluorescence staining was performed. The results revealed a colocalization of p38 and ROP18 at the PVMs (Fig.  5A), indicating that ROP18 might form a complex with p38. To confirm this assumption, we performed a GST pulldown assay. The results showed that GST-ROP18 ⌬83 was able to pull down p38 (Fig. 5B), suggesting that ROP18 interacts with p38. We next investigated the potential impact of ROP18 on p38 in cells. We found that overexpression of ROP18 ⌬83 rather than ROP18 ⌬83 -KD significantly decreased the expression level of p38 protein (Fig. 5C), indicating that the kinase activity is required for ROP18-mediated p38 degradation. To further determine whether ROP18 affects the stability of p38 in vivo, 293T cells were infected with WT RH or ROP18-Ty1 RH strain. The results showed that infection with WT RH strain significantly reduced the p38 protein level as compared with that in un-infected cells, and infection with ROP18-Ty1 RH strain resulted in a more significant reduction of p38 protein level (Fig. 5D). Together, these results suggested that ROP18 targets p38 for degradation via phosphorylation.
ROP18-mediated UBE2N Degradation via Phosphorylation-Du et al. (13) reported that ROP18 inactivates the NF-B pathway via phosphorylation and degradation of p65. Interestingly, we identified another upstream regulator of NF-B pathway, UBE2N (52,53), as a putative substrate of ROP18. UBE2N is also found as a hub in the PPI network that is directly connected to p53, PDPK1, PELI3, ZFAND4, and NME2 (Fig. 3A). UBE2N (also known as Ubc13) is a unique E2 ubiquitin-conjugating enzyme and plays an important role in host inflammatory responses (54). For example, a secreted Shigella effector, OspI, was reported to specifically target Ubc13 and deamidates Gln-100 to a glutamic acid residue, leading to the disruption of TRAF6-catalyzed poly-ubiquitylation. This results in suppressing the diacylglycerol 3 CBM 3 TRAF63 NF-B signaling pathway and dampens the host inflammatory responses (55). To validate the relevance of ROP18 and UBE2N, we expressed V5-ROP18 ⌬83 or V5-ROP18 ⌬83 -KD in 293T cells. Western blotting analysis showed that overexpression of ROP18 ⌬83 but not ROP18 ⌬83 -KD led to a significant reduction in the protein level of UBE2N (Fig. 6A), suggesting that the kinase activity of ROP18 plays an important role in ROP18-mediated UBE2N degradation. We then investigated whether ROP18 affects UBE2N expression in vivo. The data showed that, although infection with the WT RH strain significantly decreased the UBE2N expression relative to non-infection, infection with the ROP18-Ty1 RH strain resulted in a more significant decrease in the protein level of UBE2N (Fig. 6B). Together, these results suggested that UBE2N is a ROP18-targeting host protein.
ROP18-mediated Smad1 Proteasomal Degradation via Phosphorylation-Interference of the TGF-␤ signaling pathway has been shown to be highly associated with Toxoplasma infection (56). Accordingly, Smad1, one of the family members that act downstream of receptors of TGF-␤ family members (57), was identified as a substrate of ROP18 kinase in this study. To determine whether ROP18 interacted with Smad1, we performed the coimmunoprecipitation assay and observed that ROP18 could readily immunoprecipitate Smad1 in cotransfected cells (Fig. 7A). We then investigated the influ-FIG. 5. ROP18 interacts with p38 and promotes its degradation. A, HeLa cells transiently expressing V5-ROP18 ⌬83 were infected with tachyzoites of RH strain. 24 h after infection, immunofluorescence staining was performed with mouse monoclonal anti-V5 (green) and rabbit monoclonal anti-p38 (red) antibodies. DAPI (blue) was used to visualize cell nuclei. Phase, bright field; bar, 5 m. B, GST pulldown assay was performed using GST-ROP18 ⌬83 protein and 293T cell lysates. IB, immunoblot. C, 293T cells were transfected with V5-ROP18 ⌬83 or V5-ROP18 ⌬83 -KD plasmids for 36 h, and then total cell lysates were subjected to Western blotting analysis. D, 293T cells were infected without or with wild-type (WT) RH or ROP18-Ty1 RH strain for 36 h. Cellular lysates were separated by SDS-PAGE and immunoblotted with the indicated antibodies. SAG1 and GAPDH were used as loading controls. ence of ROP18 on Smad1 expression. The results showed that expression of ROP18 ⌬83 but not ROP18 ⌬83 -KD markedly reduced the protein level of Smad1 (Fig. 7B) in a dose-dependent manner (Fig. 7C). Interestingly, the protein level of Smad1 was restored in the presence of a proteasome inhibitor, MG132 (Fig. 7D), suggesting that ROP18 phosphorylation-induced Smad1 degradation was mediated by the proteasome. To confirm that Smad1 could be phosphorylated by ROP18, we carried out an in vitro kinase assay with purified ROP18 ⌬83 and ROP18 ⌬83 -KD proteins. The results clearly showed that GST-ROP18 ⌬83 , but not GST-ROP18 ⌬83 -KD, strongly phosphorylated the His 6 -tagged Smad1 protein (Fig. 7E).
To further investigate the molecular basis of ROP18-mediated Smad1 degradation, we sought to determine which residue(s) in Smad1 could be phosphorylated by ROP18. In previous studies, phosphorylation in the linker region of Smad1 has been shown to trigger its degradation (58,59). The observation that ROP18 also induced Smad1 proteasomal degradation prompted us to test the hypothesis that ROP18 might utilize a similar mechanism to regulate Smad1 protein levels. Thus, we performed site-directed mutagenesis at Ser-187, Ser-195, Ser-206, and Ser-214 in the linker region, which were identified as the phosphorylation sites in previous studies (59). By testing different combinations of mutations at these Ser residues, we found that Ser-187 to Ala mutation blocked the reduction of Smad1 protein levels induced by ROP18 (Fig. 7F), suggesting that Ser-187 phosphorylation of Smad1 by ROP18 was responsible for its degradation. To verify that Smad1 Ser-187 was phosphorylated by ROP18, we examined the phosphorylation status of Smad1 Ser-187 by performing immunoblotting analysis with a commercial phospho-specific antibody against Smad1 Ser-187 in transfected cells. The results showed that, compared with the control vector, transfection of ROP18 ⌬83 significantly increased the phosphorylation level of Smad1 Ser-187, especially considering that the Smad1 protein was strongly reduced by ROP18 ⌬83 (Fig. 7G). In contrast, no significant difference was observed upon transfection of ROP18 ⌬83 -KD (Fig. 7G). To further determine whether ROP18 phosphorylates Smad1 at Ser-187 in vivo, HFF cells were infected with the WT RH, ROP18-KO RH, or ROP18-Ty1 RH strain. Immunoblotting analysis showed that as compared with non-infection, infection with the WT RH and ROP18-Ty1 RH strain resulted in increased phosphorylation levels of Smad1 Ser-187 accompanied by decreased Smad1 protein levels, whereas infection with ROP18-KO RH strain did not alter the phosphorylation status of Smad1 at Ser-187 and had no effect on Smad1 protein levels (Fig. 7H). These results suggested that ROP18 phosphorylates Smad1 Ser-187 to facilitate its degradation. DISCUSSION In this study, we performed phosphorylation reactions on a HuProt array to globally screen for host substrates of ROP18 kinase, which is a key determinant of parasite virulence. A number of novel putative host targets involved in various cellular functions were identified. Among these proteins, four candidates, p53, p38, UBE2N, and Smad1, were further validated as ROP18 substrates.
A key to the success of this study was the purification of active ROP18 kinase proteins from the budding yeast. Although the bacterial systems have been more popular for protein expression, they suffer from a lack of proper protein post-translational modifications, and overexpression of foreign proteins in bacteria often resulted in an insoluble or inactive form of the desired proteins (28,60). Yeast cells combine the advantages of being single cells, such as fast growth and easy genetic manipulation, as well as eukaryotic features, including a secretory pathway leading to correct protein processing and post-translational modifications (61). Indeed, recombinant ROP18 proteins currently purified from E. coli are only suitable for immunity or structural analysis, requiring little or no activity (60,62). A truncated ROP18 protein that corresponds to the kinase catalytic domain was successfully expressed in E. coli, but the recombinant protein had to be refolded to obtain bioactivity. The refolded recombinant protein was rather unstable and could not be obtained in sufficient amounts (28). To our knowledge, we have provided the first demonstration that by using a yeast system, the mature ROP18 protein with high kinase activity and satisfactory quantity was successfully purified. This finding paves a way for oncoming in-depth studies, such as crystal structure analysis of phosphorylated ROP18, in vitro kinase assays, and substrate screening.
Identification of key host targets of ROP18 is crucial for understanding the pathogenesis of T. gondii. So far the discovery of ROP18 substrates mainly relies on the yeast twohybrid method, in which interactions are typically detected in the nucleus, thus restricting the types of interaction that can be detected. In addition, the two-hybrid approach detects protein-protein interaction but not the protein bioactivity (29). Protein microarray is a powerful tool for comprehensive FIG . 6. ROP18 promotes degradation of UBE2N. A, 293T cells were transfected with V5-ROP18 ⌬83 or V5-ROP18 ⌬83 -KD plasmids for 36 h, and then total cell lysates were subjected to Western blotting analysis. IB, immunoblot. B, 293T cells were infected without or with wild-type (WT) RH or ROP18-Ty1 RH strain for 36 h. Cell lysates were separated by SDS-PAGE and immunoblotted with the indicated antibodies. SAG1 and GAPDH were used as loading controls.
screens of protein functions, which has been applied for global analysis of protein modified by ubiquitylation, SUMOylation, acetylation, and S-nitrosylation (17). Additionally, protein microarray has been expanded to identify kinase substrates (18 -24). Whereas creating protein microarrays requires a specialized system, the arrays can be stored and saved for later use, providing a faster and cheaper detection tool for high-throughput studies. In this study, using the human protein microarray, we globally determined the substrates of ROP18 kinase and identified a total of 68 potential host targets.
T. gondii strictly relies on host cellular metabolism to obtain the energy and nutrients for its replication and survival (63). But how T. gondii modulates the metabolism of its host to create a more hospitable metabolic niche remains vague. Transcriptomics analysis has shown that infection of T. gondii results in alteration of protein expression in key metabolic pathways, as characterized by the down-regulation of mitochondrial respiratory chain and the metabolism of fatty acids, lipids, and xenobiotics (64). Interestingly, in this study, we found that the majority of ROP18 substrates are enriched in the metabolic process, suggesting an important role of ROP18 in manipulating the host cellular metabolism. In particular, ROP18 seems to regulate the host insulin signaling, which modulates the metabolism of glucose, lipids, proteins, and energy (65,66), as several putative substrates take part in this pathway. T. gondii preferentially utilizes host glucose as a source for the synthesis of ATP, nucleic acid, proteins, and FIG. 7. ROP18 interacts with Smad1 and triggers its proteasomal degradation via phosphorylation. A, 293T cells were transfected with ROP18 ⌬27 -GFP expression plasmid, and coimmunoprecipitation was performed with anti-GFP, followed by immunoblotting with the indicated antibodies. IB, immunoblot. B, 293T cells were transfected with V5-ROP18 ⌬83 or V5-ROP18 ⌬83 -KD and/or FLAG-Smad1 plasmids for 36 h, and then total cell lysates were analyzed by Western blotting with the indicated antibodies. C, 293T cells transiently cotransfected with the indicated amounts of V5-ROP18 ⌬83 or V5-ROP18 ⌬83 -KD and FLAG-Smad1 plasmids and cell lysates were detected by Western blotting. D, 293T cells were transfected with V5-ROP18 ⌬83 and/or FLAG-Smad1 plasmids in the absence or presence of 10 M MG132 for 24 h, and lysates were detected by Western blotting. E, in vitro kinase assay was performed to detected His-Smad1 phosphorylation with purified GST-ROP18 ⌬83 and GST-RO18 ⌬83 -KD proteins. F, 293T cells were cotransfected with FLAG-Smad1 mutants (M1, S187A, S195A, S206A, and S214A; M2, S187A, S195A, and S206A; M3, S206A; M4, S187A) and V5-ROP18 ⌬83 for 48 h, and the protein levels of Smad1 mutants were detected by Western blotting. G, 293T cells were transfected with V5-ROP18 ⌬83 or V5-ROP18 ⌬83 -KD and/or FLAG-Smad1 for 48 h, and Smad1 phosphorylation was detected by Western blotting with phospho-Smad1 (Ser-187) antibody. H, HFF cells were infected with wild-type (WT) RH, ROP18-KO RH, or ROP18-Ty1 RH strain, and Smad1 phosphorylation was detected by Western blotting with phospho-Smad1 (Ser-187) antibody. SAG1 and GAPDH were used as loading controls. lipids and intracellular replication (67). Thus, ROP18-mediated insulin pathway might be beneficial for the metabolism of T. gondii.
Induction of apoptosis is a common host response to infection. Therefore, pathogens including T. gondii have developed strategies to hamper host cell apoptosis. Consistent with this notion, ROP18 has been shown to inhibit host cell apoptosis with a not yet known mechanism. Previous studies reported that T. gondii infection alters host MAPK signaling pathway (47,48), which regulates multiple cellular processes, including apoptosis (68). Intriguingly, we identified several MAPK pathway effectors, including MAP3K5, MAPK9, MAPK10, MAPK11, PRKCA, and p53, as the putative ROP18 substrates. In addition, three substrates, PDPK1, GSK3␤, and p53, are implicated in the PI3K/AKT pathway that are important for maintaining the cellular anti-apoptotic state (69). Moreover, host proteins FLT1 (70) and PRKRA (71) involved in apoptosis and cell growth were also identified as the ROP18 targets. Overall, our data suggest that ROP18 might regulate host cell apoptosis through multiple host signaling pathways and targets.
Host immunity limits the parasite growth. In turn, T. gondii interferes with the host immune-related signaling, enabling the parasite to evade the host immune response. ROP18 has been reported to modulate the NF-B signaling pathway that is critical in regulating inflammation and innate immune responses to T. gondii infection (13). In agreement with this idea, host proteins UBE2N (72,73) and RNF25 (74), which take part in NF-B pathway, were identified as the ROP18 substrates, indicating that these proteins might be required for ROP18-mediated NF-B signaling regulation. IFN-␥ secretion is crucial for controlling T. gondii infection (75). T. gondii has been reported to down-regulate the expression of IRF1 (76,77), a transcription factor that regulates most of the other half of IFN-␥-responsive genes (78). Intriguingly, we identified IRF1 as a potential substrate of ROP18, implying that T. gondii probably regulates IRF1/IFN-␥ expression via ROP18 kinase.
Phosphorylation by protein kinases is a common way to regulate the complex signaling networks. Interestingly, we identified at least eight host kinases as ROP18 substrates, including CLK2, CSNK1G3, PDPK1, PRKCA, GSK3␤, MAPK10, MAPK11, and MAPK9. All of these kinases are key regulators involved in multiple biological processes, including cell cycle, apoptosis, cell proliferation, immunity, and metabolism, to name a few (Fig. 3A). Bioinformatics analyses based on our dataset now suggest that ROP18 may serve as a master kinase modulating multiple host signaling pathways. Notably, ATF6␤ (12) and p65 (13) that have been reported to be the substrates of ROP18 were not identified in our assays. ATF6␤ was absent in the human ORF collection used to express the proteins for HuProt array fabrication, whereas p65 was available in our collection but its protein level was not detected on the HuProt, presumably due to failure in its purification (29).
Although the possibility of false positives or false negatives due to incorrect folding or an insufficient amount of some proteins on the array cannot be excluded (29), our follow-up validation of four substrates, p53, p38, UBE2N, and Smad1, confirmed the efficacy of the screening.
The central position of p53 occupied in the PPI network formed by ROP18 substrates indicates that p53 might be a key host target required for ROP18-dependent parasite virulence. Studies have reported that infection of T. gondii induces degradation of endogenous p53, although the mechanism is not clearly understood (45,46). In this study, we found that p53 is a host target of ROP18 kinase, and overexpression of ROP18 induced p53 degradation. Combined with previous studies, our results indicate that T. gondii induces p53 degradation via active ROP18 kinase. ROP18 has been reported to inhibit the apoptosis of infected cells, thereby providing a stable environment for parasitic proliferation (41). Considering the important role of p53 in apoptosis, it is possible that ROP18 mediates p53 degradation and thus suppresses host cell apoptosis. In addition to its function in apoptosis, p53 also plays a critical role in innate immunity (43,44). It has been reported that down-regulation of p53 results in impaired expression of genes involved in interferon or toll-like receptor signaling during viral infection (79,80). Gadd45␣ is a downstream target of p53 (43). The deficiency of Gadd45␣ in mice results in impaired Th1 development due to the lack of p38 activation, IL-12, and CD40 production in dendritic cells and diminished Th1 immune response to T. gondii antigen (81). Therefore, it is also likely that during pathogen infection, ROP18 binds to and degrades p53, leading to down-regulation of Gadd45␣ expression, which in turn results in reduction of IL-12 and CD40 production, thereby conferring an immune-privileged microenvironment for parasite proliferation, but such a mechanism remains to be experimentally demonstrated.
The p38 MAPK pathways controlling gene expression and early immune response contribute to the infection and replication of T. gondii (47,48). It has been reported that parasite invasion triggers p38 activation and drives IL-12 production through an association with TAB1 (51). Braun et al. (82) also reported that during Toxoplasma infection the dense granule protein GRA24 binds to p38 and promotes p38 auto-phosphorylation and nuclear translocation, which results in upregulation of the transcription factors Egr-1 and c-Fos and the synthesis of cytokines IL-12 and MCP1. Parasite infection activates the p38 MAPK pathway; however, this activation is rapidly decreased thereafter (50), implying a feedback mechanism subverting the host p38 MAPK pathway may exist. Interestingly, in this study, we identified p38 as a host target of ROP18. Moreover, we found that ROP18 interacted with p38 and triggered its degradation. Our results indicate that ROP18 might be a key negative modulator of the p38 MAPK pathway that disrupts the autonomous host cell protecting early immune response.
Ubiquitylation is an important process modulating protein destruction. It has been reported that ROP18 induced ubiquitin-dependent degradation of host protein (13) indicates that ROP18 is involved in ubiquitylation-related signaling. Here, we demonstrated that UBE2N, a member of the E2 ubiquitinconjugating enzyme family (52), is a host target of ROP18. Previous studies have shown that UBE2N is essential for the immune signaling. In thymocytes, UBE2N regulates NF-B, MAPK, and TAK1 (TGF-␤-activated kinase 1) signaling pathways and thus regulates T cell receptor-mediated signaling and innate immunity (52,53,73,83). Conjunction of TRAF6 with UBE2N is necessary for the activation of TNF receptorassociated factor (TRAF)-dependent signal transduction pathways (NF-B, JNK, and p38 MAPK) and inflammatory responses (73,84). Given the critical role of UBE2N in innate and inflammatory responses, disruption of UBE2N by ROP18 may result in impaired immune responses, thus facilitating the intracellular growth of T. gondii. Further studies are required to validate this hypothesis.
TGF-␤ signaling, which functions in cell proliferation, apoptosis, embryonic development, angiogenesis, and immunity, plays important roles in controlling Toxoplasma infection (56,85). TGF-␤1 is an important cytokine that has a dual role in the regulation of immune response against T. gondii infection as follows: induction of immune responses by development of Th17 lymphocytes and mucosal immunity and suppression of immune responses by inducing T regulatory cells, which restrict Th17 development, and inhibiting pro-inflammatory cytokine production, such as TNF-␣ (85). Smad1 is one of the members of the intracellular effectors of TGF-␤ signaling, which has important roles in cell proliferation, inflammation, and T cell differentiation (86 -88). Smad1 is a mediator of bone morphogenetic protein signals, but it can also be activated by TGF-␤1 (89). In this study, we identified for the first time that Smad1 is a target of ROP18, but its function and mechanisms underlying the parasite virulence remain to be defined. It has been reported that manipulation of host cell targets by T. gondii contributes to its proliferation (90). Considering the important role of Smad1 in cell proliferation, we determined whether targeting Smad1 affects T. gondii intercellular growth; however, the results showed that silencing Smad1 in HFF cells was not obviously beneficial for T. gondii replication (data no shown), implying that Smad1 might mediate parasite virulence via a mechanism that is distinct from regulating the parasite proliferation. Studies have been reported that Ids, the downstream targets of Smad1 (91), are pivotal regulators of lymphocyte differentiation and proliferation, particularly the CD8 ϩ effector and memory T cells and CD4 ϩ T cells (92)(93)(94)(95). Our data showed that ROP18 interacted with Smad1 and prompted its proteasomal degradation. Therefore, it is possible that after parasite infection, ROP18 impairs the T cell development and function by down-regulating Smad1 and its downstream effectors Ids, and thus it facilitates an escape of T. gondii from the host immune sys-tem, but this hypothesis remains to be verified in the parasiteinfected Smad1 mutant mouse model.
In summary, we report for the first time that with the use of a human proteome array, the substrates of T. gondii ROP18 kinase were systemically identified. Moreover, four important proteins, p53, p38, UBE2N, and Smad1 were validated as the host targets. Notably, we demonstrated that ROP18 phosphorylates Smad1 Ser-187 to promote its proteasome-dependent degradation. Functional analysis of the novel substrates of ROP18 kinase will be of benefit for understanding Toxoplasma virulence regulatory mechanisms mediated by ROP18 and providing potential targets for therapeutic intervention. The approaches utilized in this study may offer a new strategy for discovering novel host targets during host-pathogen interactions.