The Glycolytic Enzyme Triosephosphate Isomerase of Trichomonas vaginalis Is a Surface-Associated Protein Induced by Glucose That Functions as a Laminin- and Fibronectin-Binding Protein

Abstract

Triosephosphate isomerase of Trichomonas vaginalis (TvTIM) is a 27-kDa cytoplasmic protein encoded by two genes, tvtim1 and tvtim2, that participates in glucose metabolism. TvTIM is also localized to the parasite surface. Thus, the goal of this study was to identify the novel functions of the surface-associated TvTIM in T. vaginalis and to assess the effect of glucose as an environmental factor that regulates its expression and localization. Reverse transcription-PCR (RT-PCR) showed that the tvtim genes were differentially expressed in response to glucose concentration. tvtim1 was overexpressed under glucose-restricted (GR) conditions, whereas tvtim2 was overexpressed under glucose-rich, or high-glucose (HG), conditions. Western blot and indirect immunofluorescence assays also showed that glucose positively affected the amount and surface localization of TvTIM in T. vaginalis Affinity ligand assays demonstrated that the recombinant TvTIM1 and TvTIM2 proteins bound to laminin (Lm) and fibronectin (Fn) but not to plasminogen. Moreover, higher levels of adherence to Lm and Fn were detected in parasites grown under HG conditions than in those grown under GR conditions. Furthermore, pretreatment of trichomonads with an anti-TvTIMr polyclonal antibody or pretreatment of Lm- or Fn-coated wells with both recombinant proteins (TvTIM1r and TvTIM2r) specifically reduced the binding of live parasites to Lm and Fn in a concentration-dependent manner. Moreover, T. vaginalis was exposed to different glucose concentrations during vaginal infection of women with trichomoniasis. Our data indicate that TvTIM is a surface-associated protein under HG conditions that mediates specific binding to Lm and Fn as a novel virulence factor of T. vaginalis.

FIG 1

Glucose promotes the growth of Trichomonas vaginalis and differentially modulates tvtim1 and tvtim2 gene expression and amount of TvTIM. (A1) Effects of GR (≤1 mM), NG (25 mM) and HG (50 mM) conditions on T. vaginalis growth. The initial number of parasites cultivated under the different glucose conditions was 2 × 10⁵ cells ml⁻¹. TYM (13 mM maltose) regular medium was used as a normal growth control. Cell density was determined at 6-h intervals during 24 h of growth at 37°C. (A2) Cell viability was monitored by trypan blue exclusion via hemocytometer counts after 24 h of incubation. All data (A1 and A2) are the means ± SDs of three independent experiments in duplicates. (B1) Semiquantitative RT-PCR using primers specific for the tvtim1 and tvtim2 genes (726 bp) and cDNA from parasites grown under GR (lane 1), NG (lane 2), and HG (lane 3) conditions. The β-tubulin gene (β-tub; 112 bp) was amplified as an internal control, using the cDNA from the parasites grown under different glucose conditions. RT-PCR was the negative (−) control, for which DNase-treated RNA without reverse transcriptase was used as a template. The sizes of the amplicons are given in base pairs. (B2) Densitometric analysis of the RT-PCR amplicons, as detected by ethidium bromide staining after electrophoresis in 1% agarose gels in panel B1 performed using Quantity One software (Bio-Rad). The bar graphs show the relative amounts of tvtim1 and tvtim2 transcripts normalized to the β-tubulin gene transcript level. The error bars indicate SDs, determined from three independent experiments. The asterisks (* and ***) show the significant differences (P < 0.05) of the amplicons obtained under the three glucose conditions for each tvtim gene as determined by ANOVA. (C1) SDS-PAGE and Coomassie brilliant blue (CBB) staining of 12% polyacrylamide gels were performed to assess the total protein extracts from parasites grown under GR (lane 1), NG (lane 2), and HG (lane 3) conditions. For WB assays, duplicated gels from panel C1 transferred onto NC membranes were incubated with different antibodies, including an anti-TvTIMr antibody (1:1,000 dilution) that recognized a 27-kDa band corresponding to the native TvTIM proteins. An anti-AP65 adhesin (anti-AP65) antibody (1:1,500 dilution) that recognized a 65-kDa band was used as a loading control, and preimmune rabbit serum (PI) (1:1,500 dilution) was used as a negative control. An anti-Entamoeba histolytica hexokinase (α-EhHKr) antibody (1:1,000 dilution) that recognized a 50-kDa band in T. vaginalis that corresponds to the T. vaginalis hexokinase (TvHK) was used as a control for glucose-induced modulation. An anti-enolase (α-TvENOr) antibody (1:1,500 dilution) that recognized a 48-kDa band was used as a control for the absence of glucose-induced effects. (C2) Densitometric analysis of the bands detected by WB (C1) performed with Quantity One software (Bio-Rad). The bar graphs show the relative amounts of TvTIM, TvHK, and TvENO proteins normalized to the level of the AP65 protein, which was used as a loading control. The error bars indicate the SDs of three independent experiments. The asterisks (* and ***) show significant differences (P < 0.05) among the protein bands detected by the antibodies under HG and NG conditions compared with GR conditions, as determined by ANOVA.

FIG 2

Glucose promotes the localization of TvTIM to the surface of T. vaginalis. (A) Indirect immunofluorescence and bright-field microscopy showed the cytoplasmic and surface localization and expression of TvTIM in parasites grown under HG (a to e) and GR (f to j) conditions for 24 h at 37°C. Paraformaldehyde-fixed parasites were incubated with a primary anti-TvTIMr antibody (1:50 dilution) followed by a FITC-conjugated secondary antibody (1:100 dilution). Parasites grown in HG were incubated with preimmune (PI) rabbit serum (1:50 dilution) followed by a FITC-conjugated secondary antibody as a negative control (k to o). (B) Cytoplasmic and surface localization of TvTIM in parasites grown in increasing glucose concentrations (≤1, 2.5, 5, 10, 25, and 50 mM) for 1 h at 37°C. Paraformaldehyde-fixed parasites were incubated with a primary anti-TvTIMr antibody (1:50 dilution) followed by a FITC-conjugated secondary antibody (1:100 dilution). In both panels A and B, the confocal microscopy (Zeiss) images show TvTIM labeled with FITC (in green), nuclei labeled with DAPI (in blue), and the parasite membrane labeled with DIL (1:2,000 dilution, in red). The merged images show colocalization between the TvTIM protein and the parasite surface in yellow. Bar size: 10 μm. These experiments were performed three independent times, with similar results.

FIG 3

TvTIM shows two possible unconventional trafficking routes to the plasma membrane under HG conditions. (A) Immunogold labeling of parasites grown under HG conditions using a primary anti-TvTIMr antibody at a 1:30 dilution and a secondary antibody conjugated to 10- or 20-nm gold particles. The samples were analyzed by TEM. The TEM images show parasites directly incubated with a secondary antibody conjugated to gold particles as a negative control (a). In general, the TvTIM gold labeling shows localization free in the cytoplasm (C), in vacuoles (V), and on the parasite surface, associated with the plasma membrane (PM). Additionally, TEM images show two possible unconventional trafficking routes of TvTIM from the cytoplasm to the plasma membrane (b). Panels c to e show the trafficking of TvTIM to the plasma membrane through an unconventional pathway independent of vesicles. Panels f to h show the trafficking of TvTIM to the plasma membrane through an unconventional pathway dependent on vesicles or multivesicular body (MVB)-like structures. The arrows point to the gold particles. (B) The TEM images show TvTIM labeling localized on the inner or outer face of the vesicle membranes (a). TvTIM labeling also localized in vesicles with or without cytoplasmic content and near the plasma membrane (b and c). TvTIM labeling also localized in vesicles that were in the process of fusing with the plasma membrane (d). Bar size: 500 nm. The confocal microscopy images show the green TvTIM label in compartments similar to cytoplasmic vesicles, vesicles near the plasma membrane, and vesicles in the process of fusion with the plasma membrane (g to i). The images show TvTIM labeled with FITC (g; in green), nuclei labeled with DAPI (e; in blue), and the parasite membrane labeled with DIL (f; in red). The merged images show colocalization of the TvTIM protein with the parasite surface in yellow (h and i). The arrows point to vesicular structures. The framed region in panel h is magnified in panel i. Bar size: 20 μm.

FIG 4

Binding of recombinant TvTIM proteins to ECM components. (A) Electrophoretic analysis of laminin (Lm), fibronectin (Fn), and plasminogen (Plg) proteins. SDS-PAGE and Coomassie brilliant blue (CBB) staining of 10% polyacrylamide gels show the pattern of each protein (lanes 1 to 3). (B) For WB assays, duplicated gels from panel A transferred onto NC membranes were incubated with the specific primary antibodies anti-Lm (lane 4), anti-Fn (lane 5), and anti-Plg (lane 6), followed by incubation with the HRPO-conjugated secondary antibody. Duplicated NC membranes were directly incubated with secondary antibodies conjugated to HRPO, as a negative control for nonspecific signal (lanes 7 to 9). For far-Western blotting assays, the recombinant proteins (1 μg) TvTIM1r (C), TvTIM2r (D), TvENOr (E), and ppTvCP4r (F) were subjected to SDS-PAGE, blotted onto NC membranes, and incubated with (30 μg ml⁻¹) Lm (C and D, lane 7; E and F, lane 6), Fn (C and D, lane 8; E and F, lane 7), Plg (C and D, lane 9; E and F, lane 8), or ppTvCP4r (C and D, lane 10). The specific protein-protein interactions were detected by WB using the appropriate antibody, anti-Lm (C and D, lanes 3 and 7; E and F, lanes 3 and 6), anti-Fn (C and D, lanes 4 and 8; E and F, lanes 4 and 7), anti-Plg (C and D, lanes 5 and 9; E and F, lanes 5 and 8), or anti-ppTvCP4r (C and D, lanes 6 and 10) (1:500 dilution). The positive controls (+) were incubated with an anti-TvTIMr, anti-TvENOr, or anti-ppTvCP4r antibody, accordingly (lane 2). The controls stained with Coomassie brilliant blue are shown in lane 1. The recombinant TvENO protein (TVAG_32460) was used as a specific positive control for the interaction with Plg (E, lane 8). The negative controls were obtained by incubating the unrelated recombinant protein ppTvCP4r (F) with Lm (lane 6), Fn (lane 7), and Plg (lane 8) and the corresponding antibodies. Additional negative controls (C and D, lanes 3 to 6; E and F, lanes 3 to 5) were obtained by directly incubating the recombinant proteins tested with the different antibodies used. The protein bands detected by WB were visualized using Quantity One software (Bio-Rad). These experiments were performed at least three independent times, with similar results.

FIG 5

Dot blot-binding assay confirmed the interaction of Lm and Fn with the recombinant TvTIM proteins of T. vaginalis. (A and B) For dot blot binding assays, increasing amounts (0 to 4 μg) of TvTIM1r and TvTIM2r were immobilized onto NC membranes and incubated with Lm, Fn, or Plg (30 μg ml⁻¹). The binding of Lm, Fn, and Plg to TvTIM1r or TvTIM2r was detected by WB using an anti-Lm, anti-Fn, or anti-Plg antibody, respectively, followed by a peroxidase-conjugated secondary antibody. The intensities of the black spots indicate positive interactions in relation to the amount of immobilized protein. (C) Different amounts (0 to 4 μg) of TvENOr (positive control), TvTIM1r, and TvTIM2r were immobilized onto NC membranes and incubated with 30 μg ml⁻¹ of Plg for TvENOr or 60 μg ml⁻¹ of Plg for TvTIM1r and TvTIM2r. The specific protein-protein interactions were detected by WB using an anti-Plg antibody, and positive interactions are shown as black spots. (D) Different amounts (0 to 4 μg) of TvCP4r (an unrelated recombinant protein) were incubated with Lm, Fn, and Plg (30 μg ml⁻¹). The specific protein-protein interactions were detected by WB using the corresponding antibodies. The absence of black spots indicates that TvCP4r did not bind Lm, Fn, or Plg. This assay was used as a negative control for dot blot binding. As an additional positive control, different amounts (0 to 4 μg) of TvTIM1r (A), TvTIM2r (B), and TvCP4r (D) were directly incubated with anti-TvTIMr or anti-TvCP4r antibody, respectively, followed by a peroxidase-conjugated secondary antibody. The bar graphs show densitometric analysis results for each spot as volume intensity per square millimeter, as determined using Quantity One software (Bio-Rad). The error bars indicate the SDs determined from three independent experiments. For the dot blot positive control, increasing concentrations of TvTIM1r and TvTIM2r were incubated with an anti-TvTIMr antibody, followed by a peroxidase-conjugated secondary antibody. The asterisks (** and ***) show significant differences (P < 0.05) among the spots detected by WB, as determined by ANOVA.

FIG 6

Glucose promotes the adherence of T. vaginalis to Lm and Fn. For adherence assays, live parasites (5 × 10⁵ cells per well) grown under GR, NG, or HG conditions were labeled with CellTracker Blue (CTB), added to microtiter wells coated with Lm or Fn (2 μg), and incubated for 30 min at 37°C. The light microscopy images show the attachment of the parasites to Lm (A) and Fn (B) under the different glucose conditions. The number of bound parasites was estimated indirectly by measuring the fluorescence at 466 nm using a Gemini EM spectrofluorometer. The bar graphs show the relative fluorescence units (RFUs) of the parasites attached to Lm (C) or Fn (D) under each glucose condition. The negative specificity controls were obtained using parasites attached to BSA-coated wells under each glucose condition. The error bars indicate the SDs determined from three independent experiments performed in triplicate wells for each condition. The asterisks (** and ***) show the significant differences (P < 0.05) among the attached parasites under the three glucose conditions, as determined by ANOVA.

FIG 7

TvTIM on the parasite surface mediates the specific binding of T. vaginalis to Lm and Fn, under HG conditions. (A and B) For inhibition assays performed using an anti-TvTIMr antibody, parasites grown in HG (5 × 10⁵ cell ml⁻¹) that had been previously labeled with CellTracker Blue (CTB) were incubated with increasing concentrations (0 to 400 μg ml⁻¹) of anti-TvTIMr, PI serum, or anti-ppTvCP4r IgGs before interaction with immobilized Lm (A) or Fn (B) for 30 min at 37°C. The fluorescence emission at 466 nm from the attached parasites was quantified using a Gemini EM spectrofluorometer. The direct binding of untreated CTB-labeled parasites to Lm- or Fn-coated wells (with absolute values of 171.74 and 209.48 RFU, respectively) was taken as 100% binding. IgGs from the PI serum and an unrelated anti-ppTvCP4r antibody were used as negative controls. The bar graphs show the mean percentages of adherence obtained from three independent experiments using triplicate samples. The error bars indicate the SDs, and the asterisks (* and ***) show the significant differences (P < 0.05) as determined by ANOVA. (C and D) For competition assays using the recombinant TvTIM proteins as competitors, the Lm- or Fn-coated wells were preincubated with increasing amounts (0 to 1.6 μg per well) of TvTIM1r, TvTIM2r, or ppTvCP4r as a negative control, before interaction for 30 min at 37°C with live parasites grown under HG conditions and labeled with CTB. Finally, the fluorescence emission from the attached parasites at 466 nm was quantified using a Gemini EM spectrofluorometer. The direct binding of CTB-labeled parasites to Lm- or Fn-coated wells in the absence of competitors (with absolute values of 171.74 and 209.48 RFU, respectively) was taken as 100% binding. Each point represents the mean percentage of attached parasites in the presence of a competitor (TvTIM1r, TvTIM2r, or ppTvCP4r). The error bars indicate the SDs from three independent experiments performed in triplicate wells for each condition.