Proteomic analysis of Trypanosoma cruzi secretome: characterization of two populations of extracellular vesicles and soluble proteins

Abstract

Microorganisms use specialized systems to export virulence factors into host cells. Secretion of effector proteins into the extracellular environment has been described in Trypanosoma cruzi; however, a comprehensive proteomic analysis of the secretome and the secretion mechanisms involved remain elusive. Here, we present evidence that T. cruzi releases proteins associated with vesicles that are formed by at least two different mechanisms. Transmission electron microscopy showed larger vesicles budding from the plasma membrane of noninfective epimastigotes and infective metacyclic trypomastigotes, as well as smaller vesicles within the flagellar pocket of both forms. Parasite conditioned culture supernatant was fractionated and characterized by morphological, immunochemical, and proteomic analyses. Three fractions were obtained by differential ultracentrifugation: the first enriched in larger vesicles resembling ectosomes; the second enriched in smaller vesicles resembling exosomes; and a third fraction enriched in soluble proteins not associated with extracellular vesicles. Label-free quantitative proteomic analysis revealed a rich collection of proteins involved in metabolism, signaling, nucleic acid binding, and parasite survival and virulence. These findings support the notion that T. cruzi uses different secretion pathways to excrete/secrete proteins. Moreover, our results suggest that metacyclic forms may use extracellular vesicles to deliver cargo into host cells.

Graphical Abstract

1. INTRODUCTION

Microorganisms such as bacteria, fungi, and parasites release bioactive molecules that modulate host-parasite interaction enabling pathogen survival and replication within the host. In eukaryotes, secreted proteins bear an N-terminal signal sequence that targets protein to the endoplasmic reticulum (ER)/Golgi-dependent classical secretion pathway. On the other hand, secreted proteins that do not bear an N-terminal signal sequence use non-classical secretory pathways and different routes of transportation to outside the plasma membrane.¹ Different mechanisms of intracellular trafficking have been proposed to mediate unconventional secretion, and one alternative route is through association with vesicles. There are currently three known mechanisms by which membrane vesicles are released into the extracellular environment: (i) exocytic fusion of multivesicular bodies (MVBs), resulting in exosomes; (ii) budding of vesicles directly from the plasma membrane, resulting in ectosomes or plasma membrane-derived vesicles or microparticles, and (iii) cell death, leading to apoptotic blebs.² Exosomes are small vesicles usually ranging around 20–100 nm in diameter, with a cup-shaped morphology, whereas ectosomes or plasma membrane-derived vesicles or microparticles are large vesicles ranging around 100–1000 nm in diameter, with a heterogeneous morphology.^{3, 4}

Trypanosoma cruzi is the causative agent of Chagas disease, a neglected tropical infection endemic in Latin America. Chagas disease affects almost 8 million people,⁵ with an increasing number of cases being reported in non-endemic regions, including the United States and Europe.^{6, 7} T. cruzi has four developmental stages in its life cycle; two stages are present in the hematophagous insect vector (a triatomine or reduviid bug, popularly known as the kissing bug) and two in the mammalian host. Epimastigotes are noninfective replicative forms present in the vector’s gut that later differentiate into metacyclic trypomastigotes, which are non-replicative infective forms transmitted to mammalians by the insect’s excrements during its bloodmeal or by the oral route.⁸ Metacyclic forms invade host cells by triggering signaling cascades, with a transient increase in host-cell intracellular Ca²⁺ concentration and actin cytoskeleton disruption,⁹ leading to the recruitment of lysosomes to the site of entry,¹⁰ an event required for the biogenesis of parasitophorous vacuole.^{11, 12} A few days later, metacyclic forms promote vacuole disruption and escape to the cell cytoplasm, where they transform into amastigote forms that replicate inside the cell. After a few rounds of replication, intracellular amastigotes transform into infective trypomastigotes that are released into the bloodstream upon cell disruption.

It has been reported that mammalian tissue culture cell-derived trypomastigotes (TCTs) release membranous vesicles carrying virulence factors such as members of the trans-sialidase glycoprotein multigene family,^13–16 and that these extracellular vesicles were involved in Chagas disease pathogenesis by increasing heart parasitism and inflammation.¹⁷ Also, a factor secreted by TCTs was shown to inhibit mitogen-activated protein (MAP) kinase activation and to modify host-cell gene expression.¹⁸ In addition, it has been shown that exosome-like vesicles released by Leishmania major and L. donovani inside macrophages were able to modify host cytokine profile.¹⁹ Even though several studies identified and characterized specific effector proteins excreted/secreted by T. cruzi,^{13–16, 20–26} a comprehensive proteomic analysis of T. cruzi secretome and the mechanism(s) by which this parasite releases these bioactive molecules remain elusive.

Here, we present evidence suggesting that both noninfective epimastigote and infective metacyclic trypomastigote forms of T. cruzi release proteins to the extracellular milieu in a manner associated or not with vesicles. Parasites use at least two different mechanisms for vesicle formation and release: budding from the plasma membrane and exocytose of MVBs’ content. We demonstrate by a combination of methods, such as morphological, immunochemical, and label-free quantitative proteomic analysis, that distinct proteins are enriched in each vesicle population, revealing a rich collection of excreted/secreted molecules involved in diverse cellular processes.

2. EXPERIMENTAL SECTION

2.1. Reagents

Otherwise indicated, reagents used in this study were of molecular biology-, HPLC-, or spectrometry/proteomics-grade from Sigma-Aldrich (St. Louis, MO).

2.2. Ethics Statement

This study was carried out in accordance with recommendations in the Guide for Care and Use of Laboratory Animals of the National Institutes of Health. The protocol was approved by the Committee on Animal Experiment Ethics of the Federal University of Sao Paulo (UNIFESP) (Protocol Number: CEP09555–07).

2.3. Parasite cultures and conditioned medium isolation and fractionation

T. cruzi Dm28c²⁷ clone was maintained by cyclic passage in mice and axenic cultures in liver-infusion tryptose (LIT) medium containing 10% fetal bovine serum (FBS) at 28°C.²⁸ Epimastigote forms were transfected with pGFP plasmid described by Bayer-Santos et al.,²⁹ which comprises the pTEX vector carrying the green fluorescent protein (GFP). To obtain the parasite conditioned medium (CM), epimastigotes were harvested from exponentially growing cultures, washed in Dulbecco’s modified Eagle medium (DMEM) without FBS and incubated in the same medium at a concentration of 1 × 10⁸ cells/mL for 6 h at 28 °C. Metacyclic trypomastigote forms were obtained by induction of metacyclogenesis according to Contreras et al.,²⁷ purified by anion-exchange chromatography as described by Yoshida³⁰, and incubated in TAU3AAG medium²⁷ at a concentration of 1 × 10⁸ cells/mL for 6 h at 28 °C. Following 6-h incubation, parasites were removed by centrifugation at 3,000 g for 10 min at 4 °C. The cell-free supernatant was filtered in 0.45-μm syringe filter (Millipore), transferred to 13.2-mL polyallomer tube, and centrifuged at 100,000 g for 2 h at 4 °C to obtain the first pellet, enriched in larger extracellular vesicles (V2). The resulting supernatant was transferred to another polyallomer tube and then centrifuged at 100,000 g for 16 h at 4 °C, to obtain the second pellet, enriched in smaller vesicles (V16) and the vesicle-free supernatant (VF) (see Figure 2A). All ultracentrifugation steps were carried out using a TH-641 titanium swinging-bucket rotor (Thermo Fisher Scientific) in a WX80 ultracentrifuge (Thermo Fisher Scientific).

Figure 2.

Fractionation protocol and morphological characterization of vesicle and vesicle-free fractions. (A) Schematic representation of the protocol for isolation and fractionation of parasite conditioned medium to obtain the vesicle fractions (V2 and V16) and vesicle-free fraction (VF). Vesicle fractions obtained by ultracentrifugation of epimastigotes (eV2 and eV16) and metacyclic forms (mV2 and mV16) conditioned medium had their particle size distribution determined by (B) nanoparticle tracking analysis and (C) TEM.

2.3. Parasite viability

After removal of CM, 3 × 10⁷ epimastigote or metacyclic trypomastigote cells were resuspended in 200 μL 10 mM phosphate-buffered saline (PBS), containing 20 μg/mL propidium iodide (PBS-PI), and analyzed in a Cytomics FC500 Beckman counter to assess parasite viability. Controls containing dead cells were obtained by incubating parasites with 50% ice-cold methanol in PBS for 5 min on ice, followed by washing with PBS, and staining with PI. After 6-h incubation, no significant death was detected in either epimastigote or metacyclic forms and over 98% of parasite cells were viable (data not shown).

2.4. Transmission electron microscopy (TEM)

Purified vesicles from epimastigote or metacyclic forms were absorbed onto formvar/carbon-coated copper grids, washed in deionized water, and stained with 1% uranyl acetate for 1 min. Grids were allowed to air dry and viewed on microscope. For ultrastructural analysis, parasites were fixed in 2% paraformaldehyde (PFA), 2.5% glutaraldehyde (GAA), in 100 mM phosphate buffer, pH 7.2, for 1 h at room temperature (RT). Samples were washed in PBS and post-fixed in 1% osmium tetroxide for 1 h. Samples were washed, dehydrated in a graded series of ethanol and embedded in Eponate 12 resin. Sections of 95 nm were stained with uranyl acetate and lead citrate, and observed on microscope. For immunolocalization, parasites were fixed in 4% PFA, 0.05% GAA, in 100 mM phosphate buffer, pH 7.2, for 1 h at 4 °C. Samples were embedded in 10% gelatin and infiltrated overnight with 2.3 M sucrose, 20% polyvinylpyrrolidone, in PIPES/MgCl₂ at 4 °C. Samples were frozen in liquid nitrogen and sectioned with a cryo-ultramicrotome. Sections of 50 nm were blocked with 5% FBS, 5% normal goat serum, and incubated with mouse monoclonal antibody (mAb) 10D8 (anti-GP35/50),³¹ followed by 18-nm colloidal gold-conjugated antibody (Jackson ImmunoResearch Laboratories, Inc., West Grove, PA). Sections were washed in 0.1 M PIPES buffer, pH 7.2, and stained with 0.3% uranyl acetate, 2% methyl cellulose. Parallel controls omitting the primary antibody were consistently negative at the concentration of colloidal gold-conjugated antibodies used (not shown). All samples were viewed on a JEOL 1200EX transmission electron microscope at an accelerating voltage of 80 kV. All TEM experiments were performed at the Molecular Microbiology Imaging Facility of Washington University School of Medicine (St. Louis, MO).

2.5. Nanoparticle tracking analysis

The size of T. cruzi extracellular vesicles from epimastigote (eV2 and eV16) and metacyclic forms (mV2 and mV16) was determined by nanoparticle tracking analysis (NTA), by measuring the rate of Brownian motion to particle size, using a NanoSight LM14 system (NanoSight, Wiltshire, UK). Three replicates of diluted vesicles aliquots (500 μL in PBS) were injected into the machine’s specimen chamber and vesicles were tracked and measured for 60 s at RT. Data were analyzed with NTA software (version 2.2).

2.6. Polyacrylamide gel electrophoresis and western blotting

Two micrograms of protein from each CM fraction from epimastigote (eV2, eV16, and eVF) and metacyclic forms (mV2, mV16, and mVF) were resolved in 12% SDS-PAGE and transferred onto nitrocellulose membranes. Membranes were processed as described elsewhere.³² T. cruzi surface glycoproteins GP82, GP35/50, and flagellar calcium-binding protein (FCaBP) were detected by mAbs 3F6, 10D8, and 25, respectively.^{31, 33} GFP was detected by rabbit anti-GFP polyclonal antibody (Invitrogen, Life Technologies, Grand Island, NY). It is worth pointing out that GP82 glycoprotein is expressed only in metacyclic forms, whereas GP35/50 (mucin-like glycoproteins) and FCaBP are expressed in both epimastigote and metacyclic forms.³⁴

2.7. Immunofluorescence analysis

Parasites collected after 6 h of incubation in serum-free medium were washed with PBS, fixed with 4% PFA for 10 min, placed on glass slides pre-treated with poly-lysine and incubated with 50 mM NH₄Cl. After wash, cells were blocked and permeabilized with 3% bovine serum albumin (BSA), 0.5% saponin in PBS for 30 min at RT, and incubated with primary antibodies (mAb 10D8) diluted in 1% BSA, 0.1% saponin for 1 h at RT. Cells were incubated with secondary antibodies conjugated to Alexa Fluor 488 (Invitrogen) containing 10 mM 4’,6-diamidine-2’-phenylindole, dihydrochloride (DAPI, Invitrogen, Carlsbad, CA), for 1 h at RT, protected from light. After washing cells by centrifugation (3,000 g, 10 min, 4 °C) in PBS-1% BSA, coverslips were mounted on slides using anti-fading agent (Vectashield, Vector Laboratories, Burlingame, CA) and sealed with nail polish. Images were acquired with a Leica DMI 6000 B epifluorescence microscope. To visualize vesicles released by metacyclic trypomastigotes in contact with HeLa cells, parasites were incubated for 10 min on ice with mAb 3F6, washed with PBS, diluted in DMEM-10% FBS and incubated for 2 h with HeLa cells at 10:1 multiplicity of infection (MOI). Cells were washed to eliminate parasites not attached or internalized, fixed, and processed as described above, except for incubation with secondary antibody conjugated to Alexa Fluor 594 (Invitrogen) and phalloidin-Alexa Fluor 488 (Invitrogen). Images were acquired with a Zeiss LSM 700 confocal microscope and processed with ZEN 2009 software (Zeiss, NY).

2.8. Sucrose-density gradient

Vesicles were separated by flotation in a linear sucrose-density gradient as previously described.¹⁹ Briefly, pellets containing V2 or V16 vesicles (30 μg of protein, as measured by the BCA kit, Pierce) from either epimastigote or metacyclic form were resuspended in 2.5 M sucrose and overlaid with a step-wise gradient (2.5–0.25 M sucrose, 20 mM HEPES, pH 7.4) followed by centrifugation at 200,000 g for 18 h at 4 °C. Fractions of 1 mL were collected from the bottom of the tube and density was determined in a refractometer. Each fraction was diluted with 10 mL PBS and centrifuged at 200,000 g for 2 h. Pellets were solubilized in reducing Laemmli sample buffer³⁵ and analyzed by 10% SDS-PAGE and western blotting.

2.9. Protein digestion and 2D LC-MS/MS analysis

In this study we analyzed six biological samples: three from epimastigotes (i.e., eV2, eV16, and eVF) and three from metacyclic trypomastigotes (i.e., mV2, mV16, and mVF), each in technical triplicate, thus totalizing 18 samples that were subjected to 2D LC-MS/MS. To obtain sufficient amount of protein for proteomic analysis, we used two distinct CM preparations that were mixed and filter-concentrated in a 3-kDa MWCO Centricon (Millipore). Twenty micrograms of protein from each fraction (V2, V16, and VF) were solubilized with RapiGest (Waters) and trypsin-digested as specified by Turiak et al.³⁶ Resulting tryptic peptides were desalted using reverse-phase microcolumns as described by Jurado et al.,³⁷ prior to mass spectrometry analysis. Peptide mixtures were separated by two-dimensional liquid chromatography and analyzed on-line by electrospray-ionization tandem mass spectrometry (ESI-MS/MS) using an ESI-linear ion-trap-MS (LTQ XL, Thermo Fisher Scientific, San Jose, CA), as described by Nakayasu et al.³⁸ Briefly, peptides were loaded into the SCX trap column and eluted using the autosampler by injecting increasing salt concentrations (0, 25, 50, 100, 200, and 500 mM NaCl in 5% ACN/0.5% FA, pH 2.5). Eluting peptides were automatically loaded into the C18-trap column, which was then washed with 2% ACN/0.1% FA (solvent A) at 2.5 μL/min flow rate for 130 min. In the second dimension, the peptides were separated in the capillary C18 column at 300 nL/min flow rate using a linear gradient from 5 to 40% solvent B (80% ACN/0.1% FA), over 100 min, followed by 10 min wash with 100% solvent B and 20 min with 5% solvent B. The nanoelectrospray was set to 1.8 kV, and the ion trap was set for a maximum injection time of 100 msec for full scan and 150 msec for MS/MS scan. The 10 most abundant ions were selected for collision-induced dissociation (CID) with an isolation width of 3.0 amu and normalized collision energy of 35%. Dynamic exclusion was set for fragmenting each ion twice and then excluding it for 1 min.

2.10. Data analysis

The 2D LC-MS/MS analysis of each sample from epimastigote or metacyclic form gave rise to eight raw MS/MS data files, thus the 18 samples analyzed in this study generated a total of 144 raw MS/MS data files and 27,728 MS/MS spectra. These spectra were searched using Sequest (version v.27; Thermo Fisher Scientific) and X! Tandem (version 2007.01.01.2; http://www.thegpm.org/tandem/) algorithms against sequences from Trypanosoma ssp., BSA, human keratin, and porcine trypsin (downloaded from GenBank on October 10th, 2011). Sequences (total of 99,536) were used in forward and reverse orientations to calculate the false-discovery rate (FDR). Parameters for database search were as follows: trypsin cleavage at both termini and two missed cleavage allowed; 1 Da for peptide-mass tolerance; 1 Da for fragment-mass tolerance; cysteine carbamidomethylation and methionine oxidation as fixed and variable modifications, respectively. Scaffold platform (version 3.4.3; Proteome Software, Portland, OR)³⁹ was used to validate peptide and protein identifications, which were accepted if greater than 95% for peptides and 95% for proteins containing at least 2 peptides, according to the PeptideProphet⁴⁰ and ProteinProphet⁴¹ algorithms. Using these parameters, the FDR was estimated to be <2.5%. Protein sequences containing similar peptides (e.g., isoforms) that could not be differentiated were grouped and referred to as protein groups. These groups were assigned by numbers and ordered by Scaffold based on the spectrum count for each protein in each row at the 50% identification probability. However, if two proteins had the same count Scaffold ordered them based on the protein probability, and if the protein probability were the same, the program ordered proteins alphabetically based on their accession numbers.

Protein quantification was performed using a label-free approach, where relative protein quantification was achieved by comparing the number of identified MS/MS spectra from the same protein in each of the multiple datasets.⁴² Scaffold was also used to calculate normalized spectral count values across all samples. Protein abundance analysis was performed by QSPEC⁴³ algorithm using normalized spectral count data, and multiple testing corrections were performed by the FDR estimation procedure implemented in the new version of QSPEC (QPROT). Proteins with significant differential expression were selected with the criterion of FDR ≤ 0.1. Principal component analysis (PCA) was performed using normalized spectral count values, which were centered prior analysis, and run using Pearson’s correlation matrix. Protein sequences were submitted to Gene Ontology (GO) analysis using Blast2GO.⁴⁴ To identify proteins secreted by classical or non-classical pathways, sequences were submitted to SignalP 4.0⁴⁵ and to SecretomeP 2.0⁴⁶ algorithms, respectively.

3. RESULTS AND DISCUSSION

3.1. Origin and purification of T. cruzi extracellular vesicles

Earlier studies in T. cruzi detected TCTs releasing vesicles from the plasma membrane.^14,17 To investigate whether this phenomenon occurs in epimastigotes and metacyclic trypomastigotes, we incubated these parasites in chemically defined serum-free medium for 6 h, collected the cells, and analyzed them by TEM. We observed epimastigote and metacyclic forms releasing vesicles around 100–200 nm or larger from the plasma membrane (Figure 1, A–C and H–J), suggesting that the phenomenon is indeed common to most parasite forms. Nevertheless, it still remains to be determined whether intracellular and extracellular (axenic) amastigote forms also release similar vesicles. To examine whether there was another mechanism by which T. cruzi releases extracellular vesicles, we looked for MVBs close to the flagellar pocket and for exosome-like vesicles released into it. The flagellar pocket is the main site of endocytosis and exocytosis, the only region where the parasite’s body is not covered by subpellicular microtubules.⁴⁷ Figure 1 (D–E and K–L) shows the presence of MVBs containing small vesicles (<100 nm) inside, close to the flagellar pocket of epimastigote and metacyclic forms. The vesicles found in the flagellar pocket of epimastigotes (Figure 1, F–G) and metacyclic trypomastigotes (Figure 1, M–N) have size similar to those usually found inside MVBs, and are overall smaller than plasma membrane-derived vesicles, suggesting that MVBs fuse with the flagellar pocket membrane to release its content.

Figure 1.

Origin of T. cruzi extracellular vesicles. Upper panel, representative TEM images of epimastigotes releasing vesicles from the plasma membrane (A-C) and showing a MVB close to the flagellar pocket (D-E), which contains small vesicles inside (F-G) (arrow). Lower panel, representative TEM images of metacyclic forms releasing vesicles from the plasma membrane (H-J) and showing a MVB close to the flagellar pocket (K-L), which contains small vesicles inside (M-N) (arrow). Arrowhead indicates an empty large vesicle close to the flagellar pocket that could be resultant of a MVB exocytic fusion (M). Scale bar = 250 nm, except for N (bar = 100 nm). Nu: nucleus; K: kinetoplast; F: flagellum; FP: flagellar pocket; MVB: multivesicular body.

Based on these findings we developed an ultracentrifugation protocol to separate extracellular vesicle populations according to their size, and also soluble proteins that were not associated with vesicles. A schematic representation of the fractionation protocol is shown in Figure 2A. We separated vesicle populations in one fraction enriched in larger vesicles (V2), which was expected to be mainly plasma membrane-derived; and a second fraction enriched in smaller vesicles (V16), supposed to be mainly derived from the exocytic fusion of MVBs with the flagellar pocket membrane. To verify the size distribution of isolated vesicles, we used NTA and TEM with negative staining. Figure 2B shows the size distribution of vesicles obtained after 2 h (V2) and 16 h (V16) of ultracentrifugation of CM from epimastigote and metacyclic forms. Overlapping between V2 and V16 fractions was observed, in part due to the limited resolution of ultracentrifugation method. TEM images (Figure 2C) confirmed data obtained by NTA, showing larger vesicles in V2 fraction and smaller vesicles in V16 fraction, for both parasite forms. To further investigate the origin of these vesicles, immunochemical, and proteomic analyses were performed.

3.2. Characterization of vesicles and vesicle-free fractions isolated from parasite CM

Analysis of extracellular vesicle and vesicle-free fractions by SDS-PAGE showed that T. cruzi epimastigotes excrete/secrete numerous proteins (Figure 3A). The distinct pattern between whole-cell lysate (WCL) and excreted/secreted proteins of epimastigote-derived CM fractions (eV2, eV16, and eVF) indicates that proteins detected in CM were unlikely to be artifacts present due to cell lysis during processing. Distinct SDS-PAGE profiles between WCL and CM fractions (mV2, mV16, and mVF) were observed for metacyclic trypomastigotes as well (data not shown). In order to verify whether the fractionation protocol was efficient for separating extracellular vesicles of different origins, as detected by TEM, we looked for the presence of known surface and cytosolic proteins in epimastigote and metacyclic fractions. Western blots carrying these fractions were incubated with mAbs against surface glycoproteins GP82 and GP35/50,³¹ and FCaBP,³³ which are associated with the plasma membrane by post-translational modifications such as glycosylphosphatidylinositol (GPI) anchor (GP82 and GP35/50)^{48, 49} or myristoylation and palmitoylation (FCaBP)^{49, 50} (Figure 3B). We also used transfected epimastigotes expressing the GFP reporter in the cytoplasm²⁹ to analyze in which fraction GFP would be accumulated (Figure 3C).

Figure 3.

Characterization of V2 and V16 vesicle fractions reveals correspondence between vesicle origin and protein content. (A) Representative SDS-PAGE silver stained showing the abundance and diversity of proteins in conditioned medium fractions from epimastigotes. (B) Western blot of V2, V16, and VF fractions of epimastigotes (left) and metacyclic forms (right), reacted with anti-GP82, anti-GP35/50, and anti-FCaBP mAbs. (C) Schematic representation of pTEX vector, carrying the GFP gene (top) and an image of the transfected parasite expressing GFP in the cytoplasm (bottom). (D) Western blot showing the presence of GFP in V16 fraction (top); the same membrane was reacted with mAb anti-FCaBP (bottom), revealing differential protein enrichment. Two micrograms of protein from each fraction were loaded per lane. Sucrose-density gradient fractions of metacyclic trypomastigote (E) and epimastigote vesicles (F). Gradient fractions were analyzed by western blot using mAbs against GP82 and FCaBP, and anti-GFP polyclonal antibody.

Surface glycoproteins like GP82 and GP35/50 and an inner plasma membrane leaflet protein like FCaBP were found enriched in eV2 and mV2 fractions (Figure 3B), supporting our assumption that V2 was enriched in larger extracellular vesicles derived from plasma membrane. Additionally, we found the cytosolic marker GFP accumulated in eV16 fraction (Figure 3D), which we assumed to be enriched in smaller extracellular vesicles derived from MVBs, hence supporting the intracellular origin of V16 vesicles. To confirm the differential protein enrichment between extracellular vesicle fractions, the same blot was incubated with the mAb against FCaBP, which instead accumulated in V2 fraction (Figure 3D). It is interesting to note that GP82 and FCaBP were also found in the vesicle-free (VF) fraction of both parasite stages (Figure 3B), suggesting that these could be isoforms secreted without GPI-anchor (GP82) or acylation (FCaBP). On the other hand, GP82 could be also released into the extracellular medium by the endogenous activity of a phospholipase C located at the parasite’s plasma membrane.^{25, 51, 52}

A mixture of V2 and V16 fractions, which contain both larger and smaller extracellular vesicles, was analyzed by sucrose-density gradient (Figure 3E–F). We found that metacyclic trypomastigote-derived vesicles have a density around 1.14–1.17 g/mL, as determined by mAbs against GP82 and FCaBP, and epimastigote-derived vesicles have a density around 1.08–1.14 g/mL, as determined by mAb against FCaBP and with anti-GFP (Figure 3 E–F). It is interesting to note that in epimastigotes both vesicle populations (eV2 and eV16) have the same density based on presence of the markers for eV2 (FCaBP) and V16 (GFP) (Figure 3F). These results confirm that these proteins were indeed associated with vesicles rather than being pelleted as aggregates derived from cell lysis during centrifugation.

To further confirm whether proteins enriched in V2 fraction were associated with plasma membrane-derived vesicles, we analyzed the localization of GP35/50 in epimastigote and metacyclic forms following 6-h incubation in CM. In immunofluorescence assays, mAb 10D8 stained the parasite plasma membrane, giving a punctuate pattern (Figure 4A). The same pattern was observed on the surface of epimastigotes and metacyclic forms with mAbs anti-GP82 and anti-FCaBP (data not shown). Immunoelectron microscopy images showed the presence of GP35/50 (Figure 4B) in vesicles budding from the plasma membrane of epimastigotes, and also in extracellular vesicles, confirming that proteins found in V2 fraction were indeed associated with the parasite plasma-membrane. Together, these results indicate that we were able to separate two extracellular vesicle populations, one that is mainly derived from the plasma-membrane (i.e., V2) and another that is formed intracellularly via MVBs (i.e., V16).

Figure 4.

Immunolocalization of the surface protein GP35/50 (mucin-like glycoproteins) in plasma membrane-derived vesicles. (A) Representative immunofluorescence images of epimastigotes (left) and metacyclic forms (right) incubated with mAb against GP35/50 (green), showing labeling in plasma membrane and punctuate regions that resemble plasma membrane-derived vesicles (arrows). Nucleus and kinetoplast are shown in blue (DAPI). Scale bar = 5 μm. (B) Representative EM-immunostaining images of epimastigotes showing plasma membrane protrusions and extracellular vesicles labeled with mAb against GP35/50. Scale bar = 200 nm.

3.3. Identification and overall characteristics of proteins from vesicles and vesicle-free fractions

In order to obtain a broader view of T. cruzi excreted/secreted proteins, twenty micrograms of each CM fraction from epimastigote (eV2, eV16, and eVF) and metacyclic forms (mV2, mV16, and mVF) were analyzed in triplicate by 2D LC-MS/MS, using a nanoLC coupled to an ESI-linear-ion trap-MS. All MS/MS spectra were searched against a database assembled with Trypanosoma spp. sequences, and proteins identified with two or more peptides were validated. In total, we identified 345 proteins that were excreted/secreted by epimastigotes and 265 by metacyclic trypomastigotes. Of these proteins, 102 were identified exclusively in epimastigotes, 22 were detected only in metacyclic forms, and 243 proteins were common to both forms, comprising 367 distinct proteins. Proteins were classified and grouped into 16 categories based on information from the literature and Blast2GO annotation; hypothetical proteins were classified based on the presence of family domains. A large number of proteins were common to epimastigote and metacyclic forms, and nearly the same proportion was detected in each functional category (Table 1). A summary of identified proteins is shown in Table 2, and a complete list of identified proteins and respective peptides can be found in Tables S1 and S3.

Table 1.

Functional categories of T. cruzi excreted/secreted proteins.

Functional category	Epimastigote n (%)	Metacyclic trypomastigote n (%)
Host-parasite interaction	11 (3.2%)	10 (3.8%)
Signaling	22 (6.4%)	20 (7.5%)
Proteolysis, peptidolysis	27 (7.8%)	25 (9.4%)
Transcription and protein synthesis	29 (8.4%)	16 (6.0%)
Heat shock proteins and chaperones	21 (6.1%)	15 (5.7%)
Trafficking and membrane fusion	17 (4.9%)	13 (4.9%)
Transporters	7 (2.0%)	5 (1.9%)
Cell structure and motility	18 (5.2%)	17 (6.4%)
Oxidation reduction	24 (7.0%)	22 (8.3%)
Carbohydrate metabolism	24 (7.0%)	17 (6.4%)
Lipid metabolism	8 (2.3%)	7 (2.6%)
Protein and amino acid metabolism	34 (9.9%)	28 (10.6%)
Intermediate metabolism	9 (2.6%)	7 (2.7%)
Nucleotide metabolism	7 (2.0%)	6 (2.3%)
Nucleic acid binding	38 (11.0%)	19 (7.2%)
Others	49 (14.2%)	38 (14.3%)
Total	345 (100%)	265 (100%)

Table 2.

Summary of identified proteins excreted/secreted by epimastigotes and metacyclic trypomastigotes, their functional classification and determination of fraction enrichment according to label free quantification.

Group #a	Protein description	Accession #	Stageb	Fraction enrichmentc
				eV or eVFd	mV or mVFd
Host-parasite interaction
10	cyclophilin A (peptidyl-prolyl cis-trans isomerase)	gi|6224860	E/M	eVF↑	mVF↑
35	calreticulin	gi|322823951	E/M	eV2↑	mV↑
116	proline racemase copy B	gi|30349129	E	nd	nd
214	prolyl oligopeptidase	gi|322821000	E/M	eVF↑	nd
220	surface glycoprotein GP90	gi|21321389	M	Nd	nd
237	glycoprotein 82 kDa	gi| 154347429	E/M	eV2↑	mV↑
251	mucin-associated surface protein (MASP)	gi|322829180	E	nd	Nd
255	trans-sialidase	gi|71652185	E	eV↑	Nd
274	surface protease GP63	gi|322815808	E/M	nd	mV2↑
Signaling
15	IgE-dependent histamine-releasing factor	gi|322826089	E/M	eVF↑	nd
22	calcium-binding protein	gi|487896	E/M	eV2↑	mV↑
36	14-3-3 protein I-1	gi|257900395	E/M	eVF↑	mVF↑
37	calmodulin	gi| 10386	E/M	eVF↑	mVF↑
45	nucleoside diphosphate kinase B	gi|333361352	E/M	eV↑	nd
77	activated protein kinase C receptor	gi|70882943	E/M	nd	nd
113	phosphatidylinositol 3 kinase	gi|261332938	E/M	eV↑	mVF↑
187	mitogen activated protein kinase	gi|322827816	E	nd	nd
201	inositol 5-phosphatase 1	gi|322815366	E/M	eV16↑	nd
249	serine/threonine protein kinase	gi|322821342	E/M	nd	mV2↑
Proteolysis, peptidolysis
16	calpain-like cysteine peptidase	gi|322830271	E/M	eVF↑	mVF↑
39	peptidase M20/M25/M40	gi|322827793	E/M	eVF↑	nd
56	carboxypeptidase	gi|322815312	E/M	nd	mV2↑
87	calpain cysteine peptidase	gi|71402378	E/M	nd	nd
90	proteasome activator protein PA26	gi|70866914	E/M	nd	nd
104	cruzipain	gi|1136312	E	eV↑	nd
124	aminopeptidase P	gi|322818803	E/M	nd	mV2↑
Transcription and protein synthesis
7	elongation factor 1-alpha G5	gi|52424046	E/M	nd	nd
27	eukaryotic initiation factor 5a	gi|322815486	E/M	nd	mV2↑
41	40S ribosomal protein S3a	gi|71413035	E/M	eV↑	mV↑
82	60S acidic ribosomal protein P0	g i 1322819376	E/M	eV↑	nd
95	ribosomal P2-type protein	gi| 10634	E	eV↑	nd
126	basic transcription factor 3a	gi|70871 98	E/M	eVF↑	mV2↑
Heat shock proteins and chaperones
4	heat shock protein HSP70	gi| 162117	E/M	eVF↑	nd
9	85 kDa protein (heat shock protein 83)	gi|162111	E/M	eVF↑	mV↑
20	chaperonin HSP60, mitochondrial precursor	gi|71665068	E/M	eV2↑	nd
58	Hsc70-interacting protein	gi|70884092	E/M	nd	mV2↑
60	protein disulfide isomerase	gi|322824776	E	eV2↑	nd
69	stress-induced protein sti1	gi|322827444	E/M	eVF↑	mVF↑
102	10 kDa heat shock protein	gi|322824762	E	nd	Nd
Trafficking and membrane fusion
78	transitional endoplasmic reticulum ATPase	gi|322820539	E/M	eVF↑	nd
176	dynein	gi|322821824	E/M	nd	mV↑
178	ADP-ribosylation factor 3	gi|322821587	E/M	eVF↑	mV2↑
207	small GTP-binding protein Rab1	gi|62175231	E	eVF↑	nd
258	small GTP-binding protein Rab11	gi|70874996	E/M	eVF↑	nd
280	neurobeachin/beige protein	gi|340057219	E	nd	nd
284	clathrin heavy chain	gi|57648429	E/M	eVF↑	nd
Transporters
59	proton motive ATPase	gi|2393849	E/M	eV↑	mV2↑
137	ATPase beta subunit	gi|70883049	E/M	eV2↑	nd
197	calcium-translocating P-type ATPase	gi|322827561	E	nd	nd
218	(H+)-ATPase G subunit	gi|322827345	E/M	eV2↑	mV↑
Oxidation reduction
2	dehydrogenase	gi|61741938	E/M	nd	mV↑
11	tryparedoxin peroxidase	gi|71413207	E/M	eV↑	nd
12	D-isomer specific 2-hydroxyacid dehydrogenase-protein	gi|70876009	E/M	nd	nd
13	thiol-transferase Tc52	gi|9954109	E/M	eVF↑	mV↑
29	aldo-keto reductase	gi| 189396135	E/M	nd	mVF↑
30	tryparedoxin	gi|19171158	E/M	eVF↑	mV2↑
32	Fe-superoxide dismutase	gi| 1066332	E/M	eVF↑	nd
34	dihydrolipoyl dehydrogenase	gi|71422952	E/M	eV↑	nd
68	NADP-dependent alcohol hydrogenase	gi|322827328	E/M	nd	mV↑
134	trypanothione reductase	gi|49258301	E/M	nd	nd
Carbohydrate metabolism
8	enolase	gi|71665461	E/M	nd	nd
31	transaldolase	gi|322826394	E/M	nd	mV↑
33	phosphoglycerate kinase	gi|71411859	E/M	nd	nd
42	malate dehydrogenase	gi|322821087	E/M	nd	nd
44	2,3-bisphosphoglycerate-independent phosphoglycerate mutase	gi|322820998	E/M	nd	nd
48	malic enzyme	gi|322825994	E/M	eVF↑	mV↑
49	pyruvate phosphate dikinase 1	gi|10119899	E	eV2↑	nd
54	hexokinase	gi|71655813	E/M	eV↑	nd
63	transketolase	gi|70883151	E/M	nd	mV2↑
67	isocitrate dehydrogenase	gi|70886982	E/M	eV↑	nd
Lipid metabolism
46	sterol 24-c-methyltransferase	gi|70864663	E/M	eV2↑	nd
64	myo-inositol-1-phosphate synthase	gi|322820378	E/M	eV16↑	nd
80	fatty acid desaturase	gi|70861362	E/M	eV↑	mV2↑
81	squalene monooxygenase	gi|322829427	E	eV2↑	nd
135	3-ketoacyl-CoA thiolase	gi|322827257	E/M	eVF↑	mV↑
Protein and amino acid metabolism
1	glutamate dehydrogenase	gi|2981039	E/M	nd	nd
3	tyrosine aminotransferase	gi|408894	E/M	nd	nd
23	cystathionine beta-synthase	gi|71425069	E/M	nd	mV2↑
24	arginine kinase	gi|3831705	E/M	nd	mVF↑
25	glutamamyl carboxypeptidase	gi|71403542	E/M	eV2↑	nd
26	urocanate hydratase	gi|322817862	E/M	eVF↑	nd
40	cystathione gamma lyase	gi|70883261	E/M	eV16↑	nd
61	arginase	gi|322818400	E/M	eV16↑	mV↑
72	aspartate carbamoyltransferase	gi|322826325	E/M	eVF↑	mV↑
86	S-adenosylhomocysteine hydrolase	gi|33563100	E/M	nd	nd
Nucleic acids binding
73	poly(A)-binding protein	gi|2854066	E	nd	nd
93	threonyl-tRNA synthetase	gi|322819645	E/M	nd	mV2↑
106	seryl-tRNA synthetase	gi|322825689	E/M	eVF↑	nd
123	asparaginyl-tRNA synthetase	gi|322822542	E/M	nd	nd
158	UV excision repair RAD23-like protein	gi|322823223	E	eVF↑	nd
160	helicase	gi|340056526	E/M	eVF↑	mV↑
174	aspartyl-tRNA synthetase	gi|322823627	E/M	nd	nd
177	tyrosyl-tRNA synthetase	gi|322830198	E/M	nd	mV2↑
199	mitochondrial DNA polymerase beta	gi|322827495	E	eV2↑	nd
234	LA RNA binding protein	gi|70884717	E	nd	nd
236	isoleucyl-tRNA synthetase	gi|70872783	E/M	eVF↑	nd

^aNumber assigned to a group of protein isoforms that could not be differentiated from each other by the identified peptides.

^bE, epimastigote; M, metacyclic trypomastigote.

^cStatistical analysis was used to compare protein abundance between fractions and determine in which fraction the protein was more abundant.

^dAccording to label free quantification, proteins were determined to be enriched in vesicle (V) or vesicle free (VF) fractions, and in cases where it was not possible to determine whether a protein was enriched in V or VF the protein was assigned as not determined (nd). Among the proteins enriched in V fraction, V2 or V16 was assigned when it was possible to determine in which vesicle fraction the protein was enriched.

The majority of T. cruzi secreted/excreted proteins (85.7%) fit into 15 functional categories, whereas 89 proteins (14.3%) could not be assigned to any category (Others) (Table 1). About 56% of these proteins with unassigned function (31/55) were predicted to be secreted, being 12.7% (n=7) likely secreted by the classical secretory pathway and 43.6% (n=24) by non-classical pathways (Table S1D). The majority of proteins with unassigned function are coded by genes annotated as hypothetical that have orthologues in Trypanosoma brucei and L. major. The abundance of these proteins in the secretome and the conservation among trypanosomatids provide additional evidence for their still unknown function.

Another functional category with relatively high proportion was the nucleic acid-binding proteins (11% and 7.2%, in epimastigote and metacyclic, respectively). A considerable number of proteins within this category were RNA-binding proteins, which play an essential role in the control of gene expression in trypanosomes, modulating RNA processing, stability, turnover, and translation.⁵³ The release of such proteins in the extracellular medium may suggest a possible role in intercellular communication and gene expression regulation. These findings make us speculate whether these vesicles could be used as vehicles to exchange genetic material between parasites, and between parasites and host cells as proposed by Hecht et al.⁵⁴

Transcription and translation factors, together with ribosomal proteins, represent around 7.2% of the secretome (Table 1). Although the functional basis for this finding is unclear, it is possible that the incubation of epimastigotes in medium without FBS for 6 h have triggered the metacyclogenesis process and some proteins were disposed of via secretion during parasite remodeling. Another explanation would be that polysomes carrying mRNAs may be mobilized into secreted vesicles.⁵⁵ Also, there is a possibility that these proteins may play auxiliary roles in pathogenesis, similar to what happens with elongation factor-1α in Leishmania.⁵⁶

Proteins involved in stress response (oxidation and reduction), protein folding (heat shock and chaperones), and proteolysis comprise 22.1% of the T. cruzi secretome (Table 1). Epimastigote and metacyclic forms are subjected to stress conditions in the insect and vertebrate hosts. Secreted proteins with oxidoreductase activity are likely to be involved in intracellular survival and may provide protection against free radicals. Proteases and peptidases could be involved in the degradation of other host components, e.g., extracellular matrix.

We identified proteins that have the potential to manipulate host-cell function via interference with signaling pathways, such as 14-3-3 proteins, calcium-binding proteins, kinases, phosphatases, and adenylate cyclases (Tables 2 and S1). It is interesting to note that 14-3-3 protein is known to have anti-apoptotic properties in multicellular and also unicellular organisms like yeast.⁵⁷ In addition, various proteins associated with vesicular trafficking (4.9%) were identified in the secretome. These proteins may be considered candidate virulence factors for their potential to modulate vesicle trafficking in infected cells, e.g., affecting phagosome maturation. Furthermore, our study identified proteins previously described to play a role in host-parasite interaction, e.g., calreticulin, prolyl oligopeptidase, GP82 and GP90 glycoproteins, proline racemase, surface protease GP63, and several members of the trans-sialidase (TS) family (Tables 2 and S1). Some of these proteins were already described to be secreted,^{22, 24, 25} confirming our findings. Many of the proteins detected in T. cruzi secretome were also identified in the secretome of Leishmania donovani⁵⁸ and Trypanosoma brucei,⁵⁹ suggesting conservation in secretion mechanisms and/or in the function of secreted proteins.

To define whether proteins found in T. cruzi secretome were secreted by classical or non-classical secretory pathways, we searched protein sequences for the presence of signal peptide using SignalP and determined leaderless secreted proteins using SecretomeP. Results revealed that only ~9% of the 367 identified proteins have a signal sequence and are predicted to be secreted by classical pathways (Table S1D). Alternatively, ~48% of the proteins were predicted by SecretomeP to be secreted by non-classical pathways. Although we cannot exclude that SignalP algorithm is not able to identify all T. cruzi signal sequences, our results suggest that non-classical pathways play a major role in extracellular delivery of proteins in T. cruzi, and according to our results, some of these routes are through association with vesicles.

3.4. Differential protein enrichment revealed by label-free quantification

Many proteins identified in the T. cruzi secretome were present in more than one CM fraction, which could be due to cross-contamination during fractionation, association of soluble proteins with the extracellular vesicles, or release of proteins from damaged vesicles. Therefore, we decided to perform a quantitative analysis using a label-free method to compare protein abundance between fractions rather than a qualitative analysis. In this way, normalized spectral count values from three technical replicates were used to calculate protein enrichment according to the method described by Choi et al.⁴³ A summary of protein enrichment results is shown in Tables 2 and S1A, normalized spectral count values and detailed statistical analysis can be found in Table S2A–G. To avoid loss of information due to statistical cutoff, instead of comparing protein abundance between the three fractions together, we decided to compare pairs of fractions as follows: V2 vs. V16, V16 vs. VF, and V2 vs. VF (Tables S1A and S2). In this way, a protein like nucleoside diphosphate kinase that is enriched in vesicular fractions, but has no statistical significant difference (ns) between them (V2 vs. V16), can be recognized as being enriched in vesicles (V↑) (Table S1A). In some cases it was not possible to determine whether a protein was enriched in extracellular vesicles or vesicle-free (VF) fraction, so it was assigned as not determined (nd) (Tables 2 and S1A). It is important to highlight that the label-free quantification data (Table 2) agreed with the western blot (Figure 3) results as it showed that GP82 (i.e., glycoprotein 82 kDa, protein group # 237, Table 2) and FCaBP (i.e., calcium-binding protein, protein group # 22, Table 2) were enriched in vesicle fractions.

From the total of 367 distinct proteins identified in T. cruzi secretome, 256 proteins (70%) were determined to be enriched in vesicle (V↑) or vesicle-free (VF↑) fractions (Table S1A). Moreover, 40.6% (n=149) of the 367 proteins were defined to be enriched in one of the vesicle fractions (V2↑ or V16↑). If we compare secreted proteins identified in epimastigote and metacyclic forms, 28 proteins were defined to be enriched in the same fraction (V↑ or VF↑ in both forms); and 34 proteins were enriched in distinct fractions (V↑ in epimastigote and VF↑ in metacyclic trypomastigote, or vice versa), suggesting that some proteins might have distinct cellular locations and/or they are secreted by different means depending on the parasite form. In accordance with data from western blot and microscopy (Figures 3–4), which indicated that the fractionation protocol could separate vesicles of different origins, the quantitative proteomic analysis revealed differences in protein content between fractions. In Figure 5A–C, it is possible to notice that the protein abundance in each functional category varied substantially between V2, V16, and VF fractions, some categories being more represented in one fraction than others. Proteins enriched in V2 fraction are mainly concentrated in categories comprising hypothetical proteins (18%), carbohydrate metabolism (12%), and heat shock proteins and chaperones (12%). Conversely, proteins enriched in V16 fraction are concentrated in categories involving proteolysis/peptidolysis (22%), protein and amino acid metabolism (21%), and nucleic acid binding (14%) (Figure 5). Due to the reduced number of metacyclic trypomastigote proteins enriched in each CM fraction, only the results with epimastigotes are shown. In addition, we also used normalized quantitative proteomic values to evaluate the similarity of CM fractions and to search for data clusters using the principal component analysis (PCA), which is a mathematical algorithm that reduces the dimensionality of data while retaining most of the variation in the data set, thus identifying groups with similar data behavior (clusters) inside the whole package of data.⁶⁰ In Figure 5D, it is possible to identify clusters of the same color formed by the three replicates from each CM fraction, and that these same color clusters were grouped separated from other CM clusters, supporting our previous observation that V2, V16, and VF fractions are indeed different.

Figure 5.

Differential protein enrichment in vesicle (V2 and V16) and vesicle-free (VF) fractions. Quantitative values based on spectral count data were used to compare and determine protein enrichment in conditioned medium fractions (V2, V16, and VF). Proteins that were proved to be enriched in one fraction with statistical significance were grouped according to functional categories. Epimastigote-derived proteins enriched in V2 fraction (A), V16 fraction (B), and VF fraction (C) are shown. (D) Quantitative spectral count values from all 367 proteins were used to evaluate the similarity of conditioned medium fractions and to search for data clusters using the principal component analysis (PCA).

In some cases there was a good correlation between the fraction in which proteins were enriched and their known cellular location (Table S1A), pointing to the existence of similar secretion mechanisms for delivering proteins with the same cellular function/localization to the extracellular environment. For instance, membrane proteins present in host-parasite interaction, signaling, and transporter categories were determined to be enriched in V2 fraction, supporting the plasma-membrane origin of V2 fraction. On the other hand, it is remarkable that proteins from the cytosol and intracellular compartments, such as those involved in proteolysis, peptidolysis, and amino acid metabolism, were mostly enriched in V16 fraction. Also, proteins involved in metabolism of carbohydrates and lipids were found to be enriched in V16 fraction. The protein heterogeneity observed in V16 fraction and the finding of GFP in this fraction (Figure 3D), suggest that these vesicles have different intracellular origins: vesicles created during the inward budding process of MVB formation that engulfs small portions of the cytosol,^{61, 62} and those originated from amphisome secretion, which is an unconventional secretion pathway that is autophagy-based.^{63, 64} TEM images showed vesicles resembling autophagosomes close to the parasite flagellar pocket (data not shown), supporting the hypothesis of amphisome secretion. Several of these proteins enriched in V16 fraction have been identified in the reservosome proteome,⁶⁵ which is a MVB present in T. cruzi, suggesting that vesicles derived from these organelles are present in the secretome. Additionally, we found that V16 fraction was enriched in several kinds of tRNA synthetases. In a previous study by Garcia-Silva et al.⁶⁶, it has been shown that tRNA halves accumulate in intracellular granules that partially colocalize with the protease cruzipain, which is found in T. cruzi reservosomes. These findings, together with our unpublished data revealing that both T. cruzi epimastigote- and metacyclic trypomastigote-derived extracellular vesicles contain particularly small tRNAs or tRNA halves (Bayer-Santos et al., unpublished), suggest that the granules containing small tRNAs are indeed secreted. We are currently carrying further studies to clarify the composition and role of small noncoding RNAs in these vesicles.

Many of the secreted cytosolic proteins identified in T. cruzi secretome, e.g., glycolytic enzymes, transcriptional factors, and chaperones, have been described as virulence factors and with immunostimulatory properties. Sequences encoding the 70-kDa heat shock protein and related proteins had been routinely found by immunologic screening of cDNA expression libraries.⁶⁷ Similarly, proteomic approaches to identify and characterize major immunostimulatory proteins from soluble antigens of L. donovani identified elongation factor-2, heat shock protein 70, heat shock protein 83, aldolase, enolase, and triosephosphate isomerase.⁶⁸ As mentioned above, Leishmania elongation factor-1α was reported to be a virulence factor secreted in the cytosol of infected macrophages, responsible for cell deactivation through activation of the Src homology 2 domain containing tyrosine phosphatase-1 (SHP-1).⁵⁶ Additionally, alcohol dehydrogenase was reported to be a virulence factor in the parasite Entamoeba histolytica due to binding to host extracellular matrix.⁶⁹ In the same manner, enolase was reported in the surface of several fungi and bacteria, and in Streptococcus pneumonia the surface enolase is responsible for binding to plasminogen, leading to activation of the serine protease plasmin.⁷⁰ Overall these findings reinforce the role of the cytosolic proteins as multifunctional proteins, which might act with different purposes depending on the cell localization.

3.5. Vesicles may be used by T. cruzi metacyclic forms to deliver cargo into mammalian cells

A growing body of evidence in mammalian cells indicates that vesicles can be used to mediate intercellular communication processes by directly stimulating target cells via receptor-mediated interactions or by transferring various bioactive molecules including membrane receptors, proteins, mRNAs, and microRNAs.^{71, 72} Based on findings that demonstrated the involvement of T. cruzi excreted/secreted products in host-cell modulation,^{17, 18, 73} we hypothesized whether metacyclic forms may use extracellular vesicles to deliver virulence and modulatory factors into host cells. Therefore, we infected HeLa cells with metacyclic trypomastigotes and analyzed the infection after 2 h by confocal microscopy using mAb 3F6 against the surface glycoprotein GP82 (Figure 6). We found parasites releasing vesicles while in contact with (Figure 6A) or internalized into (Figure 6B) HeLa cells. Also, we found structures resembling vesicles attached to or internalized into HeLa cells (Figure 6C). Since vesicles derived from MVBs are too small to be visualized with a confocal microscope, we believe that the vesicles observed in our images should rather be membrane-derived ones (mV2), as supported by labeling with mAb 3F6 against the surface glycoprotein GP82. In addition, we incubated metacyclic forms with LLC-MK2 cells for 2 h and collected extracellular vesicles from culture supernatant as described in Figure 2A. Proteomic analysis showed the presence of both mammalian and parasite proteins in this fraction (data not shown). These findings contrast with recent results described by Cestari et al.,⁷⁴ who were unable to detect vesicles released by metacyclic forms in culture medium or when parasites were incubated with mammalian cells. Together, our results suggest that T. cruzi infective forms release vesicles inside mammalian cells or they are released in the medium and endocytosed by cells, reaching the cytoplasm where they can eventually release their content and modulate host cell to support parasite survival and replication. Moreover, the release of bioactive molecules associated with extracellular vesicles may be an efficient strategy employed by the parasite to protect these biomolecules against extracellular degradation.

Figure 6.

Vesicles may be used by infective metacyclic trypomastigotes to deliver cargo into mammalian cells. Metacyclic forms were incubated with mAb against the surface protein GP82, washed and allowed to invade HeLa cells for 2 h. Cells were washed to eliminate parasites that were not attached or inside cells, fixed and processed for immunofluorescence. Parasites releasing vesicles were observed while in contact with (A) or internalized into (B, C) HeLa cells. Structures resembling vesicles attached to or internalized into HeLa cells were also observed (C, arrows). Red: surface glycoprotein GP82; blue: DAPI labeling of host-cell and parasite nuclei and parasite kinetoplast; green: phalloidin labeling actin microfilaments. Scale Bar = 5 μm.

4. CONCLUSION

In conclusion, this report broadens the current view on the secretome of a eukaryotic unicellular microorganism by taking into consideration the individual proteome of two extracellular vesicle populations and soluble proteins isolated from T. cruzi conditioned medium. Proteins identified in the secretome revealed many uncharacterized candidates for further studies with potential role in the virulence and pathogenesis. This study provides evidence that support the existence of at least two mechanisms of vesicle secretion in T. cruzi, which seems to be similar to mechanisms present in mammalian cells, indicating conservation throughout the eukaryote lineage. Moreover, these results provide the first step for further investigation of secretion mechanisms in protozoan parasites and to the establishment of potential protein markers for each extracellular vesicle population. In the same way, the study is an important step to substantiate further characterization of the biological functions of extracellular vesicles and to define its lipid and nucleic acid composition.