Parasite cultures

Choanomastigotes of the C. fasciculata strain LLM494 [12] were cultured at 27°C in complete medium containing RPMI 1640 medium supplemented with L-glutamine (Life Technologies, Carlsbad, CA), 10% heat inactivated foetal bovine serum (Lonza, Basel, Switzerland) and 100 µg/ml streptomycin – 100 IU/ml penicillin (Life Technologies) pH 7.2. Cell density of three replicate cultures started at 2×10⁶ cells/ml was monitored and choanomastigotes were harvested daily at 2,000 g for 10 min and washed once with PBS at 4°C.

Purification of PNA

100 g of non-roasted peanuts were submerged in 300 ml PBS at 4°C overnight, mashed and filtered through a gauze squeezing the homogenate to collect as much liquid as possible. Then, the homogenate was centrifuged at 8000 rpm for 10 min in a Sorvall RC5C centrifuge using a GSA rotor (Dupont, Stevenage, Herts, UK). The supernatant was recovered and clarified through filter paper. Then, 40% (w/v) (NH₄)SO₄ was progressively dissolved in the extract, which was then incubated at room temperature for 30 min. After 20 min of centrifugation at 6000 rpm, the supernatant was recovered and (NH₄)SO₄ progressively dissolved up to 75% (w/v). The extract was centrifuged again and the pellet recovered, resuspended in 50 ml PBS and dialyzed three times with 40 volumes of PBS at 4°C overnight. To clarify the extract, an additional centrifugation step at 6000 rpm for 10 min using an SS34 rotor and three filtration steps were carried out, the first one through filter paper, the second one through a 47 mm diameter, 2 µm pore size borosilicate fiber pre-filter (Millipore, Billerica, MA) and the third one through BioGel A2 (BioRad, Hercules, CA). Finally, the PNA was purified by affinity chromatography using melibiose immobilized in agarose beads (Sigma Aldrich, Buchs, Switzerland). Two washes with PBS were performed and the lectin eluted with 50 mM D-galactose (Sigma). The 1 ml fractions collected were quantified by the Warburg and Christian method [13].

Separation of PNA⁺ and PNA^- stationary phase choanomastigote subpopulations

Stationary phase choanomastigotes were resuspended at 2 x 10⁸ cells/ml in complete medium containing 50 µg/ml PNA in a polypropylene tube. After 30 min incubation at room temperature allowing the sedimentation of agglutination complexes, the supernatant was recovered and the sediment was resuspended again in complete medium containing PNA at the same concentration. Sedimentation of both fractions was performed again but this time at 200 g for 10 min. The resulting supernatants were mixed and centrifuged at 2,000 g for 10 min to harvest PNA^- promastigotes, whereas only the pellet obtained from the original sediment was processed as the PNA⁺ fraction. All the steps of this procedure were checked at the light microscope (63X).

Preparation and quantification of protein extracts

We described a similar proteome analysis procedure [14] and modifications are detailed herein. Each sample of parasites was resuspended in 300 µl lysis buffer (8.4 M urea, 2.4 M thiourea, 5% CHAPS, 50 mM DTT, 1% Triton X-100, 50 µg/ml DNase and Mini EDTA-free Protease Inhibitor Cocktail according to the manufacturer's instructions –Roche, Mannheim, Germany). The total protein extracts obtained were agitated by mild rotation at 4°C for 30 min, centrifuged at 8,000 g for 10 min and precipitated with methanol/chloroform [15]. All samples were dried at room temperature for 5 min and resuspended in 2X rehydration buffer (7 M urea, 2 M thiourea, 4% CHAPS, 0.003% bromophenol blue). Protein quantification was performed by the RC DC protein assay kit (BioRad). 50 µg aliquots of each sample were diluted to a final volume of 140 µl in 2X buffer containing 18.2 M DTT and 0.5% IPG buffer solution pH 3–10 (BioRad). Further confirmation was carried out by densitometric analysis of 10% PAGE-SDS gels [16] as described [14].

2DE separation and analysis of protein abundance

Isoelectrofocusing of 50 µg total protein per sample was performed on IPG strips (non-linear pH 3–10 gradient, 7 cm, BioRad) in a Protean IEF Cell system (BioRad) following the manufacturer's instructions. A seven step program was used (50 V for 12 h, 250 V for 1 h, 500 V for 1 h, 1000 V for 1 h, 2000 V for 1 h, linear ramp to 8000 V for 1 h and 8000 V up to 3500 V·h). More than a total of 12,000 V·h were reached in all runs. The second dimension was run by 12% SDS-PAGE in a pre-cooled MiniProtean 3 Dodeca Cell system (BioRad) at 0.5 W/gel for 30 min and then at 1.5 W up to 5 min after the die-front reached the bottom edge of the gels (approximately 2 h). Then, the gels were stained with SYPRO Ruby protein gel stain (BioRad) following the manufacturer's instructions. Imaging was performed with EXQuest Spot Cutter system and the analysis of differential abundance with PDQuest 2D Advanced 8.0.1 software (BioRad) following the manufacturer's instructions. First, all spots were automatically detected and thereafter manually checked by observation of single spot quantitation histograms and 2DE gel images. Normalized intensities were calculated by the Total Quantity in Valid Spots algorithm to ensure that relative quantification between gels is not biased by staining and background. The statistical analysis was performed by the Student's t-test at 0.05 significance level. Three replicates of each experiment were performed.

Protein identification by MALDI-TOF/TOF mass spectrometry

The spots selected in the previous analysis were excised with EXQuest Spot Cutter (BioRad), digested with trypsin and prepared for MALDI-TOF/TOF mass-spectrometry as we described [14]. A 0.8 µl drop of resuspended peptides from each spot was deposited in an OptiTOF Plate (Life Technologies) together with a 0.8 µl drop of a 3 g/l α-cyano-4-hydroxycinnamic solution (Sigma). The mixture was allowed to dry at room temperature. Samples were run in an ABI 4800 MALDI-TOF/TOF (Life Technologies) mass spectrometer in positive reflector mode at 25 kV for MS and 1 kV for MS/MS. The spectra were prepared with ABI 4000 Series Explorer Software 3.6 (Life Technologies). Fingerprint and fragmentation spectra were run in MASCOT 2.1 with Global Protein Server Explorer 4.9 (Life Technologies) for protein identification with both the NCBInr database and annotations on the genome sequence of C. fasciculata. The genome sequence was completed at Washington University School of Medicine in St. Louis (Stephen Beverley, Richard Wilson) and assembly and annotations at Seattle Biomedical Research Institute (Peter Myler). These data can be retrieved from http://tritrypdb.org/common/downloads/release-8.0/CfasciculataCfCl/fasta/data/. A MIAPE-compliant report and the MS data have been deposited to the ProteomeXchange Consortium [17] via the PRIDE partner repository with the dataset identifier PXD001331 and DOI 10.6019/PXD001331.

Western blot

Protein extracts were separated by SDS-PAGE in 8% slab gels (12 mA, 30 min; 30 mA, 90 min) in a MiniProtean II Cell system (BioRad). 20 µg protein extract was loaded per well including 1 µl Benzonase Nuclease HC (Novagen, Madison, WI). Blotting onto 0.45 µm nitrocellulose membranes (BioRad) was performed at 100 V for 1 h in a Mini Trans-Blot Cell wet transfer system (BioRad). Membrane blocking was carried out with 5% skimmed milk in PBS-0.1% Tween 20 (Sigma) for 1 h and washed three times with PBS-1% Tween 20 for 15, 5 and 5 min respectively. Next, membranes were incubated with 1∶500 of rabbit anti-LACK polyclonal serum for 2 h [18] or 1∶10,000 of monoclonal mouse anti-L. mexicana glycosomal GAPDH antibody kindly provided by Paul Michels (University of Edinburg) [19], washed again and incubated with 1∶2,000 HRP-conjugated goat anti-rabbit IgG (DAKO, Ely, UK) for 90 min. Once the wash steps were repeated, the immunoblots were developed using the ECL detection system (GE Healthcare, Pittsburg, PA) according to the manufacturer's instructions.

Growth kinetics of C. fasciculata choanomastigotes and 2DE-MS/MS analysis

Choanomastigote cultures reached the stationary phase within 3 days (Fig. 1B), twice as fast as Leishmania spp. promastigotes. Similar growth kinetics of C. fasciculata choanomastigote clones has been reported [20]. Total protein of 5×10⁸ choanomastigotes was extracted at early logarithmic (day 1), mid-logarithmic (day 2), late logarithmic/early stationary (day 3) and stationary phase (day 4). In addition, protein extracts were successfully obtained from the PNA⁺ and PNA^- subpopulations within the cultures in stationary phase. Protein concentrations were comprised between 4 and 9 µg/µl and this was confirmed by PAGE-SDS as described [14]. After 2DE separations, normalization with the Total Quantity in Valid Spots algorithm and manual check of all the spots, 136 changes in abundance of proteins were detected throughout the four time points of the growth curve. Some proteins showed significant differences in abundance in more than one time point of the choanomastigote growth curve. The cut-off values were: ratio to day 1, R≥1.7 or≤0.6 within the significance level inferred with Student's t test (p<0.05). Of these, 63 spots that correspond to 83 differences in abundance (Fig. 2A–D, Table 1) were excised from the 2DE gels as they were suitable to be identified by MALDI-TOF/TOF. Therefore, 10 proteins are differentially expressed at two of the time points compared. We also analyzed 43 spots containing constantly expressed proteins (Table 2), as well as 67 spots differentially expressed between the PNA^- and PNA⁺ subpopulations within the stationary phase culture (Fig. 2E and F, Table 3). All proteins could be identified when MASCOT searches were performed against the reference genome sequence of C. fasciculata (Tables 1–3) when there was sufficient amount for identification, whereas a total of 20 constantly expressed proteins (Table S2 in S1 File), 41 differentially expressed proteins in the growth curve (Table S1 in S1 File) and 30 proteins with different abundance between PNA⁺ and PNA^- choanomastigotes (Table S3 in S1 File) could be identified against the NCBInr database, which is 53.4% of the proteins analyzed by MALDI-TOF/TOF mass spectometry. Most of the identifications (73.9%) were consistent between the two databases and most of those successfully performed with the NCBInr database (63.9%) had the highest MASCOT scores for orthologue proteins of the genus Leishmania, whereas only 9.3% of them matched with a Trypanosoma spp. orthologue (Tables S1–S3 in S1 File). This is additional evidence for the hypothesis supporting very close evolutionary relationship between Leishmania spp. and Crithidia spp [6], [7], [8], [9]. Also, only 9.3% matched with Crithidia spp., as very few genes had been identified in this organism prior to the release of the reference genome sequence (Tables S1–S3 in S1 File). Sometimes, different spots represent the same type of protein. This may be due to the presence of different isoforms, post-translational modifications or protein aggregation at the conditions assayed. In the next sections, we refer to these possibilities using the term variant of a given protein.

Download:

• PPT
  PowerPoint slide
• PNG
  larger image
• TIFF
  original image

Figure 2. 2DE of total protein extracts throughout the growth curve and of the PNA⁺ and PNA^- subpopulations of C. fasciculata choanomastigotes.

2DE of 50 µg of total protein extracts of C. fasciculata choanomastigotes at (A) early logarithmic, (B) mid logarithmic, (C) late logarithmic and (D) stationary phase. (E) PNA^- and (F) PNA⁺ subpopulations. One out of three replicates is shown for each phase/subpopulation. IEF was performed in a non-linear 3–10 pH interval. Complete spot names include Cf (A–D) or Cf_p (E and F) preceding the spot numbers (see Tables 1–3).

https://doi.org/10.1371/journal.pone.0113837.g002

Download:

• PPT
  PowerPoint slide
• PNG
  larger image
• TIFF
  original image

Table 1. Differentially regulated proteins throughout the growth curve of C. fasciculata choanomastigotes.

https://doi.org/10.1371/journal.pone.0113837.t001

Download:

• PPT
  PowerPoint slide
• PNG
  larger image
• TIFF
  original image

Table 2. Constantly expressed proteins throughout the growth curve of C. fasciculata choanomastigotes.

https://doi.org/10.1371/journal.pone.0113837.t002

Download:

• PPT
  PowerPoint slide
• PNG
  larger image
• TIFF
  original image

Table 3. Differential abundance of proteins between the PNA⁺ and PNA^- subpopulations of C. fasciculata choanomastigotes in stationary phase of axenic culture.

https://doi.org/10.1371/journal.pone.0113837.t003

Peanut lectin agglutination capability of choanomastigotes

It is known that certain species of Leishmania agglutinate with PNA and a non-agglutinating subpopulation can be isolated [10], [11]. This is the first time that the agglutination properties of Crithidia spp. choanomastigotes with PNA have been assayed and the outcome has been a noticeable agglutination capability and the isolation of a non-agglutinating population in stationary phase. These findings support a modification of the Crithidia spp. lipoarabinogalactans (LAG) at the end of the growth curve analogous to that taking place in the lipophosphoglycan (LPG) of Leishmania spp., even when the structure of the LAG [21], the surface molecules that presumably agglutinate with the lectin in this genus, is quite different to the leishmanial LPG. The biological roles of the respective agglutinating surface molecules involved are probably different in these organisms given the differences in their life cycles. L. major and L. infantum promastigotes are able to agglutinate with PNA and the non-agglutinating subpopulations are more infective and lead to higher infection rates than the agglutinating ones, yielding more infected phagocytes and amastigotes per infected cell [10], [11]. Promastigotes attach to the gut epithelium by the LPG to maintain infection during bloodmeal excretion and only with differentiation signals as starvation, a developmental process ultimately leading to metacyclic promastigotes is triggered (reviewed in [22]). The differentiation process is widely recognized to be mimicked in axenic culture, where starvation also takes place when promastigotes reach the stationary phase.

The change in LAG composition of C. fasciculata is likely due to developmental processes according to the findings described herein (see below) but does not provide evidence to be associated to any process related with infectivity. In fact, the next step in the life cycle of this parasite is differentiation to the extracellular amastigote stage that attaches to the gut epithelium of the insect host.

Changes in abundance of proteins involved in glucid catabolism and the pentose phosphate pathway

According to proteome profiling, glycolysis is more active in early and mid logarithmic phase (Fig. 3, Table 1), when several protein variants alternate. In fact, two aldolase (ALD) and two enolase variants are up-regulated at day 1 but at the second day, their expression levels decay and are replaced by distinct ones, respectively one ALD and two enolases. Additionally, an hexokinase (HK) variant, the pyruvate kinase (PyrK), a putative and a glycosomal malate dehydrogenase (MDH) and the components of the pyruvate dehydrogenase complex (PDH) dihydrolipoamide dehydrogenase (DLD) and E1α are more abundant at day 2 (mid logarithmic phase), which suggests higher activity of hexose catabolic processes and malate shuttles, provided that most of the glucolytic reactions take place in the glycosome of these organisms (reviewed in [23]). In fact, other monosaccharides may be increasingly utilized by choanomastigotes in mid logarithmic phase as additional carbon and energy sources and/or to provide precursors for the biosynthesis of glycans. This is suggested on the basis of the wide substrate specificity of the HK and the up-regulation of the phosphomannomutase (PMM) and the aldose 1-epimerase (AEP). The PMM is involved in the biosynthesis of N-glycans providing manose-1-phosphate, as the reaction is reversible. The AEP is also up-regulated at the stationary phase, especially in PNA^- choanomastigotes (Tables 1 and 3). As highlighted in Fig. 3, the up-regulation of these ALD variants may yield high levels of glyceraldehyde-3-phosphate not only for the subsequent glucolytic reactions but also for the pentose-phosphate shunt, provided the up-regulation of the transaldolase B (TALDO) at early logarithmic phase (day 1) and the transketolase (TKETO) at mid logarithmic and stationary phase. These proteins are related functionally with the phosphoribosyl pyrophosphate synthase (PRPPS), which is more abundant at day 2. These findings are indicative of maximum activity of the glucolytic pathway at mid logarithmic phase providing energy and essential precursors of certain amino acids, ribonucleotides and derived coenzymes. By contrast, the highest expression levels of genes involved in glucolysis are found in L. infantum promastigotes in stationary phase [24] and differential regulation of genes involved in the pentose phosphate shunt has not been detected up to date in these pathogenic trypanosomatids.

Download:

• PPT
  PowerPoint slide
• PNG
  larger image
• TIFF
  original image

Figure 3. Differentially expressed proteins related with carbohydrate metabolic processes in C. fasciculata choanomastigotes.

Functional connection of differentially expressed enzymes involved in glucid metabolic processes. Legend: proteins/protein variants in blue are constitutively expressed throughout the growth curve; proteins/protein variants in orange are up-regulated at day 1 or 2 (logarithmic phase); proteins/variants in green are up-regulated at day 3 or 4 (late logarithmic/stationary phase); proteins in red are up-regulated in PNA⁺ choanomastigotes; proteins in purple are up-regulated in PNA^- choanomastigotes.

https://doi.org/10.1371/journal.pone.0113837.g003

The up-regulation of two alcohol dehydrogenases (ADH) at early logarithmic phase is also related with sugar catabolism, as a relative inefficiency of the respiratory process and the finding of subproducts as ethanol and lactate were described in trypanosomatids (reviewed in [25]). Alternation in up-regulation between ADH variants has been detected between PNA⁺ and PNA^- choanomastigotes (Table 3), and there are also constitutive and differentially regulated variants of ADH and MDH.

In certain cases, only proteins that catalyze irreversible reactions and/or control the kinetics of the metabolic process are increasingly abundant, whereas those catalyzing reversible and non-limiting steps are constitutively expressed. This is the case of the expression profile of glycolytic enzymes in C. fasciculata choanomastigotes as the triose phosphate isomerase (TPI) and the phosphoglycerate kinase (PGK) (Fig. 3, Table 2) are constitutively expressed unlike others mentioned above. Conversely, an L. infantum orthologue of the PGK reaches its highest abundance in stationary phase promastigotes at the transcript level [24]. The aconitase is up-regulated in the promastigote stage with respect to amastigotes in L. infantum [26] and the opposite profile was observed in L. donovani [27], but the C. fasciculata orthologue is constitutively expressed in the choanomastigote stage.

Several enzymes related with sugar metabolism are also differentially regulated between PNA⁺ and PNA^- choanomastigotes within the stationary phase. The PGM^BPI and the enolase are over-expressed, suggesting that only the second part of glycolysis is more active in PNA^- choanomastigotes in a bisphosphoglycerate-independent manner. As a difference with mammalian organisms, the trypanosomatid genomes encode the PGM^BPI, which suggests that this protein is a good drug target [28]. Interestingly, this enzyme is differentially regulated between different life cycle stages in L. infantum [24] and also in C. fasciculata according to the analysis described herein. The expression patterns of the HK, ALD and TKETO are the opposite to those of the PGM^BPI and the enolase. Consequently, the expression profile of the glycolytic and the pentose phosphate pathway are the same in the PNA⁺ subpopulation and the whole population in stationary phase (Table 3). The finding is coherent as this is the major subpopulation at that growth phase.

The heme biosynthetic enzyme coproporphyrinogen (III) oxidase (C(III)O) is up-regulated in PNA^- C. fasciculata choanomastigotes (Table 3) whereas it is constitutively expressed throughout the growth curve (Table 2). Differential expression was not observed during axenic growth of L. infantum promastigotes either and it was revealed that C(III)O is up-regulated in intracellular amastigotes [24] and axenic amastigotes obtained by temperature and pH shift [29]. The heme group is necessary for a variety of cellular functions in Leishmania spp., including not only the electron transport chain, but also the catalase, which is also over-expressed in the PNA^- subpopulation (see below).

Changes in abundance of proteins involved in lipid metabolism

Fatty acid biosynthesis, ketone body degradation, β-oxidation of fatty acids and/or branched chain amino acid degradation is probably more activated at day 2 because a putative 3-ketoacid-CoA thiolase (KAT) is up-regulated. This may be linked with the up-regulation of the succinyl-CoA:3-ketoacid-CoA transferase (SCAT), which suggests the utilization of ketone bodies in C. fasciculata mid logarithmic phase choanomastigotes, which is of unknown meaning in trypanosomatids. The existence of ketone bodies in trypanosomatids was tested by ¹H NMR in L. donovani axenic amastigotes [30]. These molecules are probably mere intermediate metabolites in these organisms. Given that the SCAT catalyzes the split of acetoacetate into two acetyl-CoA molecules, a probable explanation for its up-regulation may be that it is involved in the last step of Leu degradation, as well as KAT up-regulation suggests a role in Ile catabolism and the DHL in degradation of all branched chain amino acids. Conversely, thiolases reach their highest expression levels in stationary phase promastigotes in L. infantum [24]. The SCAT is down-regulated in the PNA^- choanomastigote subpopulation, which indicates a decrease throughout choanomastigote development taken together with the results obtained for the growth curve.

Changes in abundance of proteins involved in gene expression regulation and signal transduction

A decreased translational elongation rate is expected throughout the growth curve of choanomastigotes, as the abundance of the eukaryotic translation initiation factor 5a (eIF5a), the translation elongation factor 1β (EF1β) and the elongation factor 2 (EF2) decreases. This is likely due to the higher metabolic activity and faster growth in logarithmic phase choanomastigotes. However, other proteins involved in gene expression regulation that are differentially abundant in L. infantum promastigotes [24], [31] are constitutively expressed conversely in C. fasciculata choanomastigotes: the EF1α, a poly(A) binding protein (PABP2), two RNA-binding proteins (RNA bp) and an ATP-dependent RNA helicase (Table 2). These proteins may be involved in developmental processes rather than in growth in Leishmania spp. promastigotes. A similar hypothesis can be posed to explain the up-regulation of the eIF5a, the EF1α, the EF2 and the endoribonuclease L-PSP in the PNA^- choanomastigotes, as this minor subpopulation was isolated from culture in stationary phase, where growth conditions would not explain an increase on the translation rate. The transcript levels of PABP are also higher in L. infantum PNA^- promastigotes, but those of the EF1α were significantly lower, thus being the expression profile the opposite [11]. In fact, three variants of the EF1α protein are more abundant in stationary phase promastigotes, where the PNA⁺ promastigote subpopulation is by far the more represented [14]. Future in depth analysis of these differences between both parasites may aid to explain why their developmental processes are different.

As for protein folding, only a variant of the hsp60 chaperonin is more abundant in early logarithmic phase choanomastigotes, whereas other variant of this protein, a DnaJ domain-containing protein, two cyclophilins and the adenosine kinase domain-containing nucleoside diphosphate kinase b (Ndkb) are constitutively expressed throughout the growth curve of axenic choanomastigotes. These findings also contrast with results found in L. infantum promastigotes [11], [14], [24], [31]. To put an example, a cyclophilin is up-regulated in stationary phase promastigotes in L. infantum.

Regarding intracellular signalling, the C. fasciculata analogue (CACK) (gi3132790; CfaC1_26_3810) of the receptor of the activated protein kinase C (RACK) is up-regulated in early logarithmic phase choanomastigotes (day 1) (Fig. 2, Table 1). This expression profile has been confirmed by Western blot (Fig. 4, Figures S1 and S2 in S2 File), which has revealed the progressive descent of CACK abundance throughout the choanomastigote growth curve. The leishmanial orthologue LACK, an antigenic protein that partially protects against canine leishmaniasis [32], [33], [34], is located in the particulate fraction of the cytoplasm near the plasma membrane. LACK is up-regulated in L. infantum amastigotes but constantly expressed in promastigotes [18], [24]. Consequently, the CACK/LACK expression patterns in the motile stages of C. fasciculata and L. infantum are different, what suggests different roles in proliferation and differentiation in their respective life cycles. PKCs are translocated by their receptor (RACK) to different intracellular sites [35] and activated via phospholipase C or Ca²⁺. PKCs and RACKs are involved in a variety of characterized signal transduction cascades in mammals [36]. However, their role in specific pathways is unknown in these parasites. Although the kinomes of trypanosomatids are well characterized [37], the signaling pathways may not be necessarily the same as in other organisms like yeasts and mammals. Signaling proteins are expected to regulate gene expression but most of the specific mechanisms and pathways have not been unraveled yet and may be different given the unique features of gene expression in these parasites (reviewed in [38]). The unknown specific function of this protein in signaling may be especially important in Leishmania spp. for resistance of amastigotes inside the parasitophorous vacuole of the host phagocyte cell whereas the only colonization step of the Crithidia spp. life cycle is the infection of the gut of the insect host. CACK and LACK seem to be important for growth and/or development of the motile stages of the respective species they belong to, as up-regulation in C. fasciculata logarithmic phase choanomastigotes has been found herein and its leishmanial orthologue is one of the 50 most abundant transcripts of L. major promastigotes [39].

Download:

• PPT
  PowerPoint slide
• PNG
  larger image
• TIFF
  original image

Figure 4. CACK is up-regulated in logarithmic phase choanomastigotes.

Detection and differential expression analysis of CACK in 20 µg total protein extracts by Western blot with 1∶500 polyclonal antibody against the LACK analogue throughout the choanomastigote growth curve. The ∼60 KDa band presumably contains dimeric CACK aggregates (González-Aseguinolaza et al., 1999). gGAPDH is the protein of reference (dilution 1∶10,000 of the monoclonal antibody).

https://doi.org/10.1371/journal.pone.0113837.g004

Changes in abundance of proteins involved in thiol-based redox homeostasis

The TryP of C. fasciculata previously characterized [40] (gi3851500; CfaCl_10_1430) is over-expressed in mid logarithmic phase choanomastigotes, as well as the thiol-dependent reductase 1 (TDR1) (Table 1, Fig. 5). This is an important difference with promastigotes of L. major, an ethiological agent of cutaneous leishmaniasis in the Old World, as TryP is constitutively expressed in all the stages of its life cycle [41], [42]. It is also known that L. donovani amastigotes up-regulate the TryP with respect to promastigotes [27]. The catalase is absent in pathogenic trypanosomatids [43] but not in C. fasciculata. In fact, variants matching with the annotations CfaCl_30_0050 have been identified in several spots (Tables 1 and 3). For this reason, hydrogen peroxide removal in C. fasciculata is not necessarily dependent on trypanothione-linked peroxidases, as a difference with the pathogenic trypanosomatids. Like the TryP, this protein is more abundant at mid logarithmic phase. These data suggest higher levels of oxidative stress counteracted with TDR1, TryP and catalase up-regulation in Crithidia spp. choanomastigotes than in Leishmania spp. promastigotes at mid logarithmic phase, possibly due to the greater growth rate observed in the former (Fig. 1). In fact, the stationary phase is reached about 5–7 days in a typical growth curve of Leishmania spp. (e.g., [24]). The up-regulation of TDR1 suggests that the glutathione-ascorbate cycle is also participating in counteracting oxidative stress.

Download:

• PPT
  PowerPoint slide
• PNG
  larger image
• TIFF
  original image

Figure 5. Differentially expressed proteins involved in redox homeostasis and translation in C. fasciculata choanomastigotes.

(A) The redox control system. Legend: proteins/protein variants in blue are constitutively expressed throughout the growth curve; proteins/protein variants in red are up-regulated at day 1 or 2 (logarithmic phase); proteins in purple are up-regulated in PNA^- choanomastigotes. (B) Summary of differential abundance of translation factors and enzymes involved in redox homeostasis throughout the growth curve and in the PNA^- subpopulation of choanomastigotes in stationary phase.

https://doi.org/10.1371/journal.pone.0113837.g005

The analysis of the PNA⁺ and PNA^- subpopulations within the stationary phase has revealed that the entire redox defense system is up-regulated in the latter (Table 3). This includes the catalase (three variants), the iron superoxide dismutase (Fe-SOD) (two variants), a hypothetical protein (CfaC1_05_0410) orthologue to the trypanothione reductases (TryR) of the pathogenic trypanosomatids and two TryP variants. Therefore, the thiol-based redox defense system is over-expressed in an NADPH-dependent manner in PNA^- choanomastigotes. These findings taken together with the expression profile observed throughout the growth curve suggest that the differentiation process of C. fasciculata motile choanomastigotes involves an increase in oxidative stress in the axenic culture model that is overcome by the up-regulation of this defense system. These changes have not been observed at the transcript level in L. infantum PNA^- promastigotes within the stationary phase [11], as well as at the whole stationary phase population at the protein level in L. infantum [14] and in L. major [41], [42]. The Fe-SOD reduces the superoxide anions generated in the ribonucleotide reductase (RNR) reaction and other reactions to hydrogen peroxide. Then, the catalase and the TryP, catalyze hydrogen peroxide oxidation and peroxidative reactions (reviewed in [44]). The thiol-dependent mechanism also reinforces the reactive oxygen species (ROS) reduction (Fig. 5). This system was characterized in C. fasciculata and consists of the enzymes TryR and TryP, also including the tryparedoxin (TryX[S]/[SH₂]). The trypanothione (T[S]/[SH₂]) (reviewed in [45]). The T[S]/[SH₂] consists of two glutathione residues coupled through a spermidine molecule. The peroxiredoxin TryP acts as a catalase and also reducing a variety of other ROS and the TryR regenerates the T[SH₂] (reduced form) to T[S] (oxidized form).

To summarize, the thiol-based redox control system is over-expressed in mid logarithmic phase C. fasciculata choanomastigotes and at the PNA^- subpopulation within the stationary phase. The faster growth kinetics compared to Leishmania spp. may be related to the higher levels of oxidative stress overcome by the up-regulation of TDR1, TryP and the catalase in C. fasciculata rather than in Leishmania spp. The up-regulation of the TryR-TryP system together with the catalase and the Fe-SOD in PNA^- choanomastigotes suggests a relationship between differentiation and the capability to overcome increased levels of oxidative stress.

Differences in the proteome profiles of the motile stages of C. fasciculata and Leishmania spp

Stage-specific regulation at the transcript and protein levels in the genus Leishmania has been widely studied. Much of the information extracted is related to differentially regulated genes between amastigotes and promastigotes [27], [29], [41], [46], [47] but few analyses have provided data about differential expression throughout the promastigote growth curve [48], [49], [50]. A high-throughput transcriptome analysis was specifically focused on the differences between logarithmic and stationary phase promastigotes with amastigotes in L. infantum [24]. Comparing the proteomic data sets available for the motile stages of L. infantum [14] and C. fasciculata (this work), an important difference has been found, the up-regulation of the TryR in logarithmic phase promastigotes in the former (this gene was reported as thiol-dependent antioxidant protein, gi21307665 in the NCBI protein databank and corresponds to gene LinJ.05.0350) vs. the PNA^- subpopulation in C. fasciculata.

Important differences with Leishmania spp. can be noticed taking the information extracted from the transcriptome and the proteome as a whole. Indeed, regarding metabolic processes, the expression profiles of glucolytic enzymes and the thiolase between the motile stages of L. infantum and C. fasciculata are different. Glucolysis is active in Leishmania spp. promastigotes and probably reaches its highest activity in stationary phase, whereas expression of many of the glucolytic enzymes decays in stationary phase choanomastigotes in C. fasciculata, being increased only in PNA^- parasites.

As a difference with the L. major TryP [40], the C. fasciculata orthologue does not maintain constant expression levels in the motile stage, especially in the PNA^- subpopulation of choanomastigotes in stationary phase, where the whole thiol-based redox defense system (catalase, Fe-SOD, TryP, TryR) is up-regulated. In fact, differential expression of the TryR, the TryP and the Fe-SOD was not found in L. major and L. infantum metacyclic promastigotes, whereas the tryparedoxin levels vary between procyclic and metacyclic promastigotes in both species [11], [48], [49]. Additionally, up-regulation of the TryR observed in L. infantum logarithmic phase promastigotes has been precisely the only one not found in C. fasciculata choanomastigotes, where TDR1 is more abundant instead. The oxidative phase of the pentose-phosphate shunt is the source of NADPH for these important cellular processes and it is constitutively expressed in both organisms probably. However, the non-oxidative phase of this pathway is over-expressed in logarithmic phase only in C. fasciculata. The plethora of changes in gene expression regulation and post-translational modification of proteins observed previously in L. infantum promastigotes has not been observed in C. fasciculata choanomastigotes. Indeed, only the eIF5a, the EF1β, the EF2 and the hsp60 are up-regulated in logarithmic phase choanomastigotes (Table 1) whereas the PABP, two RNAbp, two variants of the EF1α, an ATP-dependent RNA helicase, a different hsp60 variant, the Nkdb, a DnaJ chaperone and a peptidyl-prolyl cis-trans isomerase and two cyclophilins are constitutively expressed throughout the growth curve (Table 2) according to this analysis. A different profile of EF2 expression was observed in L. major, where it is over-expressed in metacyclic promastigotes [48]. Interestingly, the EF1α is more abundant in PNA^- choanomastigotes of C. fasciculata, what is opposite to PNA^- promastigotes of L. infantum, where it decreases. This may be linked with differences between the developmental processes these parasites undergo instead of being related to growth in a rich medium and contributes to explain in part the differences between the life cycles of C. fasciculata and Leishmania spp. Promastigotes undergo a deep differentiation process inside the gut of the phlebotomine sand fly to successfully establish intracellular infection in the mammalian host. By contrast, the life cycle of Crithidia spp. is monogenetic and differentiation of the characteristic choanomastigote motile stage is probably not so complex because the amastigote stage is not intracellular. This assumption is suggested by the substantial differences found in the abundance of proteins involved in gene expression regulation and protein modification between both genera. However, both organisms are close in the crown of the phylogenetic tree of trypanosomatids and growth kinetics of their motile stages in culture is similar. For this reason, future in depth study of these differences in protein abundance may help to explain their different mechanisms of adaptation to axenic cultures that mimic the conditions of the insect host gut. The MASCOT identifications performed against the NCBInr database (Tables S1-S3 in S1 File) provide additional evidence for the hypothesis of a very close relationship in the evolutionary tree. The increased abundance of the translation factors eIF5a, EF1β the EF2 in early logarithmic phase choanomastigotes may be linked to a more active metabolic status required for growth, whereas the increase in eIF5a, EF1α and EF2 in the PNA^- choanomastigote subpopulation in stationary phase suggests a role in development, as it cannot be associated to growth under starvation conditions (Fig. 5B). The same has been observed with the redox homeostasis control system. Logarithmic phase C. fasciculata choanomastigotes up-regulate the ascorbate-dependent TDR1, as well as the TryP, the catalase and the Fe-SOD, probably because of an increased oxidative stress resulting from faster growth kinetics than in Leishmania spp., and the abundance of these enzymes decay in the ongoing of the growth curve. Then, it increases again in the PNA^- subpopulation within the stationary phase (TryR instead of TDR1), being presumably associated to development at this stage as well (Fig. 5B). It is important to notice that the expression profiles described correspond to the motile stages of these parasites in axenic culture and the developmental processes of Crithidia spp. are almost unexplored. The axenic culture model is widely accepted in Leishmania spp. because cultured promastigotes are able to stablish infection in mammalian hosts, what suggests that the main developmental processes may be reproduced in culture. In the case of Crithidia spp., it is not known whether culture greatly affects development. The main findings of this study point to similar but faster growth kinetics in C. fasciculata and different metabolic adaptations to the same culture medium conditions.