Membrane associated proteins of two Trichomonas gallinae clones vary with the virulence

Summary

Oropharyngeal avian trichomonosis is mainly caused by Trichomonas gallinae, a protozoan parasite that affects the upper digestive tract of birds. Lesions of the disease are characterized by severe inflammation which may result in fatality by starvation. Two genotypes of T. gallinae were found to be widely distributed in different bird species all over the world. Differences in the host distribution and association with lesions of both genotypes have been reported. However, so far no distinct virulence factors of this parasite have been described and studies might suffer from possible co-infections of different genotypes. Therefore, in this paper, we analyzed the virulence capacity of seven clones of the parasite, established by micromanipulation, representing the two most frequent genotypes. Clones of both genotypes caused the maximum score of virulence at day 3 post-inoculation in LMH cells, although significant higher cytopathogenic score was found in ITS-OBT-Tg-1 genotype clones at days 1 and 2, as compared to clones with ITS-OBT-Tg-2. By using one representative clone of each genotype, a comparative proteomic analysis of the membrane proteins enriched fraction has been carried out by a label free approach (Data available via ProteomeXchange: PXD013115). The analysis resulted in 302 proteins of varying abundance. In the clone with the highest initial virulence, proteins related to cell adhesion, such as an immuno-dominant variable surface antigen, a GP63-like protein, an armadillo/beta-catenin-like repeat protein were found more abundant. Additionally, Ras superfamily proteins and calmodulins were more abundant, which might be related to an increased activity in the cytoskeleton re-organization. On the contrary, in the clone with the lowest initial virulence, larger numbers of the identified proteins were related to the carbohydrate metabolism. The results of the present work deliver substantial differences between both clones that could be related to feeding processes and morphological changes, similarly to the closely related pathogen Trichomonas vaginalis.

Keywords:

DNA-binding proteins – Protein domains – Metabolic processes – Trichomonas vaginalis – Protein metabolism – Cell metabolism – DNA metabolism

Introduction

Trichomonas gallinae is a protozoan that affects the upper digestive tract of domestic and wild birds, although columbiformes are considered the principal reservoir of the parasite. The main route of transmission of this flagellated microorganism among columbiformes is feeding chicks with crop milk or sharing food or water sources. Recently, T. gallinae has been recognized as the etiological agent of epidemic outbreaks of disease affecting wild finches that spread across Europe and North America [1].

Although T. gallinae was described as the etiological agent of avian trichomonosis for some time, the advancement of molecular epidemiology has elucidated the diversity of trichomonads present in the oropharynx of birds in the recent decades. Several genotypes of T. gallinae have been described in distinct studies, employing different nomenclature [2–6]. In addition, the occurrence of new or tentative species has been reported, some of them associated with particular avian hosts [2, 5, 7–10]. However, not all of them have been related to lesions and disease. Several factors can influence the development of pathognomonic lesions or clinical signs associated with avian trichomonosis. Among them, the age, previous contact with the parasite, diet and type of host have been reported [1, 6].

Two genotypes of T. gallinae are widely distributed among several groups of birds: genotypes A and C [2] which correspond with genotypic groups Internal Transcribed Spacer-Oropharyngeal Bird Trichomonad-T.gallinae 1 (ITS-OBT-Tg-1) and Internal Transcribed Spacer-Oropharyngeal Bird Trichomonad-T.gallinae 2 (ITS-OBT-Tg-2) [6]. Although both genotypes have been described in clinical cases, ITS-OBT-Tg-1 appears to be more frequently associated with disease [3, 6, 11–13]. However, the virulence of the distinct genotypic groups has been only partially investigated [14].

So far only two studies have been carried out on the proteomics of T. gallinae [15, 16]. One of them identified cysteine peptidases related to the pathogenic effect of the parasite in cell cultures [15], while the other predicts glyceraldehyde-3-phosphate hydrogenase (GAPDH) as a promising drug target [16]. In UniProt database there are only 71 annotations for T. gallinae, corresponding to 9 proteins. On the contrary, several proteomic studies have been performed with the closely related protozoan Trichomonas vaginalis, the etiological agent of the human venereal disease called trichomonosis and more than 50000 proteins have been annotated in UniProt. Some studies have focused on the composition of surface proteins of T. vaginalis, since they are involved in several mechanisms of inter-cell communication, such as nutrient acquisition, adhesion to epithelial cells and different virulence mechanisms [17–23]. It is known that several surface protein families are present in T. vaginalis: BspA-like proteins, surface proteases, variable surface proteins and lectins, together with proteins harboring dual functions in the parasitic cell. BspA-like proteins act by binding to epithelial cells in many organisms, including Bacteria, Archea and Eukaryota [17]. Surface proteases participate in inflammation processes, neoplastic mechanisms and cell adhesion, and include cysteine proteases, serine proteases and metalloproteinases such as GP63-like proteins [17, 19, 20–22]. Proteins with dual functionality are called moonlighting proteins, since they are present in different compartments of the cell, including the plasmatic membrane. Among them, GAPDH family and adhesins, a group that englobes the malic enzyme, have been described in T. vaginalis [17, 21].

The membrane system of the cells is composed of the cell outer membrane and associated proteins, including proteins of the membrane associated organelles. Extensive vesicular traffic occurs connecting the organelles one to each other and also to the outside of the cells and many signaling pathways take place there. Although some membrane proteins are known to be present in both, T. vaginalis and T. gallinae, it is reasonable to expect that many others are specific and not easily identifiable by comparison, due to the difference in host (man and birds).

In this paper, we have chosen seven clones of the two main genotypic groups of T. gallinae, and analyzed the virulence by in vitro cell culture-based assay. According to their initial virulence, two isolates were further selected for a comparative proteomic analysis. The ability of T. gallinae to change its shape and adhere to epithelial cells, together with excretion of cysteine peptidases, are essential characteristics for virulence [1, 15]. Thus, the knowledge of the protein composition of the membrane is crucial to elucidate those processes. Therefore, the aim of the present work was to analyze differences in the enriched membrane proteins fraction after subcellular fractionation of the trophozoites.

Material and methods

Isolates, culture and preservation

Isolates of T. gallinae were obtained in the context of routine diagnosis of avian trichomonosis in the wildlife recovery center of GREFA (Madrid, Spain). Veterinary supervision and care of wild animals, as well as samples for disease diagnosis, are routine procedures for all the animals at admission. Bird samples were collected by veterinarians by gentle rubbing sterile cotton swabs in the oropharynx of birds and subsequently cultured in 5 ml of fresh TYM (Trypticase Yeast Maltose) medium pH 6.5 at 37°C containing 10% inactivated fetal bovine serum (FBS) (Sigma-Aldrich, St. Louis, Missouri, USA), antibiotics (36 mg/l each of ticarcillin, vancomycin and ceftiofur) and 10,000 IU/ml nystatin (Sigma, St. Louis, Missouri, USA). The tubes were monitored daily using an inverted microscope for the presence of viable trichomonads, for a maximum period of 15 days. Aliquots of 1 ml of each culture were kept for DNA isolation of the original isolate.

Passages were performed every 48 hours using 100 μl of sample, until bacterial or fungal contamination was controlled, usually after 3–6 passages. The presence of bacteria and fungi was examined using Nutrient agar (Scharlab, S.L., Barcelona, Spain) and Sabouraud agar (bioMérieux, Madrid, Spain) by plating 100 μl of culture on each agar plate. Nutrient agars were incubated at 37°C for 24 hours, while Sabouraud agars were maintained preserved from light at 22°C for 15 days.

Preservation of the isolates was done using 5% of dimethyl sulfoxide (DMSO, Sigma-Aldrich, St. Louis, Missouri, USA) in TYM medium. The cryovials were placed in a freezing container that provided a slow cooling rate (-1°C/min) in a -80°C freezer. For long term storage (>12 months) samples were transferred into a -150°C freezer.

Establishment of clonal cultures

The isolates used in the present work were selected based on their origin: three isolates were obtained from animals with macroscopic lesions characteristic for avian trichomonosis and four isolates were obtained from animals without macroscopic lesions (Table 1).

Each isolate was thawed in LMH culture medium consisting of medium 199 (M199) supplemented with Earle´s salts, 100 mg/ml L-glutamine, 25mM HEPES, L-amino acids (Gibco^™, Thermo Fisher Scientific, Vienna, Austria), 0.22% of sterilized rice starch (Carl-Roth, Karlsruhe, Germany) and 15% of heat inactivated fetal bovine serum (Sigma, EEUU). Escherichia coli DH5α-T1 incubated in a horizontal rotating incubator at 37°C in the supplemented medium for 24 hours was added to the cultures in a proportion of 1/20 (v/v) (LMH culture medium), before inoculation of the parasites. Prior to the establishment of the clonal culture by a previously described micromanipulation procedure [24], the original isolates were cultivated for a short period of time (maximum 7 passages). Passages of the cultures were done every 48–72 hours.

One day before the micromanipulation procedure, 0.5 ml of each Trichomonas isolate culture was transferred into 9 ml of LMH culture medium supplemented with E. coli. At 24 hours post-inoculation, 20 μl of these cultures were carefully placed on glass slides and diluted with pure M199 in order to have a clear suspension of trophozoites. The selection of single trichomonad cells was done under an inverted microscope (Diaphot 300, Nikon, Austria) by a micromanipulator (Narishige, Japan). Single active trophozoites were retained by negative pressure with a micropipette. Each single trophozoite was deposited in a drop of M199 to visualize the correct selection of the cell and then transferred to a micro tube with 0.5 ml of LMH culture medium supplemented with E. coli. The tubes were incubated at 37°C and examined every 48 hours for trophozoite growth. Positive clonal cultures were transferred into 9 ml of LMH culture medium supplemented with E. coli for successive culturing and cryo-preservation until use.

Clones of the genotypic group ITS-OBT-Tg-1, obtained from animals with lesions, including R17-12 C1 from a Eurasian eagle-owl (Bubo bubo), R193-13 C3 from a common kestrel (Falco tinnunculus) and P349-12 C11 from a common wood pigeon (Columba palumbus). Clones of the genotypic group ITS-OBT-Tg-2, obtained from animals without lesions, including P178-13 C7 from a common wood pigeon (Columba palumbus), R24-12 C10 from a common kestrel (Falco tinnunculus) and R44-12 C5 from a northern goshawk (Accipiter gentilis). One isolate (P95-13 C3) that belonged to genotype ITS-OBT-Tg-1 was obtained from a bird without clinical signs. The information about clones, including host species of origin, presence of macroscopic lesions and ITS genotypes is shown in Table 1. From each isolate, between 6 and 25 clones were successfully established. One clone from each isolate was selected for the virulence assays.

Axenization procedure

Since clonal T. gallinae trophozoites were cultured in the presence of E. coli, penicillin, streptomycin and meropenem (Sigma-Aldrich, St. Louis, Missouri, USA) were added for the axenization procedure according to Amin et al. [24]. Tryptose-soy-agar (TSA) and tryptose-phosphate-broth with 0.2% of agar (TPB, Gibco^™, Thermo Fisher Scientific, California, USA) were used to monitor bacterial and fungal contamination (Scharlab, S.L., Barcelona, Spain). TSA plates and TPB tubes were incubated 24 h at 37°C before examination. Once the trophozoites were growing axenically, they were transferred to Hollander Fluid medium at pH 7.2 (HF medium) [24].

DNA isolation and genetic characterization of the isolates and clonal cultures

DNA was extracted from 1 ml of the initial parasite culture or at passage level <3. From clonal cultures, 1 ml was also used for DNA extraction after the axenization process was completed. DNeasy blood and tissue extraction kit (QIAGEN, Valencia, California, USA) was employed following the manufacturer´s instructions and DNA was stored at -20°C until use.

ITS1/5.8S rRNA/ITS2 (ITS) PCR

The protocol of Felleisen [25] was applied for amplification and sequencing of the ITS1/5.8S rRNA/ITS2 region using primers TFR1 (5´-TGCTTCAGCTCAGCGGGTCTTCC-3’) and TFR2 (5’-CGGTAGGTGAACCTGCCGTTGG-3’). PCR reaction was carried out in 50 μl with 5 μl of 10x PCR buffer, 1.5 mM of MgCl₂, 2 mM of dNTP, 2 μM of each primer, 2.5 IU of Taq polymerase (MP, Thomas Scientific, Swedesboro, New Jersey, USA) and 5 μl of genomic DNA. PCR protocol was as follows: initial denaturation step at 95°C for 9 min followed by 40 cycles of denaturation for 30 s at 94°C, annealing at 66°C for 30 s, extension at 72°C for 30 s and a final extension step at 72°C for 15 min. Positive and negative controls were included in each PCR reaction set.

Small subunit of ribosomal RNA (SSUrRNA) PCR

Amplification of the SSUrRNA was carried out using the protocol of Ganas et al. [26] and the primers HM long f (5′-AGGAAGCACACTATGGTCATAG-3′) and HM long r (5′-CGTTACCTTGTTACGACTTCTCCTT-3′). The PCR was performed in 50 μl composed of 25 μl of HotStarTaq Master Mix kit polymerase (QIAGEN, Hilden, Germany), 16 μl of PCR water (QIAGEN, Hilden, Germany), 2 μl of each primer at 10 μM and 5 μl of genomic DNA. The temperature profile was as follows: 15 min at 95°C for initial denaturation, 40 cycles of 30 s at 94°C, 1 min at 55°C, 2 min at 72°C and a final extension step of 10 min at 72°C. Positive and negative controls were included in each PCR reaction set.

Fe-hydrogenase gene (FeHyd) PCR

For this genetic marker the same PCR reaction mix as for the SSUrRNA was prepared according to Lawson et al. [12] with the following primers: TrichhydFOR (5´-GTTTGGGATGGCCTCAGAAT-3´) and TrichhydREV (5´-AGCCGAAGATGTTGTCGAAT-3´). The thermal cycler was programmed for: 15 min at 95°C for initial denaturation, 40 cycles of 30 s at 94°C, 1 min at 58°C, 2 min at 72°C and lastly, 10 min at 72°C. Positive and negative controls were included in each PCR reaction set.

Sequencing. Amplicons were purified with a MiniElute PCR purification kit (QIAGEN, Valencia, California, USA) and submitted for Sanger sequencing to Sistemas Genómicos S.A., Valencia, Spain. An automatic sequencer (3730XL DNA analyzer, Applied Biosystems, Foster City, California, USA) was employed using the ABI PRISM BigDye© Terminator Cycle sequencing kit (Applied Biosystems, Foster City, California, USA). Sequencing was carried out in both directions of DNA and the resultant chromatograms were assembled and manually checked using Lasergene SeqMan software version 7.0.0 (DNASTAR, Madison, Wisconsin, USA). Alignment was done using the BLAST online resource of the NCBI.

Virulence assays of T. gallinae clonal cultures in LMH cells

LMH cell culture

The chicken hepatocellular carcinoma cell line (LMH) was used for virulence assays, since it provided a suitable system for the in vitro pathogenicity testing of T. gallinae [14]. LMH cells were cultured in RPMI-1640 supplemented with 10% FBS, 40,000 IU/ml penicillin and 40 mg/ml streptomycin (Gibco^™, Thermo Fisher Scientific, California, USA). Cells were incubated in 5% CO₂ and 85–90% humidity at 37°C.

Ten ml of medium with an initial inoculum of 1.4 x 10⁶ cells were cultured in 25 cm² filtered cap flasks (Nunc^™, Thermo Fisher Scientific, California, USA). The cultures were split every 3–4 days by adding 2 ml of a trypsin solution (0.05% of trypsin in 0.53 mM EDTA, Thermo Fisher Scientific, California, USA) for 3–5 min followed by washing of the cells in 50 ml of sterile PBS (pH 7.2). After centrifugation at 1,300 rpm for 5 minutes the cells were suspended in 10 ml of fresh medium. To assess cell viability, trypan blue exclusion was used (Sigma-Aldrich, St. Louis, Missouri, USA).

Inoculation of LMH cells

Prior to the inoculation with trichomonads, LMH cell cultures were washed with sterile PBS and 6.5 ml of pre-warmed M199 supplemented with 10% FBS, 10% TPB and penicillin-streptomycin solution. LMH cells were used for inoculation of trophozoites when 95% confluence was obtained (monolayer stage).

An inoculum of 10,000 axenized viable trophozoites was used in each flask of LMH cell cultures in monolayer. Three replicates were performed for each clone together with flasks with solely trophozoites as positive controls and flasks with solely LMH cells as negative controls. The cultures were incubated for five days and the number of trophozoites was determined every 24 hours.

Determination of cytopathogenic effect

The cultures were monitored by an inverted microscope for five days to assess the cytopathogenic effect. In parallel, the parasites were counted by trypan blue exclusion every 24 hours. Cytopathogenic effects (CPE) were scored according to Amin et al. [14]: CPE1: ≤25% of monolayer destruction, CPE2: 25–50% of monolayer destruction, CPE3: 50–75% of monolayer destruction and CPE4: >75% of monolayer destruction.

Statistical analysis

Wilcoxon signed rank test was employed for statistical significance in the growth of trophozoites with or without LMH cells. Categorical data for CPE scores grouped by ITS genotype were analyzed using the Chi square test of association (p-value < 0.01). All analyses were performed with the IBM SPSS Statistics software version 22 (IBM Corp., Armonk, New York, USA).

Subcellular fractionation and protein extraction and quantification

Based on the virulence assays, two of the clonal cultures of T. gallinae were selected for the proteomic analysis: at 24h p.i. clone R17-12 C1 showed the highest CPE score (4) whereas P178-13 C7 caused the lowest CPE score (1).

T. gallinae mass culture and harvest

Three biological replicates were performed with each clone. The clones were thawed in LMH culture medium supplemented with E. coli at 37°C. The parasites were progressively axenized as previously described and adapted to TYM medium. Mass cultures were generated by the inoculation of 1 x 10⁵ axenic trophozoites in 40 ml of TYM medium on 20 Falcon tubes of 50 ml and harvested at the end of the log phase (48 hours). In this case, trophozoites grew in TYM medium without antibiotics and antifungals. Clone R17-12 C1 was at passage 9 and clone P178-13 C7 at passage 11. Cell viability was assessed at 24 and 48 hours by trypan blue exclusion. Between 1.1 x 10⁸ and 1.7 x 10⁷ trophozoites were harvested from each Falcon tube by centrifugation at 2,000 rpm for 5 min and then washed 3 times with sterile PBS (pH 7.2) at 4°C to remove growth medium. The obtained cell-pellets were used for subsequent subcellular fractionation.

Subcellular fractionation

The commercial subcellular fractionation kit ProteoExtract^™ Subcellular Proteome Extraction kit (Calbiochem^™, Merck KGaA, Darmstadt, Germany) was employed for this purpose. Manufacturer instructions were followed to avoid protein degradation. Briefly, cells were suspended with protease inhibitor cocktails and a series of extraction buffers that dissolved different proteome fractions were applied. Firstly, soluble proteins corresponding to the cytosolic fraction were extracted. Then, the fraction containing membrane and membrane-bound organelles proteins was obtained. Thirdly, nuclear proteins and the cytoskeletal fraction were recovered. During the procedure the samples were kept on ice to avoid protein degradation. Incubation was done in mechanical agitators to assess the correct distribution and solubilization of the extraction buffers. The obtained fractions were kept at -80°C for further analysis. Fraction two (enriched membrane proteins fraction) was used for this study.

Protein precipitation and quantification

The enriched membrane proteins fraction was gradually thawed from -80°C to 4°C and kept on ice for acetone precipitation. Six volumes of acetone at -20°C were added to one volume of sample. The mix was kept at -20°C for 4 hours and then centrifuged at 8,000 rpm for 10 minutes. Acetone was carefully poured from the tube and the pellet was kept at -80°C. For solubilization, samples were suspended in 100 μl of 8M urea. Protein quantification was done by the RC DC^™ Protein Assay (Bio-Rad Laboratories, Hercules, California, USA). The same amount of protein was used for each replicate of the clones before performing further analysis.

SDS-PAGE preliminary analysis

A preliminary analysis by a 12% SDS-PAGE gel using 10 μg/lane of each sample from the enriched membrane proteins fraction of the two selected clones were loaded to compare the protein pattern of the clones. The same amount of protein was loaded from each replicate as a quality control for the protein quantification. Precision Plus Protein^™ Standards Dual Color de (Biorad, Alcobendas, Spain) were used as molecular weight markers.

Digestion and desalting of the peptides

Thirty mg of sample from each replicate were reduced with 10 mM of dithiothreitol (DTT) at 37°C for 30 minutes. Proteins were alkylated with 55 mM of iodocetamide in the dark for 20 min. Protein digestion was carried out with 1/30 (w/w) of trypsin (recombinant trypsin sequencing grade, Roche) and left overnight at 37°C [27]. Digested peptides were desalted and concentrated by reverse chromatography (C18 OMIX pipette tips, Agilent Technologies, Germany). Peptides were eluted with 50% acetonitrile (ACN) and 0.1% of trifluoroacetic acid. Finally, the samples were dried in a speed-vac and resuspended in 0.1% of formic acid before de Nano LC-MS/MS analysis. Samples were stored at -20°C for further procedures.

Label free experimental design

Three biological replicates were used from each clone. A total of two technical replicates of each biological replicate were analyzed, except for one technical replicate of clone R17-12 C1 (R17-R1-2) which had an insufficient amount of protein. From clone R17-12 C1 samples R17-R1-1, R17-R2-1, R17-R2-2, R17-R3-1 and R17-R3-2 were employed, while from clone P178-13 C7 samples P178-R1-1, P178-R1-2, P178-R2-1, P178-R2-2, P178-R3-1 and P178-R3-2 were used (Fig 1).

LC-MS/MS analysis (LTQ-Orbitrap-Velos-Pro)

The desalted peptides samples were analyzed by RP-LC-MS/MS in an Easy-nLC II system coupled to an ion trap LTQ-Orbitrap-Velos-Pro mass spectrometer (Thermo Fisher Scientific, California, USA). The peptides were concentrated (on-line) by reverse phase chromatography using a 0.1 mm × 20 mm C18 RP precolumn (Thermo Scientific, California, USA), and then separated using a 0.075 mm x 250 mm C18 RP column (Thermo Scientific, California, USA) operating at 0.3 μl/min. Peptides were eluted using a 180-min gradient from 5 to 40% solvent B (Solvent A: 0.1% formic acid in water, solvent B: 0.1% formic acid and 80% ACN in water). For peptide electrospray ionization (ESI) a nanoelectrospray source and 30 mm capillary emitters were employed (Nano-bore emitters stainless steel ID 30 μm, Proxeon, Thermo Fisher Scientific, California, USA).

The mass spectrometer was run in data-dependent positive ionization for automatically detection from MS to MS/MS spectra (LTQ-Orbitrap-Velos-Pro). Peptides were detected with a 30,000 resolution in full scan MS mode with 400–1,600 atomic mass units (amu). A maximum of 15 dependent MS/MS scans were selected according to their intensity. An isolation width of ± 2 mass-to-charge ratio units (m/z), normalized collision energy of 35% and dynamic exclusion were applied for periods of 30 seconds.

The mass spectrometry proteomics data have been deposited to the ProteomeXchange Consortium via the PRIDE [28] partner repository with the dataset identifier PXD013115 and 10.6019/PXD013115.

Peptide identification and relative quantification

Peptide identification

Peptide identification from raw data was carried out using licensed version of search engine MASCOT 2.3.0 thorough Proteome Discoverer^™ 1.4 (Thermo Fisher Scientific, California, USA). The non-redundant NCBI protein database with taxonomic restriction to family trichomonadidae (103,432 protein sequences) was selected for further analysis, since no specific database of T. gallinae was available and T. vaginalis is a phylogenetically closely-related organism with a broad database available. The search was optimized with the following criteria: 20 ppm of tolerance for precursor ions and 0.8 Da for MS/MS fragment ions, tryptic cleavage after arginine and lysine, up to two missed cleavage sites allowed, optional modification of methionine oxidation and fixed modification of carbamidomethylation of cysteine. Search against decoy database (integrated decoy approach) was used to False Discovery Rate (FDR) estimation and MASCOT percolator filter were applied to MASCOT results. Acceptance criteria for protein identification were FDR < 1% and at least two peptides identified with a confidence interval (CI) > 95% (p-value < 0.01).

Protein relative quantification

Progenesis QI software version 4.1 (Nonlinear Dynamics, Waters company, Newcastle upon Tyne, UK) was employed for the relative quantification of the identified proteins. This program detects and quantifies all the peptide ions produced in the LC-MS/MS and analyses protein expression. The abundance of a specific protein was calculated as the average normalized abundance of its unique peptides in each run, following the recommendations of the Progenesis users guide. Differential expressed proteins were considered those with a fold change ≥ 2 for their average value of normalized intensity ratio.

Predicted protein functions were assigned using the UniProtKB database which includes Swiss-Prot (551,987 protein sequences) and TrEMBL (66,905,753 proteins) (UniProt, 2016). BLAST analysis was performed and Gene Ontology (GO) information entries were used for the inference of biological functions (February 2019). The free software Cytoscape version 3.5.1. was used to represent the association of the identified proteins with their specific functions [29].

Results

Isolates and clonal cultures

Seven isolates obtained from the oropharynx of different species were cloned for further investigations of the virulence and subsequent proteomic experiments. Three of them were obtained from birds with lesions and one form a bird without lesions that belonged to genotypic group ITS-OBT-Tg-1. Three other isolates were obtained from animals without lesions and belonged to genotypic group ITS-OBT-Tg-2. Results of the Multilocus Sequence Typing (MLST) analysis of the clones indicated that sequences were identical to other sequences found in other studies, with the exception of 1 single nucleotide polymorphism (SNP) in the ITS region of clone P178-13 C7 (Table 1).

Virulence assays and growth curves

Virulence assays in LMH cells

For the majority of clonal cultures, trophozoites reached the maximum growth at 48 hours post-inoculum (p.i.) in the presence of Leghorn Male Hepatoma (LMH) cell line, with the exception of the clone R44/12 C5 that reached the maximum growth at 72 hours p.i. (Fig 2). After reaching the maximum growth, the number of viable trophozoites started to decrease. The average maximum number of viable trophozoites in the presence of LMH cells ranged between 1.3 x 10⁷ and 1.6 x 10⁷ parasites per flask (7 ml) for genotypic group ITS-OBT-Tg-1 clones, and between 1.0 x 10⁷ and 1.2 x 10⁷ trophozoites per flask for genotypic group ITS-OBT-Tg-2 clones (Fig 2). Trichomonads cultivated without LMH cells reached growth peaks at 48 or 72 hours p.i. (Fig 2 and S1 Data). However, the amount of trophozoites per flask was significantly lower than in the presence of LMH cells (p-value < 0.05).

Clones P349-12 C11 and R17-12 C1, of genotypic group ITS-OBT-Tg-1, produced the maximum cytopathogenic effect (CPE) score (4) as early as 24 hours p.i. (Table 2). The other two clones of the same genotypic group, P95-13 C3 and R193-13 C3, were not as virulent 24 hours p.i. and displayed CPE scores 3 and 2, respectively (Table 2). In contrast to the above mentioned clones of ITS-OBT-Tg-1, at 24 hours p.i. all three clones of genotypic group ITS-OBT-Tg-2 displayed low virulence with CPEs scores 1 and 2 (Table 2). At 48 hours, all clones of genotypic group ITS-OBT-Tg-1 reached the maximum score (4) whereas for clones of genotypic group ITS-OBT-Tg-2 the CPE score varied from 2 to 4 (Table 2). There were significant statistical differences on the average CPE score between genotypic groups ITS-OBT-Tg-1 and ITS-OBT-Tg-2 on day 1 and 2 (p-value < 0.01, Chi-square Test). However, from the day 3 p.i., all clones showed the maximum CPE (score 4) (Table 2). Two clones were selected for further studies, clone R17-12 C1, with the highest initial virulence, and clone P178-13 C7, with the lowest initial virulence.

Growth curves in TYM medium

Growth of trophozoites from the tow selected clones in TYM medium reached the maximum value at 72 hours p.i. Although a lower growth was observed in clone R1-12 C1, no significant differences were observed between clones at 24, 48 and 72 hours p.i. (S1 Fig and S1 Data).

Protein identification

SDS-PAGE preliminary analysis

The 1D-SDS-PAGE gels showed minor differences in the protein profile between the two selected clones, consequently, a label-free method of comparison and identification was carried out (S2 Fig).

LC-MS/MS analysis

Clones R17-12 C1 and P178-13 C7 were selected according to their initial virulence on cell culture (24h and 72 h to reach the maximum CPE score, respectively) for the comparative analysis of the enriched membrane proteins fraction composition. The mean percentage of matched MS/MS spectra was closed to 2%. Similar chromatogram profiles were obtained for each biological and technical replicate of the same clonal culture, but substantial differences were observed between the peak profiles of the selected clones R17-12 C1 and P178-13 C7 (S3 Fig) (ProteomeXchange: PXD013115).

Quantification of relative protein abundance and predicted functions

Proteins identified in every technical and biological replicate from the two clonal cultures, containing more than 2 peptides and a p-value < 0.01 were quantified. Those with ≥ 2 fold change difference in abundance between the two clones were considered different in the analysis and those with ≥ 3 fold change difference in abundance were further discussed. In total, of the 750 identified proteins, 302 were found with ≥ 3 fold different abundance, of which 132 were more abundant in clone R17-12 C1 and 170 in clone P178-13 C7.

For the majority of proteins, no information about their subcellular location was available (62% in clone R17-12 C1 and 73% in clone P178-13 C7) (Fig 3). Among the identified proteins more abundant in the ITS-OBT-Tg-1 clone (highest initial virulence score), a higher percentage of cell membrane, cytoskeleton and cytoplasm proteins were detected when compared with ITS-OBT-Tg-2 clone (lowest initial virulence score) (13%, 9% and 11% respectively in clone R17-12 C1 vs. 8%, 5% and 4% in clone P178-13 C7). On the contrary, ITS-OBT-Tg-2 showed a higher percentage of organelle membrane proteins and nuclear proteins (4% and 1% in clone R17-12 C1 vs. 6% and 5% respectively in clone P178-13 C7).

The group of proteins involved in the cellular metabolism showed the highest abundance, with 33% and 40% of the identified proteins in clones of genotypic groups ITS-OBT-Tg-1 and ITS-OBT-Tg-2, respectively (Fig 4). A large amount of the proteins, being present in both clones, for which no function could be predicted, was retrieved (35%). The percentages of proteins that differed in abundance between both clones and which participate in cell adhesion, DNA replication, exocytosis, signal transduction or translation were similar (2%, 4%, 3/2%, 2/3% and 4%, respectively). Among the identified proteins more abundant in the ITS-OBT-Tg-1 clone, a larger percentage from cellular components and cytoskeleton were identified (10% and 8% in clone R17-12 C1 vs. 5% and 6% in clone P178-13 C7).

Identified proteins

The largest group of identified proteins corresponded to hypothetical proteins as inferred from the genome of T. vaginalis isolate G3 [30], 42.4% in clone R17-12 C1 and 44.1% in clone P178-13 C7, respectively (S1 and S2 Tables). Although the majority could not be linked to any category or predicted function, some of the proteins were connected to cytoskeleton, protein metabolism, vacuoles, mitochondria, Golgi apparatus, or assembly of ribosomes. Identified proteins based on peptide sequences that differed in abundance between clones were grouped according to their category.

Among proteins identified as more abundant in the clone with the highest initial virulence score (R17-12 C1) (Table 3 and Fig 5) are orthologes of two proteases, the immuno-dominant variable surface antigen-like protease, and the GP63-like protein. These proteases were shown to be involved in cell adhesion of other protozoa and pointed out as virulence factors [17, 22]. In addition, two other proteases have been found more abundant in the clone with the highest initial virulence, Clan SB family S8 subtilisin-like serine peptidase and Clan CA family C2 calpain-like cysteine peptidase. Another important group of identified peptides matched with proteins that belong to the cytoskeleton, like alpha-tubulin, Tctex-1 family protein and the erythrocyte membrane-associated giant protein antigen 332 containing protein. Also, proteins of the Ras superfamily or other components acting in signal transduction were more abundant in the clone with the highest initial virulence score. This includes RhoGAP domain containing protein, ARF GAP-like zinc finger-containing protein, Ras family protein and RhoGEF domain containing protein. Three calmodulins were also recognized: EF hand family protein, putative calmodulin and putative centrin. Two of the identified proteins potentially participate in membrane trafficking, the C2 domain containing protein and Vps52 / Sac2 family protein, and other three proteins take part in vesicle-mediated transport (endocytosis and exocytosis): putative SNARE protein, CMGC family protein kinase and SacI homology domain containing protein. It is also important to mention the presence of two putative malate dehydrogenases, which are related to the carbohydrate metabolic process and are also suggested to be virulence factors [18, 23]. Finally, there are proteins involved in carbohydrate and metabolic processes, DNA replication and translation that were more abundant in clone R17-12 C1.

In the clone with the lowest initial virulence (P178-13 C7), proteins associated with stress response, such as HSP 70 and HSP 90, were identified (Table 4 and Fig 6). Only two of the proteins were related to the cytoskeleton: putative profilin and putative alpha-actinin. Two of the proteins which were found more abundant in this clone participate in signal transduction: the putative small GTP-binding protein and GTPase-activator protein. Another group of identified proteins was associated with vesicle-mediated transport (endocytosis and exocytosis) and transmembrane transport: GDP dissociation inhibitor family protein, Dynamin central region family protein, Clathrin and VSP domain-containing protein, ABC transporter family protein, hydrogenosomal carrier protein and phospholipid-translocating P-type ATPase, flippase family protein. A large group of proteins that were also more abundant in the clone with the lowest initial virulence score were related to cellular metabolism, more precisely carbohydrate metabolism. Among them, lactate dehydrogenase family protein, glucose-6-phosphate 1-dehydrogenase family protein, glucose-6-phosphate isomerase family protein, galactokinase family protein, alpha amylase, catalytic domain containing protein, glycosil transferase group 1 family protein, 6-phosphogluconate dehydrogenase decarboxylating family protein and putative fructose 1,6-bisphosphate aldolase, can be mentioned. Also, the cysteine peptidase Clan CA, family C1, cathepsin L-like and the metalloprotease Clan MH, family M20, peptidase T-like were more abundant in this clone. Finally, a group of proteins implied in protein metabolic processes, DNA replication, translation and translocation were also identified as more abundant in clone P178-13 C7.

Discussion

Considering that mixed infections have been detected during the natural course of the T. gallinae infection, the use of clonal cultures, established by micromanipulation, guarantees genetic homogeneity of the investigated parasite [3, 4, 6, 9, 11]. This is crucial in order to perform detailed studies on the parasite’s biology, especially the proteome.

The employment of cell cultures to determine the pathogenicity of T. gallinae isolates offers an in vitro approach in addition to animal trials. Some of the variables that can interfere with the results, such as the individual immune response of the animals, are not influencing the in vitro system and can therefore easier be standardized as well as better controlled in its variability. Differences in virulence between the clones were detected at 24 and 48 hours post-infection and were associated with the parasite genotype. Amin et al. [14] also observed variation in the CPE score of the T. gallinae clones employed in their study, in which one clone of genotypic group ITS-OBT-Tg-1 caused complete cell monolayer destruction whilst a ITS-OBT-Tg-2 clone induced only 50% of destruction. These results are in agreement with the observations from natural infections in some studies carried out in Europe [6, 11–13].

The membrane proteins enriched fraction of the parasite was selected for further proteomic characterization due to the potential role of the membrane proteins in the process of adhesion to host cells and signal trafficking. The adhesion is important to maintain the colonization of virulent protozoa, but also of less virulent ones. The role of membranes proteins in the establishment of infection has already been demonstrated with other protozoa, such as Giardia duodenalis and Entamoeba histolytica [31, 32]. Some of the identified and quantified proteins, such as enolase, have been also identified as moonlighting proteins with double functions in other protozoa, like Toxoplasma, Trypanosoma or Trichomonas spp. and for that reason they could have been found in different subcellular locations [33]. Vesicular trafficking is extensive in the membrane system, favoring the translocation of proteins along the cell compartments.

Among the proteins that differed in abundance, a large amount of proteins with no assigned function was revealed. This result was expected, as many of the membranes proteins are specific for each microorganism and are possibly implicated in host-parasite species-specific interactions [19].

Colonization of the host is a first step necessary for either commensal or pathogenic microorganisms [34]. In this sense, the identification of peptides from different surface proteins of the Leucine Rich Repeat family proteins and surface antigen BspA-like, with different abundance in both clones, might be related to the colonization of the mucosa and the establishment of a chronic infection, but not necessarily with virulence. These proteins are commonly present in a broad range of organisms, including all three domains of life (Eukaryota, Archae and Bacteria), all of them residing on mucosal surfaces or body cavities [17].

Some proteases or other molecules involved in cell adhesion, like the immuno-dominant variable surface antigen-like or the GP63-like, which are considerably more abundant in the clone with the highest initial virulence score (18.96 and 7.78 fold change, respectively), have been suggested as virulence factors in other protozoa, like T. vaginalis and Leishmania sp.. They act by binding and degrading various host components [17, 22]. In T. vaginalis a protein of the GP63-like group was determined to be involved in the virulence on HeLa cells [20]. The presence of the armadillo/beta-catenin-like repeat family protein in a higher amount in the clone with the highest initial virulence score also indicates that these cell adhesion molecules could be important in the pathogenesis of avian trichomonosis. Eukaryotic armadillo repeat proteins (beta-catenin-like molecules) usually play a role in cell-to-cell adhesion and cytoskeletal regulation together with intracellular signaling. They have been identified in some apicomplexan parasites, probably acting during the invasion process [35]. Malate dehydrogenase has been found more abundant in the clone with the highest initial virulence score and it could be a virulence factor since large amounts of malate dehydrogenase were found in amoeboid trophozoites of T. vaginalis bound to fibronectin, suggesting the relationship with the adherence of the parasite to the host cells [23]. Some forms of the malate dehydrogenase were also exclusively expressed in a fresh clinical isolate of T. vaginalis, (high virulence) in comparison with a long-term culture (low virulence) [18].

Although several proteases have been identified as more abundant in one or the other clone, a Clan SB family S8 subtilisin-like serine peptidase has been identified as more abundant in the clone with the highest initial virulence score. This protease has been found in higher amounts in adherent isolates of the closely related protozoan T. vaginalis [19]. Clan CA family C2 calpain-like cysteine peptidase, which is more abundant in the clone with the highest initial virulence score, has also been pointed out as a potential virulence factor in T. vaginalis [17]. On the other hand, Clan CA family C1 cathepsin L-like cysteine peptidases were more abundant in the clone with the lowest initial virulence score. Interestingly, one of these peptidases was shown to be responsible for the cytopathogenic effect of T. gallinae in permanent LMH cell cultures, however the approach (gel-based proteomics) and the analyzed protein sample (cell-free extract which corresponded to secreted proteins) employed in that study were different which might have contributed to the different result [15].

It has been proposed that T. vaginalis is able to bind to epithelial cells before transforming into an amoeboid form to increase the cell-to-cell surface contact. This fact implies a high variability of the cytoskeletal proteins [19, 23]. Our results indicate that a similar mechanism could occur in T. gallinae. Several peptides of the Ras superfamily small GTPases or components involved in their activation, like RhoGAP domain containing protein, the ARF GAP-like zinc finger containing protein; together with the Ras family protein and RhoGEF domain containing protein, have been identified as more abundant in the clone with the highest initial virulence score. In the clone with the lowest initial virulence score, only peptides mapping to two proteins related with this group were found in higher amounts: the putative small GTP-binding protein and the putative GTPase-activator protein. Overall, small GTPases and components involved in their activation are related to cell differentiation, proliferation (Ras subfamily), cytoskeletal organization, cell motility and oxidative stress (Rho subfamily), vesicle formation and regulation of vesicular trafficking (Rab, ARF and Sar subfamilies) and cell signaling in protozoa [36]. In E. histolytica, Rho GTPases, RhoGEFs and Rho effectors regulate the dynamic actin cytoskeleton and are associated with pathogenesis-related processes, such as invasion, migration, phagocytosis and evasion of the host immune response by surface receptor capping. Furthermore, Rab GTPases participate in phagocytosis as well as pinocytosis by the vesicular trafficking system and in the secretion of virulence factors, like amebapores and cysteine proteases in E. histolytica [32, 37]. In connection with the Ras superfamily proteins, it is important to note that peptides from a group of proteins involved in cytoskeleton organization, like Tctex-1 family protein or the erythrocyte membrane-associated giant protein antigen 332 containing protein, were more abundant in the membrane of the clone with the highest initial virulence score. These proteins interact with Rho subfamily proteins for regulation of the cytoskeleton re-organization. Also, two putative alpha tubulins, components of the cytoskeleton, were identified as more abundant in this clone. In the clone with delayed virulence, we only found higher amounts of peptides from two cytoskeletal proteins, in a putative profilin and a putative alpha actinin, indicating that re-organization of the cytoskeleton might be one of the mechanisms facilitating the virulence. However, more detailed studies are necessary to confirm this.

It is interesting to note that peptides from three proteins with calmodulin-like predicted function have been identified as more abundant in the clone with the highest initial virulence score. These proteins can be found in organelle membranes and play a role in the activation of phosphorylase kinases or phosphatases, necessary for cytoskeletal re-organization. Calmodulin is involved in the movement of Giardia intestinalis [38] and a calmodulin-like protein has been identified in E. histolytica playing a role in the phagocytosis of erythrocytes and cytoskeleton dynamics [39].

A group of peptides from proteins implicated in vesicular trafficking, differed in abundance between both clones. In the clone with the highest initial virulence score, peptides from a putative SNARE protein, a CMGC family protein kinase and a SacI homology domain containing protein were more abundant, whereas the opposite situation was observed with dynamin central region family protein and clathrin and vacuolar protein sorting (VPS) domain. All of these proteins are involved in endocytosis or exocytosis processes. The difference in abundance of proteins related to vesicular trafficking might be due to variations in the excretion process or feeding process between both clones including cell phagocytosis, autophagy of other trophozoites and pinocytosis of the culture medium components. Autophagy has been shown to occur in T. vaginalis cultures during the restriction of nutrients [40]. It is also noteworthy that the large group of proteins (n = 7) associated with the carbohydrate metabolism was more abundant in the clone with the lowest initial virulence score in contrast to only one protein in the clone with the highest initial virulence score. This observation can be related to the higher growth rate of the clone with the lowest initial virulence score in TYM medium as compared to the clone with the highest initial virulence score (S1 Fig).

A difference in abundance of peptides from proteins involved in plasmatic and organelles membrane transport was evident: the C2 domain containing protein and Vps52/Sac2 family protein were more abundant in the clone with the highest initial virulence score, whereas the hydrogenosomal carrier protein, the phospholipid-translocating P-type ATPase flippase family protein and the ABC transporter family protein were more abundant in the clone with the lowest initial virulence score. Interestingly, this result is in agreement with de Miguel et al [19], who identified an ABC transporter protein to be more abundant in less adherent isolates of T. vaginalis. This fact could also be related to the better growth of the clone with the lowest initial virulence score, since energy processes seem to be more active.

Conclusions

To the best of our knowledge, this is the first approach to identify and quantify membrane proteins of T. gallinae. Differences in the abundance of some peptides between the two clones were analyzed, which provides basic knowledge on disparities of strains within the same parasite species. The differences could be related to the distinct genotype, to the variations in their initial virulence or even to both characteristics. Some of the identified proteins in the present study can be potential virulence factors, similar to their orthologs in T. vaginalis and other pathogenic protozoa. Among them, we can highlight the metalloprotease GP63-like, the immuno-dominant variable surface antigen-like, the group of small GTPase Ras superfamily proteins, the calmodulins, the proteases Clan SB family S8 subtilisin-like serine peptidase and the Clan CA family C2 calpain-like cysteine peptidase. The clone with the highest initial virulence had also a higher amount of proteins related to cell adhesion, cytoskeleton organization and membrane and vesicular trafficking processes. On the contrary, the clone with the lowest initial virulence score showed higher abundance of proteins related to metabolic processes, which might be related to the better growth in the medium. These results point towards possible changes in the trophozoite shape of T. gallinae in a similar way as the human pathogen T. vaginalis.

Supporting information

S1 Fig [eps]
Growth of P178-13 C7 and R17-12 C1 clones in TYM.

S2 Fig [eps]
SDS-PAGE profile of clones.

S3 Fig [tif]
LC-MS/MS chromatograms.

S1 Table [n]
Hypothetical proteins more abundant in P178.

S2 Table [n]
Hypothetical proteins more abundant in R17.

S1 Data [xlsx]
Original data of and .