The TvLEGU-1, a Legumain-Like Cysteine Proteinase, Plays a Key Role in Trichomonas vaginalis Cytoadherence

The goal of this paper was to characterize a Trichomonas vaginalis cysteine proteinase (CP) legumain-1 (TvLEGU-1) and determine its potential role as a virulence factor during T. vaginalis infection. A 30-kDa band, which migrates in three protein spots (pI~6.3, ~6.5, and ~6.7) with a different type and level of phosphorylation, was identified as TvLEGU-1 by one- and two-dimensional Western blot (WB) assays, using a protease-rich trichomonad extract and polyclonal antibodies produced against the recombinant TvLEGU-1 (anti-TvLEGU-1r). Its identification was confirmed by mass spectrometry. Immunofluorescence, cell binding, and WB assays showed that TvLEGU-1 is upregulated by iron at the protein level, localized on the trichomonad surface and in lysosomes and Golgi complex, bound to the surface of HeLa cells, and was found in vaginal secretions. Additionally, the IgG and Fab fractions of the anti-TvLEGU-1r antibody inhibited trichomonal cytoadherence up to 45%. Moreover, the Aza-Peptidyl Michael Acceptor that inhibited legumain proteolytic activity in live parasites also reduced levels of trichomonal cytoadherence up to 80%. In conclusion, our data show that the proteolytic activity of TvLEGU-1 is necessary for trichomonal adherence. Thus, TvLEGU-1 is a novel virulence factor upregulated by iron. This is the first report that a legumain-like CP plays a role in a pathogen cytoadherence.

Iron is an essential nutrient for growth, metabolism, and virulence of T. vaginalis [36]. e environment of the human vagina, especially its nutrients and the iron concentration, is constantly changing throughout the menstrual cycle.
T. vaginalis may respond to varying iron concentrations by differential gene expression through poorly understood mechanisms [20,21,37] in order to survive, grow, and colonize the vaginal hostile environment.
We previously reported that some of the CPs of the 30-kDa region are involved in cytoadherence [4,5,7]. is region is formed by at least six spots with proteolytic activity that correspond to two distinct CP families: the papainlike family of clan CA, represented by four spots with pI between 4.5 and 5.5, and the legumain-like family of clan CD, represented by two spots with pI 6.3 and 6.5 [38] that are differentially regulated by iron at the transcript and proteolytic activity levels [21].
Among the ten legumain-like CP genes reported in the dra of the T. vaginalis genome [9], we have cloned and sequenced two cDNAs coding for the TvLEGU-1 and TvLEGU-2 precursor proteinases of 42.8-and 47.2-kDa. ese CPs were classi�ed within the asparaginyl endopeptidase (AE) subfamily of the family C13, belonging to the clan CD [38]. e family C13 of peptidases includes two distinct subfamilies with different functions, the glycosylphosphatidylinositol (GPI): protein transamidase and the asparaginyl endopeptidase. Interestingly, TvLEGU-1 and TvLEGU-2 share ∼30% amino acid identity with the AE subfamily and ∼26% with the GPI: protein transamidase subfamily [38]. We also showed that the amount of TvLEGU-1 transcript is positively regulated by iron, whereas the TvLEGU-2 mRNA is not affected by it [21]. Additionally, TvLEGU-1 is one of the most immunogenic proteinases detected by trichomoniasis patient sera [15]. us, the main goal of this work was to identify, characterize, and determine the function of TvLEGU-1. Our data show that TvLEGU-1 is a surface proteinase upregulated by iron, with affinity to the surface of HeLa cells that plays a major role in trichomonal cytoadherence. Hence, TvLEGU-1 is a novel virulence factor of T. vaginalis that is also released in vaginal secretions during infection.

Parasites and HeLa Cell Cultures. e fresh clinical
T. vaginalis isolate CNCD 147 [7,15,29] was used in this study. Parasites were kept in culture at 37 ∘ C up to two weeks by daily passage in trypticase-yeast extractmaltose (TYM) medium [39] supplemented with 10% heatinactivated horse serum (HIHS) (TYM-HIHS), containing ∼20 M iron [36]. Parasites in the logarithmic phase were grown either in iron-rich or in iron-depleted medium by the addition into the culture medium of 250 M ferrous ammonium sulfate or 150 M 2-2 dipyridyl (Sigma Co., St Louis, MO, USA) an iron-chelator, respectively, as previously reported [30]. HeLa cells were grown in Dulbecco's Modi�ed Eagle Medium (DMEM) (Gibco Laboratories, Grand Island, NY) supplemented with 10% HIHS at 37 ∘ C for 48 h in a 5% CO 2 atmosphere to obtain con�uent cell monolayers [6].  Rabbits were subcutaneously inoculated four times at twoweek intervals with 0.3 mg of the affinity-puri�ed TvLEGU-1r protein [15] in the presence of Freund's complete adjuvant (Gibco) for the �rst immunization. Booster injections were given in Freund's incomplete adjuvant (Gibco). e immune serum (anti-TvLEGU-1r) was obtained seven days aer the last immunization [40]. is antiserum was used in western blot (WB) analysis, indirect immuno�uorescence, and cytoadherence inhibition assays. Preimmune (PI) serum was obtained before the immunization schedule began and was used as a negative control in all the experiments with antibodies.

Papain Fragmentation of IgG to Fab.
To obtain the Fab fragment, 0.5 mg/mL puri�ed IgGs from the anti-TvLEGU-1r or PI serum [40,41] in PBS pH 8.0 were digested with 0.2 mg/mL papain in digestion buffer (PBS pH 8.0 containing 0.02 M cysteine and 0.02 M EDTA) at 37 ∘ C for 6 h. e reaction was stopped with 0.3 M iodoacetamide in PBS pH 8.0. Aer digestion, samples were dialyzed in PBS pH 7.0 for 18 h at 4 ∘ C and incubated with protein A agarose during 2 h to eliminate the Fc fraction and recover the unbound Fab fragment [40].

Two-Dimensional Gel Electrophoresis (2DE)
. e 2DE for protease-rich extracts was performed as recently described [22]. Brie�y, for the �rst dimension, supernatant from lysed parasites (6 × 10 7 cells/mL equivalent to 500 g protein) in rehydration solution (Bio-Rad) was loaded onto a 7 cm Ready immobilized pH gradient (IPG) strips (linear pH gradient 4-7; Bio-Rad). IPG strips were actively rehydrated for 16 h at 4 ∘ C. Isoelectric focusing (IEF) of proteins was performed in three steps: 250 V for 20 min, 4 000 V for 3 h, and a gradual increase up to 10 000 V-h. For reduction and alkylation, strips were equilibrated in buffer I and II (Bio-Rad) for 10 min at room temperature each. Proteins were resolved by SDS-PAGE using 12% polyacrylamide gels, silverstained, or transferred onto nitrocellulose (NC) membrane for WB detection. Gels and NC membranes were documented using the ChemiDoc-XRS (Bio-Rad) and analyzed using the Quantity One soware (Bio-Rad). A tridimensional analysis using the PD Quest (Bio-Rad) and Melanie soware was also performed for differentially expressed proteins. ree independent protein preparations were done, each obtained from an independent parasite culture, and similar results were observed.
2.�. Proteinase Identi�cation. Identi�cation of protein spots was performed at the Protein Unit of the Columbia University (NY, USA) as before [15]. Protein spots of interest were manually excised from silver-stained gels, distained, and prepared for in-gel digestion with trypsin. Resulting peptides were analyzed by MALDI-TOF mass spectrometry (MS) peptide mass mapping method on a Voyager DE promass spectrometer in the linear mode (Applied Biosystems). Peptide masses were searched against the National Center for Consecutive number assigned to the identi�ed peptides. b Position in amino acids (aa) residues of the identi�ed peptides (start-end) in the aa sequence of the T. vaginalis TvLEGU-1 [38]. c Arbitrary nomenclature used to describe the ten identi�ed peptides in T. vaginalis TvLEGU-1 (see Supplementary Figure 1S in Supplementary Material available online at doi:10.1155/2012/561979). d Peptide mass average (av) identi�ed by MALDI-TOF-MS aer tryptic digestion of the three protein spots obtained from 2DE of protease-rich extracts from T. vaginalis grown in normal iron conditions ( Figure 1). e Presence (+) or absence (−) of the peptides identi�ed by MALDI-TOF-MS in the three TvLEGU-1 protein spots analyzed. f Amino acid sequence of the peptides obtained from a theoretical tryptic digestion of the deduced aa sequence of TvLEGU-1 [38] with identical masses to the experimental one (Supplementary Figure 1S). T 2: Densitometric analysis of the three TvLEGU-1 protein spots observed in silver-stained gels and WB NC membranes from parasites grown in high and low iron concentrations.  Biotechnology nonredundant database (NCBInr) using the MASCOT program (http://matrixscience.com/).
To determine the presence of phosphorylations on TvLEGU-1, anti-phospho-Ser, -Tyr, and -r monoclonal antibodies at 1 : 500 dilution (Zymed) were used in 2DE-WB assays over NC membranes containing protease-rich extracts and developed by chemiluminescence. ese experiments were performed at least three times with similar results.

In Vitro Secretion Kinetic
Assay. e in vitro secretion assay was performed as previously described [22,33]. Brie�y, aer 18 h of growth in iron-rich conditions, parasites were harvested, washed three times with PBS pH 7.0, and suspended in PBS-0.5% maltose at 1 × 10 6 cells/mL parasite density. Parasites were incubated for 15, 30, 60, and 90 min at 37 ∘ C, collected by centrifugation at 700 g, and supernatants were analyzed directly by substrate-gel electrophoresis and by WB aer TCA-precipitation. e viability of trichomonads was assessed by trypan blue exclusion throughout the assay.

2.�. In�irect Imm�no��orescence Assay.
For confocal microscopy, parasites grown in iron-rich conditions were �xed with 4% paraformaldehyde for 1 h at 37 ∘ C, washed with PBS, and half of them were treated with 50 mM NH 4 Cl/PBS pH 7.0 for 10 min, washed with PBS, and with 1 N HCL for 1 h and permeabilized with 0.2% Triton X-100 for 10 min. e other half was used as nonpermeabilized parasites. Permeabilized and nonpermeabilized parasites were blocked with 1% fetal bovine serum for 15 min and with 0.2 M glycine for 1 h at room temperature. en, trichomonads were incubated for 18 h at 4 ∘ C, with the anti-TvLEGU-1r or PI serum used as a negative control, both at 1 : 1 000 dilution. Parasites were incubated with the secondary antibody, �uorescein isothiocyanate-con�ugated anti-rabbit immunoglobulins (Pierce) at 1 : 200 dilution for 1 h at 37 ∘ C, washed, mounted with Vectashield mounting solution (Vector Laboratories), and visualized by confocal microscopy with a Leica LSM-SPC-5 Mo inverted confocal microscope �tted with HC�PLapo lambda blue 63 × 1.4 oil immersion lens. Time series were captured and processed using the confocal LAS AF soware (Leica). Also, live HeLa cells were incubated with 10 g/mL TvLEGU-1r or supernatant from an in vitro secretion assay for 30 min at 37 ∘ C, washed with PBS, �xed, blocked, and treated with antibodies as the parasites described above for immuno�uorescence assays.
For lysosomal colocalization assays, the acidic compartments of T. vaginalis were stained with 1 M LysoTracker RED DND-99 (Invitrogen) for 12 h at 37 ∘ C in TYM medium supplemented with 10% heat-inactivated horse serum. Aer that, parasites were processed for indirect immuno�uorescence with the anti-TvLEGU-1r antibody as described in the previous paragraph.
For immunogold labeling assays, parasites were �xed overnight at room temperature in 0.5% glutaraldehyde, 4% formaldehyde in 0.1 M cacodylate buffer. Aerwards, cells were dehydrated in ethanol and embedded in Unicryl. Ultrathin sections were harvested on 300 mesh nickel grids. e samples were washed and incubated with 50 mM ammonium chloride for 30 min in order to quench free aldehyde groups. e sections were incubated in a series of blocking solutions (PBS containing 1% bovine albumin (BSA), 3% PBS/BSA, and 0.2% Tween-20, pH 8.0) for 10 min on each step. Cells were incubated with the anti-TvLEGU-1r antibody at 1 : 50 dilution, overnight. Aer several washes in 1% PBS/BSA, the sections were incubated with 10 nm gold-labeled goat antirabbit IgG (BB International, UK). As control some samples were incubated only with the secondary antibody. Finally, sections were stained with 5% uranyl acetate and 1% lead citrate and then observed with a JEOL 1210 transmission electron microscope.

Cell-Binding Assay for Proteinases.
To detect the affinity of the native and recombinant TvLEGU-1 proteins to the surface of host cells, we performed cell-binding assays as previously described [7]. Brie�y, a clari�ed detergent extract from 2 × 10 7 parasites or 25 g of TvLEGU-1r was incubated for 18 h at 4 ∘ C with 1 × 10 6 glutaraldehyde-�xed HeLa cells. e native and recombinant TvLEGU-1 proteins bound to the surface of �xed-HeLa cells were eluted with Laemmli sample buffer for 20 min at 37 ∘ C. e released proteins were analyzed by SDS-PAGE and blotted onto NC for WB detection with the anti-TvLEGU-1r antibody.
e cytoadherence inhibition assays with proteinase inhibitors were performed over con�uent live HeLa cell monolayers on 12 mm coverslips as recently described [41]. Brie�y, cell monolayers (5 × 10 5 cells/coverslip) were incubated with live parasites (1 × 10 6 cells/well) previously labeled with 25 mM CellTracker Blue CMAC (Molecular Probes) in serum-free DMEM-TYM (2 : 1) medium and incubated at 37 ∘ C for 30 min and 5% CO 2 . For inhibition experiments before interaction with HeLa cell monolayers labeled parasites were incubated for 20 min at 4 ∘ C with different CP inhibitors (1 mM TLCK, 0.2 mM leupeptin, or 0.18 mM E-64; all purchased from Sigma) used as controls. e Aza-Peptidyl Michael Acceptor (Mu-Ala-Ala-AAsn-CH=CH-CON, kindly donated by Dr. James Powers), a speci�c inhibitor for legumains [43], was also used at 5, 10, and 50 M. Aer interaction with CP inhibitors parasites were washed and added to HeLa cell monolayers. Aer the interaction, the coverslips were washed with warm PBS, �xed with 4% paraformaldehyde, and mounted on slides. Each condition was performed in triplicate, and ten �elds with a 40x magni�cation were analyzed per coverslip. Fluorescent parasites adhered to host cells (in blue) were counted using an Eclipse 80i epi�uorescence microscope (Nikon) and the NIS-Elements BR 2.1 soware (Nikon) ( Table 3). e experiment was repeated at least two independent times with similar results.

Measurement of the Proteolytic Activity of Live Parasites.
To detect the proteolytic activity of live trichomonads, 2.5 × 10 5 parasites were incubated for 20 min at 4 ∘ C with 1 mM TLCK, 0.2 mM leupeptin, 0.18 mM E-64, or 50 M Aza-Peptidyl Michael Acceptor [43]. e activity was measured with two CPs substrates (Z-Phe-Arg-AMC for papains; and Cbz-Ala-Ala-AAsn-AMC for legumains). Release of free 7amino-4-methylcoumarin (AMC) was measured by emission at excitation wavelengths of 355 and 460 nm, respectively, in a luminometer (BioTek) using a Gen5 2.0 Data Analysis Soware. e linear regression of the substrate hydrolysis curves was used to calculate initial velocities. e experiment was repeated at least three independent times with similar results. e viability of trichomonads was assessed by trypan blue exclusion throughout the assay.   a None corresponds to the control parasites without treatment with CP inhibitors. e number of parasites attached to the HeLa cell monolayer was taken as 100% adherence for comparative purpose. b MA corresponds to the Aza-Peptidyl Michael Acceptor, a legumain-speci�c inhibitor. ese differences were statistically signi�cant with a P < 0.001 (Figure 8(B)).

Identi�cation of the TvLEGU-1 Proteinase.
To identify the TvLEGU-1 CP in T. vaginalis proteinase-rich extracts polyclonal antibodies were produced against the recombinant TvLEGU-1 protein (anti-TvLEGU-1r) previously cloned and expressed in Escherichia coli [15]. By WB assays, the anti-TvLEGU-1r antibody reacted with the recombinant TvLEGU-1 protein used as antigen (Figure 1(A)). It also recognized a 30-kDa band in T. vaginalis protease-rich trichomonad extracts and two bands of 30-and 20-kDa in T. vaginalis total protein extracts of parasites grown in normal iron conditions. A light band of 60-kDa was also observed (Figure 1(B)). As expected, the PI serum used as a negative control had no reaction ( Figure 1).
To con�rm the identity of the protein spots corresponding to TvLEGU-1 in a protease-rich trichomonad extract (dubbed "active degradome") [15], parasites grown in normal iron conditions were analyzed by 2DE and WB assays. e anti-TvLEGU-1r antibody recognized three spots in the 30-kDa region with pI ∼6.3, ∼6.5, and ∼6.7 (Figure 1(C)). ese protein spots were identi�ed by MALDI-TOF MS analysis as TvLEGU-1 proteins; ten of the 25 peptides obtained by tryptic digestion of the three spots had identical masses to TvLEGU-1 peptides. e MS analysis showed that protein spots 1, 2, and 3 were identi�ed with MASCOT scores of 72, 85, and 150, respectively, and sequence coverage of 14, 33, and 31%, respectively; reinforcing their identi�cation as part of the TvLEGU-1 protein ( Figure 1; Table 1; Supplementary Figure 1S).

TvLEGU-1 Is Phosphorylated and Upregulated by Iron.
To investigate whether phosphorylation could be an explanation for the three protein spots with the same size but distinct pI identi�ed as TvLEGU-1 in 2DE WB assays, protease-rich extracts from parasites grown in normal iron concentrations were analyzed by 2DE WB assays with antiphospho-Ser, -r, and -Tyr antibodies. Figure 1(D) shows that while none of the protein spots have phosphorylation in Ser residues, the protein spot 1 did not show either in r or Tyr residues, the protein spots 2 and 3 showed in r, and only spot 3 in Tyr residues. Interestingly, the intensity of the protein spot 3 was greater with the anti-Tyr than with the anti-r antibody (Figures 1(D) and 1(E)). us, the three protein spots of TvLEGU-1 are isoforms with distinct type and degree of phosphorylation, as was predicted in its amino acid sequence [38].
To check the effect of iron at the protein expression level of TvLEGU-1, protease-rich extracts from parasites grown in iron-rich and iron-depleted conditions were analyzed by silver-stained 2DE and by 2DE WB. A densitometric analysis was also performed ( Table 2). Figure 2 shows that the three protein spots in the 30-kDa region are present in both iron conditions but protein spot 1 with less intensity in iron-depleted than in iron-rich parasites (Figure 2(A)). ese differences are well appreciated in the densitometric analysis ( Figure 2(B), Table 2). Furthermore, the surface localization of TvLEGU-1 in nonpermeabilized parasites grown in different iron concentrations was explored. e data show that TvLEGU-1 surface localization is positively modulated by iron (Figure 3). Only a light surface localization of TvLEGU-1 was observed in iron-depleted parasites. ese data together with the 2D WB results suggest that the major differences of this protein due to the iron concentration could be observed at its surface localization. erefore, the rest of the experiments involving the anti-TvLEGU-1r antibody were performed with parasites grown in iron-rich conditions, except when indicated.

TvLEGU-1 Is Localized in the Cytoplasm and on the Surface of T. vaginalis.
To explore the total real distribution of TvLEGU-1 in trichomonads immuno�uorescence assays were performed using the anti-TvLEGU-1r antibody with �xed nonpermeabilized or permeabilized parasites grown in iron-rich condition and analyzed by confocal microscopy. Figure 4 shows that �uorescence with the anti-TvLEGU-1r antibody (in green) was detected on the surface, colocalizing with the membrane marker (Dil, in red; panels e-h) and in the cytoplasm of trichomonad parasites (panels i-l) as compared with the PI serum used as a negative control (panels a-d).
A very interesting labeling in the Golgi complex and vesicles that could be lysosomes was observed in the cytoplasmic localization of TvLEGU-1.
To explore the hypothesis that TvLEGU-1 is also a lysosomal CP, we performed colocalization assays using LysoTracker as a lysosomal marker in addition to the anti-TvLEGU-1r antibody. Figure 5 shows that indeed TvLEGU-1 colocalized (∼60%) with the lysosomal marker in ironrich parasites, as could be expected for legumain-like CPs ( Figure 5(A)). Moreover, immunogold localization assays con�rmed the cytoplasmic localization of TvLEGU-1 in vacuoles/lysosomes containing degrading material and in the Golgi complex ( Figure 5(B)), suggesting that this is an excreted/secreted proteinase. ese data suggest that TvLEGU-1 could have multiple functions that will depend on its cellular localization possible modulated by the iron concentrations.

TvLEGU-1 Binds to the Surface of HeLa Cells.
To determine whether TvLEGU-1 binds to the surface of HeLa cells, cell-binding and WB assays were performed with proteaserich extracts from parasites grown in iron-rich medium and �xed HeLa cells. WB assays showed that the anti-TvLEGU-1r antibody reacted with the trichomonad 30-kDa band that bound to �xed HeLa cells (Figure 6(A)), suggesting the presence of TvLEGU-1. is was con�rmed in a cellbinding assay using the recombinant TvLEGU-1 protein that also bound to the surface of �xed HeLa cells (Figure 6(B)), whereas the bovine serum albumin (BSA) used as a negative control did not bind as expected (Figure 6(B)). Additionally, TvLEGU-1r was recognized by the antinative CP30 antibody [7] in WB assays ( Figure 6(C)). Together, these data show that TvLEGU-1 is one of the CP30 proteinases that interact with the surface of HeLa cells [7]. Furthermore, to con�rm it, immuno�uorescence assays using �xed and live HeLa cells incubated with the TvLEGU-1r protein and the anti-TvLEGU-1r antibody were performed. Confocal microscopy images showed that indeed TvLEGU-1r bound to the surface of �xed and live HeLa cells, whereas HeLa cells directly incubated with the anti-TvLEGU-1r antibody used as a negative control had no reaction as expected (Figures 6(D) and 6(E)).

TvLEGU-1 Participates in T. vaginalis Cytoadherence.
To study the role of TvLEGU-1 in trichomonal adherence, we performed adherence inhibition assays over HeLa cell monolayers by preincubating [ 3 H]-thymidine-labeled iron-rich parasites with varied concentrations of the anti-TvLEGU-1r IgG or Fab fractions. Figure 7 shows that the anti-TvLEGU-1r antibody inhibited the levels of T. vaginalis adherence to HeLa cell monolayers in a concentration-dependent manner. A maximum inhibition of ∼45%, using 100 g/mL of IgGs or Fab fractions, was observed. IgGs or Fab fractions from PI serum used as a negative control did not affect trichomonal cytoadherence. ese results illustrate that TvLEGU-1 is a virulence factor that plays a role in cellular attachment as one of the 30-kDa CPs required for trichomonal adherence [7].

TvLEGU-1 Proteolytic Activity Is Necessary for T vaginalis Cytoadherence.
To determine whether TvLEGU-1 proteolytic activity was required for cellular attachment, live nonradioactive-labeled parasites [41] were treated with distinct CP inhibitors (TLCK, leupeptin, or E-64) or with increasing concentrations (0, 5, 10, and 50 g/mL) of a spe-ci�c legumain inhibitor the Aza-Peptidyl Michael Acceptor (Mu-Ala-Ala-AAsn-CH=CH-CON) [43] before interaction with live HeLa cell monolayers. Figures 8(A) and 8 show that the speci�c legumain inhibitor decreased the levels of T. vaginalis adherence to HeLa cell monolayers in a concentration-dependent manner up to ∼80%, whereas TLCK, leupeptin, and E-64 inhibited ∼60, ∼40, and ∼50%, respectively. e average number of parasites without treatment attached to HeLa cells per coverslip was higher than the number of parasites treated with inhibitors (Table 3), and these differences were statistically signi�cant (Figure 8(B)). ese results suggest that both legumain and papain-like CP proteolytic activities are necessary for trichomonal cytoadherence, especially the legumain-like activity. ese data are consistent with previous reports [4].
To check the effect of these inhibitors over the CP proteolytic activity of live parasites �uorescent substrates for  control (a, b, c, and d). Nonpermeabilized (NP; e, f, g, and h) and permeabilized (P; i, j, k, and l) parasites were incubated with the anti-TvLEGU-1r antibody (1 : 100 dilution). Anti-rabbit IgG-FITC (in green) was used as a secondary antibody (1 : 100 dilution) (a, e, and i). Parasite membranes were labeled with DIL (in red; b, f, and j). Nuclei were labeled with DAPI (in blue; c, g, and k). Merge (d, h, and l) in yellow indicates colocalization. Bars: 15 m (d, h, and l).
papain-like (Z-Phe-Arg-AMC) and legumain-like (Cbz-Ala-Ala-AAsn-AMC) CPs were used. Live untreated parasites used as control showed proteolytic activity for both substrates, which were taken as 100% activity. Parasites treated with the speci�c legumain inhibitor (Aza-Peptidyl Michael Acceptor) abolished the legumain-like proteolytic activity (Figure 8(C)) and has no effect on the papain-like proteolytic activity (Figure 8(D)). TLCK, a potent inhibitor of papainlike and legumain-like CPs, greatly reduced both proteolytic activities (∼80% and ∼95%, resp.) of treated parasites as   (Figures 8(C) and 8(D)). E-64 and leupeptin, potent inhibitors of papain-like CPs, greatly reduced the papainlike proteolytic activity (between ∼80 to ∼90%) of treated parasites (Figure 8(D)) and had a minimal or no effect on the legumain-like proteolytic activity (Figure 8(C)) of live parasites. erefore, both types of CP proteolytic activity are present in live parasites and are necessary for trichomonal adherence to host cells (Figure 8).

e TvLEGU-1 Proteinase Is Expressed during Infection and Is Present in Vaginal Secretions of Patients with
Trichomoniasis. To investigate the relevance of TvLEGU-1 during trichomonal infection, we analyzed vaginal washes from vaginitis patients with [Tv (+)] or without [Tv (−)] T. vaginalis (Table 4) for the presence of TvLEGU-1 by TCA-precipitation and WB assays using the anti-TvLEGU-1r antibody. Figure 9 shows that the anti-TvLEGU-1r antibody    detected the presence of protein bands that ranged from 35to ∼30-kDa in Tv (+) VWs (Figure 9(A)), but none in those Tv (−) with other vaginitis (Figure 9(B)), used as negative controls. ese data illustrate that the TvLEGU-1 is expressed and might be released during infection.
To con�rm this, in vitro secretion kinetic assays were performed (0, 15, 30, 60, and 90 min). Zymograms of the supernatants showed a 30-kDa band with proteolytic activity released through time (Figure 9(C)) from parasites that exhibited ∼95 to ∼99% viability measured by trypan blue exclusion assays. WB assays of the TCA-precipitated supernatants using the anti-TvLEGU-1r antibody con�rmed the presence of TvLEGU-1 among the released proteins and its amount increased through time. e anti--tubulin antibody used as a negative control gave no reaction as expected, suggesting that no signi�cant parasite lysis occurred.
Moreover, immuno�uorescence assays were also performed using the anti-TvLEGU-1r antibody with cells obtained from Tv (+) VWs and with live HeLa cells incubated with TvLEGU-1-containing parasite supernatants from the in vitro secretion assays (Figure 9(C)). e confocal microscopy images showed that endogenous TvLEGU-1 decorates the surface of live HeLa cells and cells obtained from Tv (+) VWs (Figure 9(D)). ese data show that TvLEGU-1 is part of the excretion/secretion products from live trichomonads that also bound to the surface of cells present in vaginal secretions (Figure 9).

Discussion
Numerous thiol-proteinases, including cathepsin L-and legumain-like proteinases, are encoded in the T. vaginalis genome. However, few have been characterized at either the molecular, biochemical, or functional level. Understanding the role of CPs, especially those of the legumain-like type in this parasite, is relevant, as they are known to be involved in numerous biological processes including the host-parasite interplay.
In this study, we identi�ed and characterized one of the ten legumain-like proteinases described in the dra of the T. vaginalis genome sequence [9], TvLEGU-1 [38], that showed multiple localizations, in the lysosomes and Golgi complex when in the cytoplasm and also at the parasite surface in the presence of iron (Figures 3-5). We con�rmed its positive iron regulation [21] at the protein level ( Figure 2) and surface localization (Figures 3 and 4). Moreover, we demonstrated its role in trichomonal adherence (Figures 7 and 8) and its presence in vaginal secretions during trichomonal infection (Figure 9). e three protein spots in the 30-kDa region, with distinct pI identi�ed by 2DE WB and M� as TvLEGU-1 ( Figure 1, Table 1), are in agreement with the protein spots identi�ed as TvLEGU-1 in the trichomonad active degradome [15]. ese represent isoforms with a different type and level of phosphorylation (Figure 1(D)) as previously suggested [38]. ese �ndings are consistent with  legumains from other organisms such as the Cs-legumain from Clonorchis sinensis with a similar pattern in 2DE gels [44], which is also phosphorylated.
Interestingly, the anti-TvLEGU-1r antibody showed a different recognition in the WB assays depending of the protein preparation. is could be due to the presence of fewer amount of protein in the protease-rich extract corresponding to the different protein bands that the antibody was unable to detect them. However, previous data showed that a 60-kDa protein spot was also observed in the 2DE gels from protease-rich extracts, and it was identi�ed as part of the TvLEGU-1 by MS [15]. At this point, we do not have an explanation to this high molecular size TvLEGU-1 protein. It is something that needs to be studied further. Although the 20-kDa protein was not observed nor identi�ed by MS in the protease-rich extract 2DE gels [15], we can speculate that this protein could represent a processing stage of TvLEGU-1 or a degradation product. Additionally, we could not discard that these proteins are the intracellular forms of TvLEGU-1. It is also something that deserves further investigation to help to understand the way this protein is processed, activated, localized, and showed different functions that could be modulated by the iron concentrations in the microenvironment.
TvLEGU-1 is one of the 30-kDa CPs localized on the surface of iron-rich trichomonad parasites (Figure 3) that also bound to the surface of HeLa cells (Figures 6 and 9). ese are properties consistent with the proteolytic activity necessary for trichomonal cytoadherence [4,5,7]. As expected, the antibodies against the recombinant TvLEGU-1r inhibited trichomonal adherence in a similar range (Figure 7) as the anti-CP30 antibody [7]. Additionally, T. vaginalis cytoadherence reduction by the speci�c legumain inhibitor supports that the TvLEGU-1 proteolytic activity may play a major role in the host-parasite interaction during trichomonal adherence ( Figure 8). erefore, this �nding is consistent and corroborates our previous observations that the CP30 proteolytic activity is necessary for trichomonal adherence [4,5,7] and a legumain-like CP, TvLEGU-1, is part of it. Additionally, TvLEGU-1 is positively regulated by iron at the protein level ( Figure 2) and surface localization (Figures 3  and 4), similar to the iron upregulation of the trichomonad adhesins [6,41,42,45]. is behavior could be expected for molecules that participate in the same trichomonal virulence property, cytoadherence [4,6,7,45]. Interestingly, the lack of surface recognition of the anti-TvLEGU-1r antibody in the Triton X-100-permeabilized parasites could be explained based on previous reports. ese demonstrate that nonionic detergents, such as Triton X-100, redistribute and solubilize phospholipid anchored proteins, even in previously �xed cells. Detergents such as Triton X-100 also have signi�cant adverse affects on the immunochemical analysis of gangliosides and GPI-anchored proteins [46,47]. Consistent with this explanation recent unpublished data suggest that TvLEGU-1, lacking transmembrane domains [38], is on the parasite membrane through a putative phospholipid anchor (work in progress).
e localization of the TvLEGU-1 in different compartments, particularly in lysosomes, suggests the typical role for TvLEGU-1, participating in the lysosomal degradation of food [48][49][50] or internal organelles during an autophagy process for remodelling of the cellular components [51] or in the parasite metabolism. is is in addition to its new role as a virulence factor involved in cytoadherence, as has been demonstrated in here, supporting previous data [6,41,42]. Moreover, the localization of TvLEGU-1 in the Golgi complex suggests that this CP undergoes part of its processing and maturation steps in this organelle as occurs with other legumains. Commonly, these proteinases are translated as a preproform, transferred through the Golgi complex as the proform of legumain with a molecular mass of 56-kDa, and localized in late endocytic compartments as the mature enzyme with a molecular mass of 46-kDa [52], as occurs in lysosomal cysteine proteinase [53]. e legumain CPs, which belong to the clan CD, are distinct from those of all other clans (with regards to their amino acid sequence, tertiary structure fold, protein substrates, effect of inhibitors, and biological functions). eir discovery has led to a reassessment of the relevance of roles played by CPs in parasitic protozoa [23,24]. Moreover, legumains have greater speci�city in their functions than the clan CA enzymes. us, the fact that legumain-like CPs have a very restricted type of substrates [48], as compared with the papain-like CPs, which are very promiscuous [23,24], suggests that this could be one of the major proteolytic activities necessary for trichomonal adherence to host cells [4,5,7]. We can speculate that this type of CPs such as TvLEGU-1 will unmask the surface of T. vaginalis by degrading the host proteins covering the adhesins, as previously suggested [4,5,54]. However, it is important to mention that as shown here the proteolytic activity of CPs of both clans, CA and CD, is necessary for trichomonal cytoadherence (Figure 8). One explanation is that both types of CPs could be necessary for directly processing the same protein targets such as the host proteins that cover the trichomonad surface [4,5,54]. Another possible explanation is through a similar mechanism as the one described for hemoglobinolysis of parasitic organisms such as apicomplexan and nematodes, where the participation of several proteinases is necessary in a cascade of proteolytic activation. A clan CD cysteine peptidase of the legumain type is implicated in the �rst step activating proteinases of clan CA directly involved in hemoglobin degradation [55][56][57].
erefore, we could speculate that in T. vaginalis an activation cascade could also be occurring to uncover the parasite surface, in which TvLEGU-1 will participate in the �rst step activating the papain-like CPs involved in host protein degradation. Once trichomonad CPs digest the proteinaceous cover on the T. vaginalis surface, by any of the two proposed pathways, the adhesins will then be free to interact with the host cell receptors for attachment [6,41,42]. us, further studies will be required to determine whether TvLEGU-1 directly participates in host protein degradation or in the �rst step of a putative activation pathway that will degrade particular host proteins as a prerequisite for cytoadherence. It also will be relevant to identify the substrate proteins for TvLEGU-1. Further work is needed to examine these questions.
It is relevant to mention that the recombinant protein TvLEGU-1r obtained in this study did not have proteolytic activity and could not be activated (data not shown) for biochemical and direct functional assays, using the same reported conditions [58]. is could be due to the fact that the tvlegu-1 gene was expressed in bacteria and lacked the sequence coding the �rst 10 aa residues of the N-terminal region, which may be required for its correct folding and activation. In spite of that, the recombinant TvLEGU-1 interacted with �xed HeLa cell, as the native protein did, suggesting that, in TvLEGU-1, the cell-binding and catalytic domains are different and could function separately. Moreover, identi�cation of TvLEGU-1 in vaginal washes from women with active trichomoniasis is consistent with the presence of the CP30 proteolytic activity [7] in trichomoniasis symptomatic patients [7,59], suggesting that, during infection, T. vaginalis releases several CPs, including TvLEGU-1, which is highly immunogenic [15]. Interestingly, we observed several bands (35-to ∼30-kDa) speci�cally recognized by the anti-TvLEGU-1r antibody in the Tv (+) VWs analyzed, suggesting the presence of different processing states of TvLEGU-1 during infection. Further work is needed to determine the processing steps of TvLEGU-1.

Conclusion
In this work, we have demonstrated that indeed CP proteolytic activity is necessary for trichomonal adherence to host epithelial cells that is consistent and corroborates our previous observations. One of these CPs is the TvLEGU-1, a legumain-like CP that is located in lysosomes, Golgi complex, and at the parasite surface in the presence of iron and shows different levels of phosphorylation. It will be interesting to identify the particular substrates for this CP, in addition to determine the phosphorylation or dephosphorylation effect on the proteolytic activity and its impact on cytoadherence. is CP was also found in vaginal secretions of patients with trichomoniasis, supporting its potential as biomarker. is work is the �rst paper that shows that a legumain-like CP plays a role in a pathogen cytoadherence.