Abstract
Cytoadherence and consequential migration are crucial for pathogens to establish colonization in the host. In contrast to the nonadherent isolate of Trichomonas vaginalis, the adherent one expresses more actin-related machinery proteins with more active flagellate-amoeboid morphogenesis, amoeba migration, and cytoadherence, activities that were abrogated by an actin assembly blocker. By immunoprecipitation coupled with label-free quantitative proteomics, an F-actin capping protein (TvFACPα) was identified from the actin-centric interactome, with an atypically greater binding preference to G-actin than F-actin. TvFACPα partially colocalized with F-actin at the parasite pseudopodia protrusion and formed the protein complexes with α-actin through its c-terminal domain. Meanwhile, TvFACPα overexpression suppresses F-actin polymerization, amoeboid morphogenesis, and cytoadherence in this parasite. Ser2 phosphorylation of TvFACPα enriched in the amoeboid stage of adhered trophozoites was reduced by a CKII inhibitor. The site-directed mutagenesis and CKII inhibitor treatment revealed that Ser2 phosphorylation acts as a switching signal to alter TvFACPα actin-binding activity and consequent actin cytoskeleton behaviors. Through CKII signaling, TvFACPα also controls the conversion of adherent trophozoite from amoeboid migration to flagellate form with axonemal motility. Together, CKII-dependent Ser2 phosphorylation regulates TvFACPα binding actin to fine-tune cytoskeleton dynamics and drive crucial behaviors underlying host colonization of T. vaginalis.
Introduction
Trichomonas vaginalis is a pathogenic protist causing trichomoniasis which is one of the most prevalent non-viral sexually transmitted diseases, with approximately 180 million new infections worldwide annually (1).
A successful pathogenic infection includes cytoadherence to establish colonization, followed by migration for population spread. Numerous studies on Trichomonas vaginalis have focused on the cytoadherence mechanism in adhesion molecules, like cadherin (2), rhomboid protease (3), legumain protease (4), BAP proteins (5), TvAD1 protein (6), and surface-expressed hydrogenosomal proteins (7, 8, 9, 10). However, the effects of these reputed adhesins in cytoadherence are limited when analyzed by the gain- or loss-of-function assays (2–10). Thus, we postulated that the cytoadherence of T. vaginalis might be regulated by pathways other than adhesion molecules. In mammalian adhesion cells, transmembrane integrins link peripheral focal protein complexes underneath the cell membrane for focal adhesion, which is the site that connects the extracellular matrix to transmit traction forces required for cell migration and activates downstream signaling followed by local cytoskeleton reorganization (11, 12, 13). A few studies have used ligand competition or antibody neutralization to demonstrate the involvement of integrin-like molecules in the cytoadherence of T. vaginalis (14, 15, 16). Recently, the adherence of clinical T. vaginalis isolates to the plastic surface or host cells was shown to be influenced by an actin polymerization blocker (17), implying that the actin cytoskeleton might coordinate cytoadherence in T. vaginalis, but the regulatory mechanism was unknown.
Furthermore, flagellate-amoeboid transition immediately after contact with a solid surface or human vagina epithelium cells (hVECs), is another striking feature in adherent isolates of T. vaginalis (18). Upon morphological transformation, the free-swimming flagellar trophozoite converts to an adherent trophozoite that crawls over a solid surface by pseudopodia-like protrusions referred to as amoeboid migration. A similar flagellate-amoeboid transition was observed in the pathogenic amoeba, Naegleria fowleri. This free-living trophozoite builds lamellipodia-like protrusions for phagocytosis and migration driven by actin cytoskeleton machines (19), in which actin expression correlates with its virulence (20).
The actin cytoskeleton is a complex network of actin filaments and actin-associated proteins that shape cell morphology, drive cellular locomotion, and confer cell adhesion (21, 22, 23). The globular actin monomer (G-actin) polymerizes into filamentous actin polymers (F-actin), which are further organized into bundles or branched into three-dimensional networks for complicated cytoskeleton activities. In the polarized F-actin filament, growth initiates from the assembly of the Arp2/3 nucleation complex (24), then G-actin is continuously added at the fast-growing barbed end or dissociated from the pointed end (25). The cellular actin cytoskeleton dynamics are tightly modulated by a variety of accessory effectors for actin polymerization, depolymerization, branching, and reorganization (26). In high eukaryotes, F-actin capping protein (CP) is heterodimerized from α (CPα) and β (CPβ) subunits to form a mushroom-shaped structure capping the fast-growing barbed end of F-actin to block off G-actin access and subsequent polymerization. The C-terminal regions of CPα and CPβ form as two tentacles to bind actin (27, 28, 29). A set of regulatory proteins binds to the barbed end of F-actin to prevent the binding of CP, or several proteins directly bind CP to spatially guide subcellular localization or allosterically alter actin capping activity for instant regulation of cytoskeleton remodeling (30, 31).
Post-translational modifications like phosphorylation and acetylation on the interacting interface within the c-terminal tentacle of CPβ alter the actin-binding dynamics (32). Human CPα forms a protein complex with Casein kinase II-interacting protein (CKIP-1) and Casein kinase II (CKII). CKII phosphorylates the Ser9 of CPα coordinating CKIP-1 to inhibit capping activity, but this inhibitory effect seems to be independent of Ser9 phosphorylation (33, 34). The capacity of CP binding actin filaments is tightly regulated in a spatial or allosteric manner to fine-tune the actin assembly dynamics in cells.
The mechanisms of actin cytoskeleton regulation in T. vaginalis have not been fully elucidated. TvFimbrin1 protein (TvFim1) has been identified in vitro to accelerate actin assembly and in vivo to co-localize with F-actin at the cell membrane periphery in the pseudopod-like structures of T. vaginalis upon phagocytosis or migration (35). In this study, a putative F-actin capping protein subunit α (TvFACPα) was identified and characterized from the α-associated protein complexes in T. vaginalis.
Results
Differential morphogenesis, cytoadherence, and motility of T. vaginalis
The differential host-parasite interaction between nonadherent T1 and adherent TH17 isolates was evaluated by the cytoadherence, morphogenesis, and motility. CFSE-labeled trophozoites were co-cultured with the hVECs monolayer at MOI of 2:1. Post 60-min infection, ~80% of TH17 but little T1 trophozoites bound to the hVECs monolayer (Figure 1A). Most T1 trophozoites maintained an oval-shaped flagellate form, but ~60% of TH17 trophozoites transformed into a flat disk or irregular ameboid form and tightly adhered to the slide surface (Figure 1B). To further observe the dynamics of host-parasite interaction, the trophozoites co-cultured with hVECs were monitored by time-lapse imaging (Figure 1C and Videos 1 and 2), showing that nonadherent T1 trophozoites maintained a flagellate form and swam by flagellar locomotion only randomly coming into contact with the hVECs. By contrast, adherent TH17 trophozoites rapidly transformed into an amoeboid form within 10 min of contact with the glass slide and crawled toward hVECs via pseudopod-like protrusions, referred to as amoeboid migration. In contrast to T. vaginalis nonadherent isolate, the adherent isolate displayed more active cytoadherence and amoeboid morphogenesis and migration.
(Please see the attached Video 1)
Video1 Dynamics of amoeboid morphogenesis and migration in the adherent isolate of T. vaginalis. The trophozoites from TH17 adherent isolate were co-cultured with hVECs. The dynamics of trophozoite activities were recorded by time-lapse imaging at the capturing rate of one frame per 30 sec over time as defined.
(Please see the attached Video 2)
Video2 Dynamics of migration in the nonadherent isolate of T. vaginalis. The trophozoites from nonadherent T1 isolate were co-cultured with hVECs. The dynamics of trophozoite activities were recorded by time-lapse imaging at the capturing rate of one frame per 30 sec over time as defined.
Differential expression of actin-related proteins in T. vaginalis
Cytoadherence and migration in T. vaginalis correlate with actin cytoskeleton (17, 35), therefore, the expression of α-actin and α-actinin, the respective major component and actin bundle linker protein in the cytoskeleton were investigated. The expression of α-actin and α-actinin varied between isolates and was higher in adherent TH17 and T016 isolates compared to nonadherent T1, special in a fresh adherent isolate from a clinical vaginitis patient (Figure 2A). The overexpression of HA-Tvactin in a nonadherent isolate did not induce amoeboid morphogenesis or cytoadherence (Figure 2-Figure Supplement 1), suggesting that α-actin might be determinant but insufficient to confer cytoadherence in a nonadherent isolate. Additionally, no detectable α-actin on the adherent parasite surface indicates that α-actin is unlikely to act as an adhesion molecule (Figure 2-Figure Supplement 2).
The immunostaining of α-actin was more intense in TH17 than T1 trophozoites and detected in tiny punctate or short bundles in the cytoplasm of the free-swimming flagellate TH17 but in dense fine networks in the cytoplasm with sporadic clumps underneath the plasma membrane of the amoeboid-adhered TH17. However, the expression of α-actin and α-actinin was similar between the two forms of TH17 trophozoites according to western blotting (Figure 2C). The validated phalloidin binding sites are conserved in α-actin of T. vaginalis (Figure 2-Figure Supplement 3) (36). F-actin was double-stained by TRITC-conjugated phalloidin and anti-α-actin antibody (Figure 2D), showing prominent F-actin and α-actin signals concentrated in the juxtanuclear region, referred to as perinuclear actin cap (39), with intense staining underneath the cell membrane of the leading edge in protrusive pseudopods, and less intense staining in the cytoplasm. The signal colocalization of α-actin and phalloidin had a Pearson’s correlation coefficient value of 0.95 (Figure 2D, bottom panel). To evaluate F-actin assembly in cells, G-actin, F-actin, and co-sediments were fractionated, and western blotting analysis revealed an F-actin ratio of ~70% in the adherent isolate and ~30% in the nonadherent one (Figure 2E), similar to α-actinin. It is speculated that F-actin polymerization is more active in the adherent than the nonadherent isolate, and that the actin assembly pattern is also distinct.
Actin-based morphogenesis, migration, and cytoadherence in T. vaginalis
Latrunculin B (LatB) binding sites are conserved in α-actin of T. vaginalis (Figure 2-Figure Supplement 3), therefore adherent TH17 trophozoites were treated with LatB to study the role of F-actin in cytoskeleton behavior and cytoadherence. LatB treatment reduced the ratio of F-actin assembly (Figure 3A) and morphogenesis (Figure 3B) in the parasite compared to the DMSO control, as well as decreasing the wound closure rate (Figure 3C) and cytoadherence 60-min post-infection (Figure 3D), showing that F-actin disorder retarded morphogenesis, amoeboid migration, and cytoadherence of this parasite. To rule out the effects from the reputed adhesion molecules (2, 7, 10), the expression of AP65 and PFO (Figure 3E), and their surface distributions (Figure 3F) were analyzed, showing that there was no change in adhesion molecules in the trophozoites with or without LatB treatment. Under IFA permeation condition, hydrogenosomal colocalization of AP65 and PFO proved their surface signal specificities (7, 10). Also, the surface localization of HA-tagged Cadherin-like protein (CLP) was not affected by LatB treatment (Figure 3G). Taken together, actin polymerization is positively associated with the parasite morphological transition, amoeboid migration, and cytoadherence. Also, LatB-inhibited cytoadherence might be independent of adhesion molecules.
TvFACPα as an α-actin effectors
Since α-actin is not sufficient to promote the cytoadherence in T. vaginalis nonadherence isolate, we attempted to identify the regulatory proteins in the α-actin-associated complexes. HA-Tvactin was immunoprecipitated from transgenic TH17 trophozoites and subjected to mass spectrometry analysis (Figure 4A), identifying 41 α-actin-associated proteins with an emPAI score above 0.25 or specific in the immunoprecipitant of HA-Tvactin (Table 1). These proteins were classified by function into multiple cellular pathways, including cytoskeleton proteins (22%), chaperones (5%), membrane trafficking and transporter (10%), protein binding or modification (7%), DNA/RNA regulation and translation (17%), metabolism enzymes (37%), and uncharacterized proteins (2%) (Figure 4-Figure Supplement 1). The top five abundant protein identified in IP proteome were listed in Figure 4B. Bait HA-Tvactin was identified with an emPAI score of ~9.5, an F-actin CP subunit α homolog, referred to as TvFACPα (TVAG_470230), had an emPAI score of ~9.7 (Figure 4B) and 40% identified peptide coverage (Figure 4C), supporting the possibility of a strong protein-protein interaction between TvFACPα and Tvactin. The in-silico protein sequence analysis revealed that TvFACPα encodes 267 amino acids with a molecular weight of 29.1 kDa and a PI value of 5.43 and shares 17% identity and 63% similarity with CPα from high eukaryotes (Figure 4D). TvFACPα contains a highly conserved actin-binding domain at C-terminus spanning amino acids from 237 to 261. By a phosphorylation site prediction algorithm (NetPhos 3.1 Generic phosphorylation prediction: Services.healthtech.dtu.dk/service.php?NetPhos-3.1), Ser2, Ser46, Ser88, Ser106, and Ser223 were predicted as CKII phosphorylation sites. The sequence of 2SESE5 fits the putative CKII phosphorylation motif (pS/pTDXE) possibly recognized by a phospho-CKII substrate antibody. In the TrichDB database, BLAST analysis identified two CPα homologous proteins (TVAG_470230 and TVAG_212270) with 32% sequence similarity (Figure 4-Figure Supplement 2) but whether they are functionally redundant in this parasite remains to be studied.
The non-canonical interaction of TvFACPα to α-actin
Immunoprecipitation was performed to examine whether TvFACPα forms the protein complexes with α-actin in T. vaginalis, with two major bands at ~30 and ~32 kDa recognized by an anti-HA antibody in the total lysates. A 42-kDa α-actin and a 110-kDa α-actinin band were co-immunoprecipitated from the trophozoites overexpressing HA-TvFACPα but not the non-transfectant control (Figure 5A). To further confirm the direct interaction of TvFACPα and α-actin, His-TvFACPα, His-△237 mutant, GST and GST-Tvactin were purified to homogeneity for the pull-down assay (Figure 5B, left panel). When an equal amount of His-TvFACPα and His-△237 were reacted with GST or GST-Tvactin for the pull-down assay, the signal from His-TvFACPα but not His-△237 was pulled down with GST-Tvactin, showing that the c-terminal domain is vital for the direct binding of TvFACPα and α-actin (Figure 5B, right panel).
The function of TvFACPα in actin assembly was analyzed by an in vitro polymerization assay. When over ~95% G-actin polymerized into F-actin in F-buffer in the absence of His-TvFACPα (Figure 5C left panel), F-actin polymerization ratio was only ~25% in the presence of His-TvFACPα, of which 25% of His-TvFACPα co-sedimented with F-actin. By contrast, the polymerization ratio was 75% in the presence of His-△237, and less than 5% of His-△237 could be co-sedimented with F-actin (Figure 5C), indicating that TvFACPα directly interacts with actin molecules to attenuate polymerization. Of note, only 25% of TvFACPα co-sedimented with F-actin but it inhibited over ~70% F-actin formation in the polymerization assay, suggesting that TvFACPα also binds G-actin to inhibit its polymerization.
To determine the kinetics of TvFACPα binding G-actin and F-actin by a solid phase binding immunoassay, two forms of actin were reacted with various concentrations of His-TvFACPα or His-△237 mutant to measure the Kd and Bmax values (Figure 5D). The binding signal increased with increasing concentration of His-TvFACPα or derived mutant, and plateaued in the presence of over 20 μM of His-TvFACPα. The binding curves show that His-TvFACPα binds F-actin with a Kd of 8.36 μM and Bmax of 1.35 and G-actin with a Kd of 4.45 μM and Bmax of 1.5. By contrast, His-△237 binds both F-actin and G-actin with a similar Bmax of ~0.5, only one-third of His-TvFACPα (Figure 5D inset table). In contrast with the canonical F-actin binding preference for high eukaryotic CPα, the in vitro assays demonstrated that TvFACPα bound G-actin with an affinity greater than F-actin to suppress actin polymerization.
TvFACPα represses F-actin assembly in T. vaginalis
By western blotting, the anti-TvFACPα antibody identified a ~30-kDa protein band in the total lysate from TH17 trophozoites (Figure 6A) and colocalized with TRITC-phalloidin with the Pearson’s correlation coefficient value of 0.96, indicating the colocalization of TvFACPα with F-actin in this parasite (Figure 6B). To further study if TvFACPα regulates F-actin polymerization, HA-TvFACPα wild type or actin-binding domain deletion mutant △237 were overexpressed in TH17 trophozoites. By IFA, HA-TvFACPα was detected as network-like structures extending extensively into the cytoplasm and slightly intense immunostaining condensed near the cell membrane (Figure 6C). In the non-transgenic control TH17, α-actin was distributed in the cytoplasm as fine-dense tubular networks. However, cytoplasmic α-actin was observed as numerous stubby rods with punctate signals in HA-TvFACPα-overexpressed TH17, and the pattern in △237 transfectants was similar to that in the non-transgenic TH17 control, indicating that TvFACPα overexpression may alter α-actin organization in this parasite.
In western blotting, HA-TvFACPα or △237 were overexpressed at a level ~5-fold higher than the endogenous form in the non-transgenic control, and the former inhibited endogenous TvFACPα expression in the transfectant (Figure 6D), suggesting that this parasite may have a feedback mechanism to maintain cellular TvFACPα levels. Western blotting showed that the expression of α-actin or α-actinin did not change between transfectants (Figure 6D). Actin fractionation revealed that ~45% F-actin co-sedimented with ~25% TvFACPα in the non-transgenic TH17 control. In transfectants overexpressing HA-TvFACPα, the F-actin level reduced to ~25% but co-sedimented HA-TvFACPα was ~2-fold higher than the endogenous form of the non-transfectant. In the △237 mutant, the F-actin ratio was slightly higher but co-sedimented △237 was lower than the non-transfectant, therefore, TvFACPα may repress actin polymerization. A similar trend was observed for α-actinin. By immunoprecipitation, co-precipitated α-actin and α-actinin were detected in HA-TvFACPα but much less in the △237 mutant (Figure 6E), indicating that actin-binding activity is essential for TvFACPα to inhibit actin assembly.
TvFACPα function in actin polymerization is regulated by CKII signaling
Compared to the nonadherent T1 isolate, more TvFACPα and α-actin were detected in adherent TH17 isolates but less TvFACPα co-sedimented with F-actin (Figure 7-Figure Supplement 1). The immunostaining of α-actin was different between the flagellate and amoeboid forms of the adherent isolate (Figure 2B), with equal amounts of TvFACPα, α-actin, and α-actinin detected in the total lysates (Figure 7A). The F-actin ratio in the amoeboid trophozoites was two-fold higher than the flagellate form (Figure 7A), whereas the TvFACPα co-sedimented with F-actin in amoeboid trophozoites was two-fold lower than flagellate form. A similar trend was observed for α-actinin, indicating that adhered-amoeboid T. vaginalis displays more active F-actin polymerization and less TvFACPα binding α-actin.
Regarding the post-translation modifications of TvFACPα, TvFACPα Ser2 was previously predicted as a CKII phosphorylation sites (Figure 4D) potentially recognized by a phospho-motif (pS/pTDXE)-specific antibody, referred to as TvFACP(pS2). When TvFACPα was equally immunoprecipitated from the trophozoites, more TvFACP(pS2) but less α-actin and α-actinin were co-pulled down from the amoeboid trophozoites than the flagellate form (Figure 7B). Ser2 hyper-phosphorylation enriched in the amoeboid form trophozoites, in which TvFACPα binding α-actin or α-actinin was low. To confirm the role of Ser2 phosphorylation in the complex formation of TvFACPα and α-actin, hypo-phosphorylation mimic S2A or hyper-phosphorylation mimic S2D mutant were introduced into TH17 trophozoites for actin fractionation and immunoprecipitation. The overall level of α-actin and α-actinin were similar in the total lysates from TvFACPα, S2A, and S2D transfectants. Compared to the non-transgenic control, both HA-TvFACPα and S2A overexpression repressed F-actin levels in the transfectants, with higher levels of co-sedimented HA-TvFACPα or S2A in the F-actin fraction (Figure 7C). By contrast, a similar level of F-actin was detected in the non-transfectant and S2D mutants but co-sedimented S2D in the F-actin fraction was lower than HA-TvFACPα (Figure 7D). Similar results were obtained for α-actinin. Furthermore, α-actin signals co-immunoprecipitated from the S2A and S2D mutant were three-fold higher and 70% lower respectively than HA-TvFACPα (Figures 7E and 7F), with the low signal intensity of α-actin co-immunoprecipitated with △237 mutant, implying that Ser2 phosphorylation is crucial for the actin-binding activity of TvFACPα. Meanwhile, the low intensity of TvFACP(pS2) signal precipitated from △237 mutant implying that the actin-binding domain integrity might be important for Ser2 phosphorylation. Ser2 phosphorylation is a major signal for the dissociation of TvFACPα and α-actin. The undetectable TvFACP(pS2) signal in the S2A or S2D mutant proves the antibody specificity.
To verify whether Ser2 phosphorylation is regulated by CKII signaling, TH17 trophozoites overexpressing HA-TvFACPα were treated with DMSO or TBB for immunoprecipitation and actin fractionation, showing that the overall expression of HA-TvFACPα or α-actin was not influenced by TBB treatment. When an equal amount of HA-TvFACPα was immunoprecipitated from the trophozoites treated with or without TBB, decreasing TvFACP(pS2) but increasing α-actin signals were detected in the co-immunoprecipitants from the parasite treated by TBB (Figure 7G). Furthermore, the overall expression of TvFACPα, α-actin, and α-actinin remained constant in TH17 trophozoites with or without TBB treatment, and when F-actin in TBB-treated parasite was inhibited to one-third of the basal level, the TvFACPα co-sedimented with F-actin was three-fold higher than the DMSO control (Figure 7H). In summary, CKII-dependent Ser2 phosphorylation triggers dissociation of TvFACPα and α-actin to evoke actin polymerization.
TvFACPα in morphogenesis and cytoadherence of T. vaginalis
To examine the role of Ser2 phosphorylation on cytoskeleton behaviors, the morphogenesis of TH17 trophozoites overexpressing HA-TvFACPα and derived mutants was observed by phase-contrast microscopy. Morphogenesis in the trophozoites overexpressing HA-TvFACPα and S2A was reduced to ~20% compared to ~70% morphogenesis in the non-transgenic control, whereas it was restored to ~70% in the △237 and S2D mutants (Figure 8A). TBB treatment also reduced the morphological transformation of TH17 trophozoites from ~80% in DMSO control cells to ~30% in the TBB-treated trophozoites. Notably, the TBB effect inhibiting morphogenesis was abolished in the S2D transfectant, suggesting that CKII-dependent Ser2 phosphorylation in TvFACPα is crucial to the regulation of morphogenesis in T. vaginalis (Figure 8B). The differential cytoadherence of various HA-TvFACPα transfectants was monitored over time, showing that the non-transgenic TH17 trophozoites achieved ~100% cytoadherence 60-min post-infection, reducing to ~40% in HA-TvFACPα and S2A transfectants and increasing to ~80% in △237 and S2D transfectants. TBB treatment also significantly reduced the cytoadherence 60-min post-infection and this effect was abrogated in the S2D transfectant (Figure 8D). Notably, the overexpression of HA-TvFACP and related mutants or TBB treatment did not affect the cytoadherence at the initial 20-min infection (Figure 8C). These findings were consistent with our previous observation that LatB only perturbed cytoadherence from the staging 60-min post-infection (Figure 3D). The data strongly supports that CKII-dependent Ser2 phosphorylation regulates TvFACPα function in cytoskeleton-mediated morphogenesis and consequential cytoadherence of T. vaginalis. The morphogenesis capacity of this parasite tightly correlates its cytoadherence.
The function of TvFACPα in amoeboid migration
The conversion of morphology and motility is the dominant features in adherent isolates (Video 1) and retarded by LatB (Figure 3C), so we investigated the role of TvFACPα in amoeboid migration. Since cytoskeletal disorder retarded the morphogenesis and reduced the adherent activity of T. vaginalis, the conditional trophozoites had to be sufficiently cultured in the T25 flask until forming a confluent parasite monolayer for the wound heal assay. The wound recovery rate was significantly suppressed in the TH17 trophozoites overexpressing HA-TvFACPα but the rate was similar in the non-transgenic control and △237 mutant, indicating that actin-binding activity is essential for TvFACPα to reduce the amoeboid migration (Figure 9A). Also, the wound recovery rate in the non-transgenic parasite was inhibited by TBB to a similar level to the HA-TvFACPα transfectant. By contrast, the wound closure rate in the S2D mutant was similar to the non-transgenic parasite and not influenced by TBB treatment (Figure 9B), revealing that the S2D mutant counteracted the TBB inhibitory effect on amoeboid migration. This observation indicates that CKII-dependent Ser2 phosphorylation might play a key role in TvFACPα-regulated amoeboid migration.
TvFACPα regulates motility switching in T. vaginalis
Next, we tested whether parasite motility is changed with the morphology transition using the trans-well system (Figure 9C). The relative GAPDH signal in the western blotting indicates the relative amount of migratory trophozoites between the bottom wells and top inserts. When GAPDH expression was equal in the input trophozoites, the HA signal was also similar between the transfectants. Focusing GAPDH signal from the bottom well of the 30-min trans-well plate, HA-TvFACPα was higher but △237 and S2D mutants were lower than the non-transgenic control, revealing that more trophozoites with HA-TvFACPα overexpression migrated into the bottom well in a short time (Figure 9D). As observed by microscopy, the trophozoites in the bottom well displayed the morphology at the free-swimming flagellate form (Figure 9-Figure Supplement 1), suggesting that HA-TvFACPα overexpression may retain the parasite in the flagellate form with faster movement driven by motile flagellum. The motility conversion involved actin binding activity regulated by Ser2 phosphorylation. TBB inhibited Ser2 phosphorylation in TvFACPα (Figure 7G), with the GAPDH signal from the TBB-treated trophozoites in the bottom well higher than in the DMSO-treated trophozoites (Figure 9D). Together, TvFACPα Ser2 hypo-phosphorylation retarded amoeboid migration in the adhered trophozoites but expanded the population of free trophozoites that rapidly moved via flagellar locomotion (Figure 9E).
Discussion
TvFACPα was identified as an actin-binding protein that suppressed actin polymerization via the direct interaction with G-actin monomers and F-actin polymers. Furthermore, CKII-dependent signaling plays a key in the switch from morphology and motility. These cytoskeleton-mediated behaviors are crucial for optimizing the cytoadherence and population spread of this parasite. In the human urogenital tract, the intermittent flushing action of body fluid generates a mechanical barrier to either impair or eliminate the retention of uropathogenic microbes, therefore switching to the opportune motility mode to instantly counteract the environmental challenges or physical defenses would be beneficial for T. vaginalis colonization (37).
Unfortunately, the real-time tracking system for fluorescence protein within a living parasite did not work in our assay system, so the overall actin assembly and cytoskeleton activities were evaluated by western blotting and IFA to show the relevance of TvFACPα and actin dynamics in the adherent trophozoites under a steady-state condition.
The DNA sequences of the tvfacpα gene from nonadherent and adherent T. vaginalis isolates share 100% identity (38), thus the differential cytoskeleton behaviors between isolates are unlikely to be attributed to sequence polymorphism in TvFACPα. Meanwhile, α-actin overexpression dose not promote adherence in the nonadherent isolate, thus cytoskeleton-dependent cytoadherence is unlikely to be determined by one single molecule.
Compared to the nonadherent T1 isolate, more TvFACPα and α-actin were detected in adherent TH17 isolates but less TvFACPα co-sedimented with F-actin (Figure 7-Figure Supplement 1), possibly explaining why the adherent isolate displays more active cytoskeleton behaviors than the nonadherent isolate. Furthermore, the adherent isolate may require a larger TvFACPα reservoir to immediately modulate cytoskeleton dynamics in response to sudden environmental challenges.
The perinuclear actin cap was observed in the trophozoite with dividing nuclei. One of the known functions of the perinuclear actin cap is to govern nuclear location and movement during nuclear division (39), therefore F-actin may function in the nuclear division of this parasite. When there was colocalization of TvFACPα and F-actin at the leading edge of the extending pseudopodia, there was less colocalization observed near the actin cap (Figure 6), suggesting that F-actin bundle assembly in peripheral motile structure is presumably manipulated by TvFACPα, distinct from that in the central juxtanuclear actin cap. Human CKIP-1 protein containing pleckstrin homology domain directs CPα to the cell membrane periphery and bridges the interaction of CPα with CKII kinase to co-regulate cell morphology (33, 34).
TvFACPα Ser2 identified as a CKII phosphorylation site is conserved with Ser9 on human or yeast CPα (Figure 4D) (33, 34, 40). Human CPα Ser9 has been demonstrated to be phosphorylated by CKII kinase but does not directly affect actin assembly (33), indicating that the regulation of human CPα is divergent to TvFACPα in this early evolutionary-branched protozoan. Also, yeast CPα Ser9 resides in the stalk domain but not the actin-binding domain, thus Ser2 phosphorylation may not directly interfere with TvFACPα actin-binding, instead altering function by an allosteric effect or binding with other CPα interacting partners to co-regulate actin dynamics (40).
Iron was previously found to trigger a protein kinase A-dependent signaling to activate the Myb3 transcription factor sequential phosphorylation and ubiquitination essential to its nuclear translocation (48). However, iron was observed to slightly change T. vaginalis morphogenesis long-term cultured in iron-restricted growth medium, so whether iron triggers the CKII pathway to regulate cytoskeleton dynamics in this parasite remains to be elucidated.
When the gain-or loss-of-function assay was employed to study the role of Ser2 phosphorylation, F-actin assembly was repressed in the hypo-phosphorylation mimic S2A mutant but restored to near the basal level instead of exceeding it in the hyper-phosphorylation mimic S2D mutant. This implies the existence of additional pathways promoting F-actin assembly under our tested conditions. For example, TvFim1 protein reveals an opposite function to TvFACPα to accelerate F-actin polymerization that favors phagocytosis and migration in T. vaginalis (35). In TrichDB database, BLAST analysis identified two CPα homologous proteins (TVAG_470230 and TVAG_212270) with 32% sequence similarity (Figure 4-Figure Supplement 2) but whether they are functionally redundant in this parasite remains to be studied.
A previous proteomic study reported that surface fibronectin-binding might change actin expression in this parasite. In this report, α-actin expression was constant in the free-swimming flagellate or adhered-amoeboid forms, implying less involvement of fibronectin-binding in the morphogenesis and cytoadherence under our test condition (15).
Mass spectrometry data revealed GAPDH as a major interacting partner of Tvactin. In chicken neuron cells, GAPDH acts as a chaperone for α-actin and co-translocates with α-actin to specialized axon sites for polymerization (41). In yeast, GAPDH associates with α-actin and RpB7 subunit of RNA polymerase II to regulate transcription (42, 43). The significance of GAPDH complexed with the actin cytoskeleton in T. vaginalis remains to be studied.
The EC50 of TBB is varied in different cell types. Numerous CKII alpha subunit (CKIIα) proteins predicted from TrichDB shared less sequence consensus in the TBB binding pocket to high eukaryotic CKIIα (38, 44), possibly explaining why Ser2 phosphorylation and downstream cytoskeleton activities were partially inhibited by TBB treatment. Again, the S2D mutation was unable to promote actin polymerization efficiency beyond that of the non-transgenic parasite, suggesting that actin filament growth might be modulated by additional pathways.
The opportunistic amoeba, Naegleria fowleri, exists in three life stages: flagellate trophozoite, amoeba trophozoite, and cyst. Multiple environmental factors, like growth temperature, cation level, steroid hormone, or chemical agents, affect the flagellate to amoeba transformation (45, 46, 47). In T. vaginalis, other than the contact-dependent effect (14, 18, 35), the factors that trigger the morphological transition are virtually unknown. Overexpression of actin increases the phagocytosis and cytotoxicity of N. fowleri (20) but does not affect T. vaginalis. (Figure 2-Figure Supplement 1). Although they have the cognate behavior of flagellate-amoeba conversion, their regulation in these two protozoa is distinct. The immediate conversion to motility may allow the parasite to rapidly respond to environmental fluctuations or flushing by humoral fluid flow in the urogenital tract (37).
LatB had little effect on the initial 20-min cytoadherence and surface expression of adhesion molecules, AP, PFO, and cadherin, thus we speculate that the adhesion molecules on the cell surface may play roles in the initial cytoadherence, thereafter actin-based morphogenesis reinforces cytoadherence at the later stage of cytoadherence (49). In our previous study, TvCyP2 was demonstrated to shuttle between intracellular membrane compartments, involving the endoplasmic reticulum, Golgi apparatus, and hydrogenosome before translocation onto the cell membrane (50). The cell surface presentation of adhesion proteins may occur through similar endomembrane trafficking routes.
Conclusion
In conclusion, TvFACPα directly binds G- or F-actin to block actin filament extension (Figure 10), with Ser2 phosphorylation on TvFACPα decreasing actin-binding activity and triggering actin polymerization. In adherent T. vaginalis trophozoites, TvFACPα spatially colocalizes with actin molecules at the membrane periphery of motile protrusive pseudopodia, where TvFACPα regulates actin assembly dynamics to control the cytoskeleton behaviors of motility switching, amoeboid migration, or cytoadherence consequent to the morphogenesis. The Ser2 phosphorylation status is crucial for TvFACPα function in the regulation of cytoskeleton behaviors. The cytoskeleton-driven activities are also inhibited by a cytoskeleton (LatB) or CKII (TBB) inhibitor. These findings may provide potential therapeutic targets for cytoskeleton aspects to prevent T. vaginalis colonization and transmission.
Materials and Methods
Cell cultures
T. vaginalis trophozoites were cultured in TYI medium at 37°C (48). Two T. vaginalis isolates, nonadherent T1 (48) and adherent TH17, were used in this study. T1 with only flagellate trophozoites freely swim in the medium suspension. TH17 displayed vigorous morphogenesis and tightly adhered on glass surface of culture tube. Once the void surface is saturated by adhered trophozoites, the unbound parasite at the flagellate form freely swims in the medium suspension (Figure 1 and Videos 1 and 2). The flagellate trophozoites in the medium suspension and adherent trophozoite on the culture tube surface were collected for analysis as described below. Human vaginal epithelium cells (hVECs, VK2/E6E7) were cultivated in Keratinocyte-Serum Free medium (Thermo Fisher Scientific, Massachusetts, USA) at 37°C in 5% CO2 as the suggestion by ATCC.
Lysate preparation from adherent-amoeboid and nonadherent-flagellate trophozoites
Approximately 2×107 trophozoites from TH17 adherent isolate were inoculated into culture tube with 15 ml of medium and incubated at 37°C for 2 hr. The free-trophozoites in suspension were transferred to a new tube and recovered by centrifugation. The cell pellet was lysed in 1 ml lysis buffer (1% Triton X-100, 1× Protease inhibitor cocktail, 1×Phophatase inhibitor cocktail, 100 μg ml-1 TLCK, 5 mM EDTA, in TBS). The trophozoites adhering to the glass tube were directly lysed by adding 1 ml lysis buffer and vigorously vortexing for 5 min at 4°C.
Plasmid construction
The full-length coding sequence of the tvfacpα gene (TVAG_470230) was amplified from T. vaginalis genomic DNA using the primer pair of TvFACPα-BamHI-5′ and TvFACPα-XhoI-3′. The PCR product was gel-purified, then digested by BamHI/XhoI, and ligated into BamHI/XhoI-predigested Flp-HA-TvCyP2 or pET28a backbone plasmid to obtain Flp-HA-TvFACPα or pET28-His-TvFACPα plasmid. Following a similar procedure, the DNA fragments were amplified from Flp-HA-TvFACPα individually using the primer pairs, TvFACPαS2A-5′ and TvFACPα-XhoI-3′ for the S2A mutation, TvFACPαS2D-5′ and TvFACPα-XhoI-3′ for the S2D mutation, and TvFACPα-BamHI-5′ and TvFACPα△237-3′ for the actin-binding domain deletion mutant (△237). The PCR products were gel-purified and subcloned into Flp-HA-TvFACPα or pET28a backbone with BamHI/XhoI sites to generate Flp-HA-TvFACPα(S2A), Flp-HA-TvFACPα(S2D), Flp-HA-TvFACPα(△237), or pET28-His-TvFACPα(△237) plasmid.
To express HA-tagged α-actin in T. vaginalis or glutathione S-transferase (GST) fused-α-actin for the GST pull-down or actin polymerization assays, the full-length coding sequence of the tvactin gene (TVAG_337240) was amplified from T. vaginalis genomic DNA by the primer pair of Tvactin-BamHI-5′ and Tvactin-XhoI-3′. The gel-purified PCR product was digested with BamHI and XhoI, then ligated into BamHI and XhoI-predigested Flp-HA-TvFACPα or pGST-TvCyP2 plasmid (50) to generate Flp-HA-Tvactin, or pGST-Tvactin plasmid.
The TvCadherin expression plasmid was constructed, the coding sequence of the tvcadherin gene (TVAG_393390) (2) was amplified from T. vaginalis genomic DNA by the primer pair of TvCadherin-BamHI-5′ and TvCadherin-XhoI-3′, and subcloned into Flp-HA-TvFACPα backbone vector with BamHI and XhoI sites to produce Flp-HA-TvCadherin plasmid.
The primer oligonucleotides used in this study.
Cytoadherence binding assay
hVECs was cultured in a 24-well plate to an 85% confluent monolayer. Mid-log phase T. vaginalis prelabeled with 5 μM of carboxyfluorescein diacetate succinimidyl ester dye (CFSE; CellTraceTM, Thermo Fisher Scientific, Massachusetts, USA), were inoculated by a multiplicity of infection (MOI) of 2:1 into hVECs culture. At the specific time point, the medium was aspirated and unbound trophozoites were removed by washing two times with PBS for 5 min each. Samples were fixed in 4% formaldehyde for fluorescence microscopy.
Real-time microscopy
The activity of trophozoites on the confluent hVECs monolayer in a glass-bottom culture dish was monitored in real-time by confocal microscopy (LSM-700, Zeiss, Oberkochen, Germany) under a phase-contrast mode with the sampling rate at one frame per 15 sec over time as indicated.
Inhibitor treatment
1 μM of LatB (Sigma-Aldrich, Massachusetts, USA) or 250 μM of TBB (Sigma-Aldrich, Massachusetts, USA) was added into the T. vaginalis culture and incubated at 37°C for 2 hr before analysis.
Morphology analysis
Trophozoites were cultured on a glass slide in a humid chamber at 37°C for 1 hr and the morphology was observed by phase-contrast microscopy (CKX31, Olympus, Tokyo, Japan). The percentage of flagellate or amoeboid form was measured from 600 trophozoites within 12 random microscopic fields.
Immunofluorescence assay (IFA)
T. vaginalis was fixed with 4% formaldehyde and permeabilized with 0.2% Triton X-100. The samples were then incubated with the primary antibodies: rabbit anti-α-actin (200×, GenScript, New Jersey, USA), mouse anti-α-actin (400×, Abcam Ac-40, Cambridge, UK), mouse anti-HA (200×, Sigma-Aldrich HA-7, Massachusetts, USA), mouse anti-AP65 (7), rabbit anti-PFO (10), rabbit anti-TvFACPα, followed by reaction with FITC or Cy3-conjugated goat anti-mouse or rabbit IgG secondary antibodies (200×, Jackson ImmunoResearch, Pennsylvania, USA). The specimens were air-dried and mounted in medium with DAPI (Vector laboratories, California, USA) for observation by confocal microscopy (LSM-700, Zeiss, Oberkochen, Germany).
F-actin staining
Trophozoites were fixed with 4% formaldehyde, then permeabilized with 0.2% Triton X-100. The sample was incubated with 20 μg ml-1 of TRITC-conjugated Phalloidin (Sigma-Aldrich, Massachusetts, USA) diluted in PBS with 1% BSA in the dark at room temperature for 1 hr. After washing three times with PBS, the glass slide was air-dried and mounted in anti-fade medium (Vector laboratories, California, USA) for fluorescence microscopy (BX-60, Olympus, Tokyo, Japan).
Signal colocalization evaluation
The fluorescent intensity distributed in the fluorescence assays was measured by plot analysis of ImageJ (Version 1.53q, National Institutes of Health, Maryland, USA). Pearson’s correlation coefficient was calculated to evaluate the signal co-localization, with a value of 1 indicating perfect colocalization, −1 indicating anti-correlation, and 0 representing no correlation.
Western blotting
The protein samples denatured in 1x SDS sample buffer were separated by sodium dodecyl sulfate-polyacrylamide gel electrophoresis (SDS PAGE) in a 12% gel before blotted to polyvinylidene difluoride (PVDF) membrane by the wet transblot system (Bio-Rad, California, USA). The blocked membrane was incubated with the primary antibodies: mouse anti-HA (2,000×, Sigma-Aldrich HA-7, Massachusetts, USA), mouse anti-α-actin (20,000×, Abcam Ac-40, Cambridge, UK), mouse anti-TvCyP2 (5,000×) (50), mouse anti-α-tubulin (10,000×, Sigma-Aldrich DM-1A, Massachusetts, USA), rabbit anti-TvFACPα (3,000×), mouse anti-6×His (2,000×, Abcam AD1.1.10, Cambridge, UK), rabbit anti-phospho-CKII substrate [(pS/pT)DXE] (1,000×, Cell Signaling Technology, Massachusetts, USA), rabbit anti-PFO (5,000×) (10), mouse anti-AP65 (10,000×) (7), mouse anti-GAPDH (10,000×) (51) and mouse anti-α-actinin (5,000×) (51) at 4°C overnight, followed by HRP-conjugated anti-mouse or rabbit IgG secondary antibodies (5,000×, Jackson ImmunoReaearch, Pennsylvania, USA) at 37°C for 1 hr. The membrane reacted with the enhanced chemiluminescence substrate (ECL, Thermo Fisher Scientific, Massachusetts, USA) were detected and quantified by UVP image system (ChemiDoc-It 815 Imager, VisionWorksLS 8.6 software, Analytik Jena Company, Jena, Germany).
Immunoprecipitation
Briefly, 6×107 trophozoites were lysed in 1ml of lysis buffer (1% Triton X-100, 1×Protease inhibitor cocktail, 1×Phophatase inhibitor cocktail, 100 μg ml-1 TLCK, 5 mM EDTA, in TBS) and centrifuged to remove unbroken cell debris, before the addition of 20 μl of anti-HA antibody-conjugated agarose beads (Sigma-Aldrich, Massachusetts, USA), then incubated on a rotator at 4°C overnight. The beads were recovered by centrifugation and washed three times with 1ml lysis buffer. The precipitates were denatured in 1× SDS sample buffer for western blotting or staining (48, 50).
Label-free quantitative proteomic analysis
The proteins separated by SDS-PAGE were fixed in methanol for SYPRO Ruby staining (Thermo Fisher Scientific, Massachusetts, USA) and visualization by the Typhoon9410 imaging system (GE healthcare, Illinois, USA). Each gel lane was equally cut into 4 pieces, then sliced into smaller 1-mm3 cubes. The gel cubes were desalted by five washes sequentially in 1 ml of 20 mM triethylammonium bicarbonate buffer (TEABC) and 1 ml of 20 mM TEABC with 50% acetonitrile, with the vigorous vortex. The samples were sequentially reduced in 20 mM dithiothreitol (DTT) at 56°C for 1 hr, alkylated in 55 mM iodoacetamide in the dark at room temperature for 30 min and digested with trypsin (Promega, Wisconsin, USA) at 37°C overnight. The tryptic peptides were extracted by vortexing three times sequentially in 20%, 50%, and 100% acetonitrile, then dried in a vacuum concentrator (SpeedVac, Thermo Fisher Scientific, Massachusetts, USA) for LC-MS/MS analysis (48). The protein abundance from the mass spectrometry data was analyzed by a label-free quantitative method by Mascot search, which provides an automated calculation of the Exponentially Modified Protein Abundance Index (emPAI) to estimate the coverage of the identified peptides and abundance for each protein in a dataset. The identified proteins with an emPAI above 0.25 or specific in the co-pull-down sample with their function category are summarized in Table 1. The mass spectrometry proteomics raw data have been deposited to Dryad (https://datadryad.org/stash/share/e30mZQElM-nBNmJOniuiGSBJWBkB7V4-t0XzQ891cX8) or the ProteomeXchange Consortium via the PRIDE (www.ebi.ac.uk/pride/) (52) partner repository with a dataset identifier number of PXD034359.
PRIDE Reviewer access account details:
Username: reviewer_pxd034359@ebi.ac.uk, Password: XpCqEnqW
In silico analysis of protein sequence and function
The functions of the proteins identified by mass spectrometry were categorized by Protein Analysis Through Evolutionary Relationships (PANTHER) Classification System (www.pantherdb.org/). The TvFACPα protein homologue was searched in TrichDB (trichdb.org/trichdb/app). The multiple protein sequence alignment was analyzed by the Vector NTI AdvanceR 11.5.1 software (Thermo Fisher Scientific, Massachusetts, USA). The protein search was performed by the Basic Local Alignment Search Tool (BLAST, blast.ncbi.nlm.nih.gov/Blast.cgi) or UniProt (www.uniprot.org/).
Production of recombinant protein
The recombinant protein was produced as previously described (48, 50). The majority of GST-Tvactin was expressed in the inclusion bodies of E. coli (BL21). For the GST-pull-down assay, the inclusion bodies from 200 ml of E. coli culture were dissolved in 1 ml of 8 M urea at 4°C for 20 min to solubilize the proteins. Then, 1 ml of lysate was immediately added to 14 ml PBS and incubated at 4°C for 30 min to refold proteins. After the removal of the insoluble pellets by low-speed centrifugation at 23,000× g, soluble GST-Tvactin was incubated with glutathione-conjugated sepharose beads as suggested by the supplier (GE healthcare, Illinois, USA) at 4°C for 3 hr and then eluted in GST elution buffer (50 mM Tris-HCl, 10 mM reduced glutathione, pH 8.0). For the solid-phase binding and in vitro actin polymerization assays, the bacterial inclusion bodies were solubilized in 8 M urea and directly reconstituted in 14 ml of G-buffer (0.2 mM CaCl2, 0.2 mM ATP, 0.5 mM DTT, 5 mM Tris-HCl pH 8.0) at 4°C overnight. The insoluble materials were removed by ultracentrifugation at 100,000× g to recover the soluble G-actin in the supernatant (53). Soluble G-actin was further purified by glutathione-conjugated sepharose beads and eluted in G-buffer with 10 mM reduced glutathione.
GST pull-down assay
GST pull down assay was performed as previously described (50). Briefly, 80 picomoles of GST or GST-Tvactin immobilized on 20 μl of glutathione conjugated-sepharose beads 4B (GE healthcare, Illinois, USA) was incubated with 80 picomoles of His-TvFACPα or derived mutants in 1 ml GST binding buffer (PBS contains 0.2% Triton X-100 and 1 mM EDTA) at 4°C rotation overnight The GST beads were washed three times by 1 ml of GST binding buffer then denatured in 1× SDS sample buffer for further analysis.
TvFACPα antiserum production
The recombinant His-TvFACPα full-length protein was produced and purified by a standard protocol as suggested by the supplier (QIAGEN, Hilden, Germany) (48, 50). Using the purified His-TvFACPα protein to immunize rabbits for antiserum production is a customized service provided by the manufacturer (Genetex, California, USA). The antibody specificity of anti-TvFACPα was tested by western blotting as shown in Figure 6A.
In vitro actin polymerization and co-sedimentation assay
Insoluble GST-Tvactin denatured in 0.5 ml of 8M urea was reconstituted in 7.5 ml of G-buffer (0.2 mM CaCl2, 0.2 mM ATP, 0.5 mM DTT, 5 mM Tris-HCl, pH 8.0) at 4°C overnight to ensure that the thorough GST-Tvactin depolymerizes into the G-actin form. Then, 80 picomoles of G-actin in 1 ml of G-buffer were added 1/10 volume of 10× F-buffer (500 mM KCl, 20 mM MgCl2, and 10 mM ATP, 100 mM Tris, pH 7.5) at 4°C for 1 hr to trigger actin polymerization. F-actin was recovered from the pellet by 100,000× g ultracentrifugation, whereas soluble G-actin in the supernatant (54). The ratio of F-actin versus G-actin was evaluated with Coomassie blue staining or Western Blot detection. Alternatively, F-actin and co-sediments were recovered from the pellet by 100,000× g ultracentrifugation of the in vitro actin polymerization assay in the presence of 80 picomoles of His-TvFACPα wild type or His-△237 at 4°C for 1 hr.
ELISA-based solid-phase binding assay
The solid-phase binding assay was performed as described previously, with a few modifications (54). Briefly, 100 μl of 2.5 μM G-actin in G-buffer or F-actin in F-buffer was added to a 96-well microplate and incubated at 4°C with gentle shaking for 8 hr. After three washes with PBST (0.05% Tween 20 in PBS), the samples were blocked in PBST with 5% non-fat milk at 37°C for 2 hr before 100 μl aliquots of different concentrations of His-TvFACPα (0, 2.5, 5,10, 20, 40, and 80 μM) were added to the wells and incubated at 4°C with gentle agitation overnight for protein-protein interaction. Unbound protein was removed by three washes with PBST and the plate was incubated with mouse anti-6×His primary antibody (10,000×, in PBST containing 5% non-fat milk) at room temperature for 2 hr, followed by three washes with PBST. The wells were incubated with HRP-conjugated goat anti-mouse IgG secondary antibody (5000× in PBST containing 5% non-fat milk, Jackson ImmunoReaearch, Pennsylvania, USA) at room temperature for 2 hr. The wells were washed before the addition of 100 μl/well of 3, 3’, 5, 5’-tetramethylbenzidine (TMB, Sigma-Aldrich, Massachusetts, USA) substrate at room temperature for 5 min. The colorimetric reaction was stopped by the addition of 100 μl/well 1N HCl and the absorbance was detected by spectrophotometry at OD450 (Molecular Device, California, USA). The absorbances at OD450 were plotted against the concentrations of His-TvFACPα to generate Scatchard plots and calculate Kd and Bmax values (54).
Actin biochemical fractionation
G- and F-actin were fractionated and enriched using a commercial in vivo assay biochem kit (Cytoskeleton Inc, Colorado, USA), according to the manufacturer’s instructions with minor modifications. Briefly, around 3× 107 trophozoites were incubated in cell lysis buffer (Cytoskeleton Inc, Colorado, USA) with vigorous agitation at 4°C for 30 min and homogenized by a 23-gauge needle on a 5-ml syringe. The total lysate was centrifuged at 1,000× g to remove the unbroken cell debris, followed by ultracentrifugation at 100,000× g for 1 hr to separate the insoluble F-actin and associated proteins in the pellet from soluble G-actin in the supernatant. In western blotting, α-tubulin and TvCyP2 were respectively detected as purity markers for F-actin and G-actin fractions.
Cell migration assay
For the wound healing assay, adherent T. vaginalis trophozoites were cultured to a confluent monolayer in a T25 flask. A scratch (200-μm to 1-mm wide) was generated by scraping the trophozoite monolayer with a P200 tip. After removal of cell debris by washing once with the growth medium, the culture flask was incubated at 37°C and images were captured in a defined area at an interval of 30 min over 2 hr. The wound closure area in each image was measured by ImageJ software (Version 1.53q, National Institutes of Health, Maryland, USA). For the trans-well migration assay, ~1×107 trophozoites suspended in 2 ml of TYI medium were inoculated into the top insert divided by a polyester membrane with 3-μm pores (4.6 cm2, JET Biofil, Guangzhou, China). The top insert was placed in a 6-well culture plate containing 2 ml of TYI medium and cultured at 37 °C for 30 min. The trophozoites in the top insert and bottom well were collected for microscopic observation and western blotting.
Statistical analysis
Statistical significance of data collected from control and conditional samples was analyzed by Microsoft Office Excel 2019 software with Student’s t-test. P< 0.05 is considered as significant difference.
Author contributions
Kai-Hsuan Wang: Investigation, Validation, and Methodology.
Jing-Yang Chang: Investigation, Validation, and Methodology.
Fu-An Li: Investigation, Validation, and Methodology.
Yen-Ju Chen: Investigation, Validation, and Methodology.
Kuan-Yi Wu: Investigation, Validation, and Methodology.
Tse-Ling Chu: Investigation and Validation
Jessica Lin: Investigation and Validation
Hong-Ming Hsu: Investigation, Validation, Project Administration, Supervision, Funding Acquisition, Conceptualization, Writing-Original Draft Preparation, and Writing-Review & Editing.
Statement of conflict of interest
The authors declare that they have no competing interests in this manuscript.
Data availability Section
All data generated or analyzed during this study are included in the manuscript and supplementary data; Source Data files have been provided for Table 1, the statistical analysis of quantification, and raw gel or blot images generated in this study. The proteomics raw data have been deposited to Dryad (https://datadryad.org/stash/share/e30mZQElM-nBNmJOniuiGSBJWBkB7V4-t0XzQ891cX8) and PRIDE (www.ebi.ac.uk/pride/) with a dataset identifier number PXD034359. (PRIDE Reviewer access account details: Username: reviewer_pxd034359@ebi.ac.uk, Password: XpCqEnqW).
Supplement data and legends
Acknowledgement
We are grateful to Dr. Jung-Hsiang Tai (Institute of Biomedical Sciences, Academia Sinica, Taiwan) for T. vaginalis T1 isolate, Dr. John Alderete (Washington State University, USA) for the anti-α-actinin, anti-GAPDH, and anti-AP65 antibodies, and Dr. Rossana Arroyo (CINVESTA, Mexico City, Mexico) for the anti-PFO antibody. Also, we are grateful to the Proteomics Core Facility (Institute of Biomedical Sciences, Academia Sinica, Taiwan) for the LC-MS/MS analysis. This work was supported by grants from the Ministry of Science and Technology of Taiwan (110-2320-B-002-048-and 110-2320-B-002-076-).