Trypanosoma cruzi: Gene Expression Surveyed by Proteomic Analysis Reveals Interaction between Different Genotypes in Mixed In Vitro Cultures

We have analyzed the comportment in in vitro culture of 2 different genotypes of Trypanosoma cruzi, the agent of Chagas disease, pertaining to 2 major genetic subdivisions (near-clades) of this parasite. One of the stocks was a fast-growing one, highly virulent in mice, while the other one was slow- growing, mildly virulent in mice. The working hypothesis was that mixtures of genotypes interact, a pattern that has been observed by us in empirical experimental studies. Genotype mixtures were followed every 7 days and characterized by the DIGE technology of proteomic analysis. Proteic spots of interest were characterized by the SAMESPOT software. Patterns were compared to those of pure genotypes that were also evaluated every 7 days. One hundred and three spots exhibited changes in time by comparison with T = 0. The major part of these spots (58%) exhibited an under-expression pattern by comparison with the pure genotypes. 32% of the spots wereover-expressed; 10% of spots were not different from those of pure genotypes. Interestingly, interaction started a few minutes after the mixtures were performed. We have retained 43 different proteins that clearly exhibited either under- or over-expression. Proteins showing interaction were characterized by mass spectrometry (MALDI-TOF). Close to 50% of them were either tubulins or heat shock proteins. This study confirms that mixed genotypes of T. cruzi interact at the molecular level. This is of great interest because mixtures of genotypes are very frequent in Chagas natural cycles, both in insect vectors and in mammalian hosts, and may play an important role in the transmission and severity of Chagas disease. The methodology proposed here is potentially applicable to any micropathogen, including fungi, bacteria and viruses. It should be of great interest in the case of bacteria, for which the epidemiological and clinical consequences of mixed infections could be underestimated.


Introduction
Chagas disease is a parasitic infection caused by Trypanosoma cruzi. Although control has improved, Chagas disease remains a serious public health problem in most Latin American countries. It is responsible for blood transfusion accidents in the United States, and it is now spreading in Europe, particularly in Spain. Available treatments are toxic and their efficiency in the chronic phase of the disease is limited. Vaccines are not available [1]. Basic research at the level of molecular mechanisms involved in parasitic infection is therefore needed to help design effective treatments and controls.
As a result of predominant clonal evolution (PCE), with occasional bouts of genetic recombination [2][3][4][5], T. cruzi natural populations are distributed into 6 major genetic subdivisions [6] or ''near-clades'' [4,5]. Mixtures of genotypes pertaining to distinct near-clades are frequent in natural Chagas cycles, in the insect vector [7] as well as in the human host [8]. It has been proposed that such mixtures could play a major role in the transmission and severity of Chagas disease [9] and of other infectious diseases [9,10].
Interactions between different T. cruzi genotypes have been evidenced in experimental infections of vectors [11] and laboratory mice [12]. However, these studies did not explore the molecular mechanisms involved in mixed infection interactions.
A protocol of gene expression characterization by proteomic analysis has been developed by us [13] and has made it possible to show that T. cruzi near-clades exhibit specific gene expression patterns.
In the present study, we have applied this protocol to the analysis of the interaction in in vitro cultures between two T. cruzi genotypes pertaining to distinct near-clades.

Results
With the analysis of 2D-DIGE gels with the Progenesis SameSpots 3.1 software (Nonlinear Dynamics), we found that 103 spots exhibited change over time and were considered for further analysis. The 40 spots retained according to the 2 criteria exposed in material and methods, are shown on figure 1. Figure 2 shows 2 illustrative examples of spots of interest. A majority (58%) of the spots of interest exhibited an under-expression by comparison with the fastest genotype, while a minority (32%) showed over-expression. Figure 3 shows different patterns of protein expression kinetics, either over-or under-expression, taking as reference the dominant genotype.
The 40 spots of interest corresponded to 43 different proteins. Some spots corresponded to 2 different proteins, as evidenced by mass spectrometry. Table 1 gives the list of protein categories identified by mass spectrometry in the spots of interest and indicates whether the proteins are either over-expressed or under-expressed.
Numerous different kinds of proteins are included. The main categories are proteolysis proteins, metabolic proteins, biosynthesis proteins, structural proteins and stress proteins. However, the 2 dominant categories, which comprised about 50% of the total, were tubulins and heat shock proteins (HSPs).
It is interesting to note that some strong differences from the theoretical T = 0 (average expression of each of the 2 pure genotypes) were observed as soon as 10 mn after the actual mixing, which suggests that interaction between the 2 genotypes starts very early. This was evidenced in 76 spots out of the 103 spots that showed changing over time. Out of these 76 spots, 12 (16%) exhibited over-expression, while 64 (84%) exhibited underexpression a few minutes after mixing by comparison with the theoretical T = 0 value.

Discussion
The goal of this study was to explore whether gene expression of T. cruzi genotype mixtures is different from a mere juxtaposition of the gene expression of pure genotypes. The working hypothesis was that mixed genotypes interact and exchange biochemical signals. This hypothesis is important, since mixtures of genotypes are frequent in natural Chagas cycles, both in the host [8] and in  the vector [7]. Evidence of interaction in infectious disease natural cycles has been found in other pathogens. In fungi, mixtures of closely related genotypes of the same species lead to more successful transmission [14]. Mixed infections of Plasmodium falciparum, the agent of the most malignant form of malaria, lead to less frequent fever episodes in young subjects [15].
The present in vitro experiment dealt with 2 stocks that correspond to 2 major T. cruzi near-clades [4,5]. They were selected for 3 reasons: (i) they represent « polar genotypes », phylogenetically distantly related [6]. Accordingly, their gene expression is expected to be neatly different from each other, as evidenced by proteomic analysis [13]; (ii) mixtures of these genotypes are very frequent in Chagas natural cycles [7,8]; (iii) their behavior in in vitro cultures is quite different : P 209 cl1 grows much faster and is more virulent in mice than Mn cl2 [16,17].
The present study made it possible to identify 43 different proteins that exhibit either over-expression or, more frequently, under-expression in mixed genotypes. This shows that the genotypes have interacted in these mixtures, through either reciprocal stimulation (over-expression) or inhibition (underexpression). These interactions start very early, a few minutes after the mixture was performed.
Our working hypothesis is that mixtures of genotypes play an important role in the transmission and severity of Chagas disease, through hypothetical mechanisms, which we have called « clone hitchhiking » [9]. Other authors have proposed that T. cruzi nearclades could have different histotropisms in the same host [18].
We do not have reliable hypotheses right now to explain why the proteins identified in the present study exhibit modified expression in genotype mixtures. The 2 categories of proteins that were most frequently identified in our study were heat shock proteins (HSPs) and tubulins. HSPs are stress proteins and their presence here could be attributed to the culture process itself. As a matter of fact, in vitro cultures are far from natural conditions for the parasite and obviously are a stress factor for it. This does not explain why HSPs are either over-or under-expressed in mixtures by comparison with pure genotypes, especially the fastest growing genotype P 209 cl1, which, according to previous experiments, should be overdominant in the mixtures. Tubulins are also either over-or under-expressed in mixtures. They are structural proteins of the microtubules, a major component of the cytoskeleton.
Tubulins are highly involved in the motility of the flagellum [19]. It might be the case that clones interfere at the level of their motility, either to inhibit each other or to stimulate each other, for accessing nutriments.
Over-and under-expression are also observed for most proteins identified in the present study (table 1). It seems to be the case that mixture of genotypes leads to either stimulation or inhibition.
Work is in hand to explore further these phenomena through more complete collections of genotypes in in vitro and cell cultures, and in animal models. Further experiments with in vitro cultures will seek to isolate from the medium itself the hypothetical biochemical signals which mediate interactions.
We suggest that mixtures of genotypes, which have been thoroughly evidenced in Chagas natural cycles, occur also in other parasites, as well as in other pathogens [9,10], including fungi, bacteria, and viruses. Mixtures of genotypes might play a general role in the transmission and severity of infectious diseases through clonal « hitchhiking » and cooperation [9]. The experiments reported here can potentially be applied to any pathogen to test this working hypothesis, provided that experiments are feasible. Such experiments could be especially relevant in the case of pathogenic bacteria, for which the phenomenon of mixed infection and its epidemiological and medical implications may have been underestimated.

Trypanosoma Cruzi Stocks
Two stocks were used in this study, namely P209 cl1 and Mn cl2. They belong to the T. cruzi genetic subdivisions or near-clades TC I, and TC V, respectively [6]. They have been cloned in the laboratory, with verification of the cloning under the microscope. Their genotype has been verified at the beginning of the study by Multilocus Enzyme Electrophoresis (MLEE) and Random Amplified Polymorphic DNA (RAPD). They have been selected for the following reasons: (i) TC I and V correspond to radically distinct groups of T. cruzi genotypes; (ii) mixtures of these genotypes are frequent in Chagas natural cycles [7,8]; (iii) they exhibit very distinct phenotypes, since P209 cl1 has a faster growth in vitro cultures, and a higher virulence and parasitemia in mice than Mn cl2 [16,17].  Epimastigote stages were cultured in LIT (Liver Infusion Tryptose) with 10% fetal calf serum, in 15 ml tubes containing 1 ml of parasites for 3 ml of culture medium. Cultures were done at 27uC. Parasites were passaged every 7 days. About 2610 6 cells were passaged in 3 ml of fresh culture medium.

Bulk Cultures for Proteomic Analysis
Plastic flasks of 75 cm 3 were used for bulk cultures. After growth, 30 ml of parasite culture were transferred into 50 mlcentrifugation tubes and centrifuged at 1,500 g for 15 mn at 20uC. The supernatants were discarded and the pellets were resuspended in 10 ml of PBS (Phosphate Buffered Saline). Then the tubes were again centrifuged at 1,500 g for 15 mn at 20uC. The supernatants were discarded and the pellets were resuspended in 1 ml of PBS and transferred into 1.5 ml eppendorf tubes. The eppendorf tubes were centrifuged at 1,500 g for 10 mn at 20uC. Supernatants were discarded. Pellets were weighted and stored at 280uC until use.

Protocol for Analysis of Mixed Stocks
Mixtures of 50% of each stock were used. Mixtures were analyzed at T = 0, T = 7 days and T = 14 days. Pure genotypes were also analyzed at T = 0, T = 7 days and T = 14 days. At T = 7 and 14, the patterns of the mixtures were compared to the patterns of P 209 cl1, which is the fastest growing strain. According to previous mixture experiments analyzed by isoenzymes and random primer amplified polymorphic DNA, P 209 cl1 becomes very fastly overdominant on MN cl2 (unpublished data). Patterns of the mixtures, if they were no interactions, should therefore be similar to the patterns of P 209 cl1. Mixtures were prepared by counting parasite cells with a Thoma chamber. T = 0 corresponds to the moment when the mixture is done. The theoretical value taken for the mixture at T = 0 for proteomic analysis is the average value of the 2 genotypes considered separately. However, as we will expose further, interaction between the genotypes starts as early as a few minutes after the mixture is made. We, therefore, give the proteomic results of the mixture a few minutes after T = 0.
Three replicates were performed for each measure.
Protein extraction. Parasite pellets were resuspended into 500 ml of a solubilization buffer solution, then the parasite cell walls were lysed by immersion in liquid nitrogen 3 times for a few seconds each time. Benzonase Nuclease was then added in order to degrade DNA and RNA in the samples. Samples were then centrifuged at 20,000 g for 15 mn at 4uC. PH was then adjusted at 8.5.
According to the manufacturer's recommendations, 50 mg of proteins were labeled with 400 pmol of marker. After labeling, samples were kept in the dark for 30 mn. Reaction was then stopped by saturating the solution with 1 ml of lysine. Samples were then put again in the dark for 15 mn.
Protein separation by 2-dimensional electrophoresis (2D-PAGE). (i)1st dimension: separation by isoelectrofocalization (IEF). IEF separates proteins according to their isoelectric point (IP). IEF separation was performed on strips (IPG-strips, Amersham) with a PH gradient of 4 to 7. Then, the strips were rehydrated with 450 ml of rehydration buffer covered with mineral oil to avoid dehydration (Drystrip Cover Fluid, Amersham). Proteins were then isoelectrofocused at 30V for 3 h, then at 30V to 1000 V for 5 h, then from 1000V to 8,000 V for 4 h, and finally to 8,000 V for 5 h before coming back to 30 V again. (ii) 2 nd dimension: SDS-PAGE (Sodium Dodecyl Sulfate Polyacrylamide Gel Electrophoresis). SDS-PAGE separates proteins according to their molecular weight (MW). After IEF, the proteins contained in the strips were equilibrated in a buffer containing SDS (Sodium Dodecyl Sulfate), allowing the proteins to be negatively charged, so that they have the same electric charge and MW is the only parameter acting in SDS-PAGE. Gels were then set in a tank (Ettan Dalt II system, Amersham). Electrophoresis was performed at 20 mA/gel overnight at 4uC. Gels were then revealed by using a scanner (Typhoon Variable Mode Imager 9400). For each gel, pictures were checked by the ImageQuant software (GE Healthcare Life Sciences).
Selection of spots of interest. It was performed by comparing proteomic profiles with the Samespot software (Nonlinear Dynamics), which is specifically designed to assess the reproducibility of changes in protein levels determined by 2-D DIGE. The first step aimed to compare all spots at times T = 0, 7 and 14 days. The null hypothesis was that there is no significant protein expression change over time. Spots that differed significantly from that of the theoretical T = 0 at p#0.05 were selected. The second step aimed to compare these spots to the spots of each of the pure genotypes. The null hypothesis was here that protein expression of the mixture should be similar to that of the genotype that has the fastest growth rate, i.e. P209 cl1 (see above). The spots of interest that were finally retained followed these 2 criteria: (i) change over time; (ii) protein expression different from the fastest growing genotype. For material reasons, we finally retained only 40 spots, which proved to correspond to 43 proteins. Identification by mass spectrometry. After coloration of the spots by Coomiassie G-250 blue, out of 1,850 spots identified by Samespot, the 40 spots of interest were removed from the gels. Enzymatic in-gel digestion was performed automatically (Tecan freedom evo proteomics) according to Shevchenko et al's modified protocol [20]. Briefly, protein spots were digested using 150 ng of trypsin, peptide extraction was performed using 5 sonication cycles of 2 mn each, and peptides were concentrated 1 hr at 50uC in a heat block. Peptide samples were automatically spotted (Tecan freedom evoH proteomics). For this step, 0.5 ml of peptide sample and 0.5 ml of alpha-cyano-4-hydroxy-trans-cinnamic acid (a saturated solution prepared in acetonitrile/trifluoroacetic acid, 50:0.1%, vortexed, sonicated 30 s and microcentrifuged 30 s with a 1/3 dilution of the supernatant used as the matrix) were deposited on a 384-well MALDI anchorship target using the drydroplet procedure [21] and air dried at room temperature. Peptide samples were then desalted using a 10 mM phosphate buffer and dried again at room temperature. Peptides samples were then analyzed by mass spectrometry MALDI-TOF.
MALDI-TOF MS analysis was performed using UltraFlex MALDI TOF-TOF mass spectrometer (Brucker Daltonics, Bremen, Germany) in the reflectron mode with a 26 kV accelerating voltage and a 50 ns delayed extraction. The AutoXecute module of Flexcontrol v3.0 (Bruker Daltonics) (laser power ranged from 32 to 40%, 600 shots) was used to acquire mass spectra. Spectra were analyzed using FlexAnalysis software v3.0 (Bruker Daltonics) and calibrated internally with the autoproteolysis peptides of trypsin (m/z: 842.51; 1045.556; 2211.10). Peptides were selected in the mass range of 900-3000 Da.
Peptide Mass Fingerprint identification of proteins was performed by searching against the Trypanosoma entries of either the Swiss-Prot or TrEMBL databases (v20090506, http://www. uniprot.org/) and by using the MASCOT v2.2.05 algorithm (Matrix Science Inc. http://www.matrixscience.com/) with trypsin enzyme specificity and one trypsin missed cleavage allowed. Carbamidomethyl was set as fixed cystein modification and oxidation was set as variable methionine modification for searches. A mass tolerance of 50 ppm was allowed for identification. Matching peptides with one missed cleavage were considered as pertinent when there were two consecutive basic residues or when arginine and lysine residues were in an acidic context. MASCOT scores higher than 57 were considered as significant (p,0.05) for Swiss-Prot and TrEMBL database interrogations.