Trypanosoma cruzi Coexpressing Ornithine Decarboxylase and Green Fluorescence Proteins as a Tool to Study the Role of Polyamines in Chagas Disease Pathology

Polyamines are essential for Trypanosoma cruzi, the causative agent of Chagas disease. As T. cruzi behaves as a natural auxotrophic organism, it relies on host polyamines biosynthesis. In this paper we obtained a double-transfected T. cruzi parasite that expresses the green fluorescent protein (GFP) and a heterologous ornithine decarboxylase (ODC), used itself as a novel selectable marker. These autotrophic and fluorescent parasites were characterized; the ODC presented an apparent Km for ornithine of 0.51 ± 0.16 mM and an estimated Vmax value of 476.2 nmoles/h/mg of protein. These expressing ODC parasites showed higher metacyclogenesis capacity than the auxotrophic counterpart, supporting the idea that polyamines are engaged in this process. This double-transfected T. cruzi parasite results in a powerful tool—easy to follow by its fluorescence—to study the role of polyamines in Chagas disease pathology and in related processes such as parasite survival, invasion, proliferation, metacyclogenesis, and tissue spreading.


Introduction
Protozoan parasites cause disease, death, and severe social and economic losses, particularly in developing countries. Trypanosoma cruzi is the causative agent of Chagas disease, which affects 16-18 million people mainly in Latin America [1,2].
T. cruzi is a digenetic parasite whose life cycle involves an insect vector (from Reduviidae family) and a vertebrate host. The parasite undergoes deep morphological, biochemical, and physiological changes as an adaptive response to the extreme environmental variations that it faces around the life cycle. The epimastigote form (E) is a proliferative and noninfective extracellular stage, present in the insect gut. At the insect rectum, it differentiates to metacyclic trypomastigote (MT), an infective stage present in the contaminated insect faeces, which enter the vertebrate host through the insect wound or via mucosal membranes. Once inside, the trypomastigotes invade different kinds of cells, including macrophages, fibroblasts, and muscle cells [3]. There, they transform to amastigote stage and replicate to reach a critical number of parasites per cell, then they differentiate to blood stream trypomastigotes (BT), host cells burst, and release trypomastigotes, which either infect other cells or get sucked up by a bug vector [4][5][6]. The recent construction of stable green fluorescent protein (GFP) expressing T. cruzi strains has resulted in a useful tool to follow processes as cell invasion and infection progression [7,8].
Polyamines are ubiquitous cell compounds essential for macromolecular biosynthesis [9,10] and normal cell survival, proliferation and differentiation [11][12][13]. Several studies related to polyamine metabolism in pathogenic protozoa 2 Enzyme Research have been carried out with the aim to affect the proliferation, infectivity, or differentiation of these organisms and find new therapeutic targets against parasitic diseases [1,14,15]. With this kind of approach it was discovered that the acute infections of Trypanosoma brucei brucei in mice as well as the human African sleeping sickness, caused by T. brucei gambiense, can be treated by α-difluoromethylornithine (DFMO), the specific and irreversible ornithine decarboxylase (ODC) inhibitor [16][17][18].
The naturally occurring polyamines in most cell types, including mammalian cells, are putrescine (actually a diamine precursor), spermidine and spermine, while in trypanosomes this last polyamine was only found in trace amounts [11,19,20]. There are some differences related to polyamine biosynthetic enzymes between mammalian cells and parasites; however, polyamines are essential for both kinds of organisms. In fact, only two kinds of wild-type cells, Halobacteriales and Methanobacteriales, have been described with the ability to grow in the absence of a detectable polyamine concentration [21,22].
The next step in polyamine biosynthesis is catalyzed by spermidine synthase (SpdS; EC 2.5.1.16) that transfers an aminopropyl group from the donor decarboxylated S-adenosylmethionine (dcAdoMet) to putrescine, producing spermidine. Spermine synthase (SpmS; EC 2.5.1.22) transfers a second aminopropyl group from dcAdoMet producing spermine. This enzyme is generally absent in the protozoan parasites; however, T. cruzi would be an exception since its genome presents two aminopropyltransferase genes, one of them a putative SpmS. This function has been proposed in silico but not functionally validated yet [1,30].
In trypanosomatids, spermidine plays a central role because it is the precursor of the derivative N1, N8-bis-glutathionyl-spermidine (trypanothione). This compound, unique in these protozoan organisms, is essential to maintain the redox equilibrium in the parasite cell and plays the metabolic and antioxidant functions of glutathione [33][34][35]. For this and other reasons, it is not surprising that polyamines are essential for survival and proliferation of T. cruzi [36][37][38][39][40]. However, T. cruzi is one of the extremely few eukaryotic organisms auxotrophic for polyamines. We have previously demonstrated that T. cruzi epimastigotes are unable to synthesize putrescine de novo because they contain neither ODC nor ADC enzymatic activity, due to the absence of ODC and ADC genes [38,41,42]. This conclusion is also supported by data from the T. cruzi genome project [30]. Polyamines seem to be crucial for the other T. cruzi stages, particularly in trypomastigote internalization process in, at least, some cellular types, and infection progression as well as for other related parasites infection [6,[43][44][45][46]. Since parasites live inside other organisms, numerous essential metabolites can be obtained directly from the host without the need to synthesize them [47]. It seems to be the case for polyamines, where its uptake systems from the parasite together with the ODC enzymatic activity from the host are a concerted and critical system for T. cruzi survival and host infection.
In this work we obtained a double-transfected T. cruzi strain that expresses the green fluorescent protein (GFP) and a heterologous ODC enzyme. The latter is encoded by the ODC gene, an auxotrophy rescuing gene used here as a selectable marker. Then we selected a Y-GFP-ODC clone with high expression of both GFP [7] and ODC activity [38]. These autotrophic and fluorescent parasites were characterized as well as the heterologous ODC enzyme activity. As T. cruzi is a naturally auxotrophic organism for polyamines, the autotrophy acquisition could provide information about the role of polyamines in processes such as parasite survival, invasion, proliferation, metacyclogenesis and tissue spreading, under different conditions of polyamines availability natural or induced with polyamine inhibitors; the green fluorescent protein allows to easily follow those processes microscopically. Therefore, the double-transfected T. cruzi parasite is a powerful new tool to study the role of polyamines, naturally provided by the host ODC, in the parasite cell cycle. This would help to better understand Chagas disease pathology and to find targets for new therapies.
To determine the growth kinetics of each strain, a part of Y-GFP culture was maintained in BHT while others were placed in SDM79 medium in absence or presence of 1 mM putrescine or in Diamond in absence or presence of 2 mg/mL glucose. Y-GFP-ODC strain and the derived clone were maintained in SDM79 medium without or with the addition of 5 mM α-difluoromethylornithine (DFMO). Parasite growth was followed by cell counting. Cultures were diluted to 10-20 × 10 6 cells/mL every 48-72 h, when the stationary phase was reached.

Parasite Transfection.
Epimastigote forms of Y strain had been previously transfected with pROCKGFPNeo and characterized [7,48]. This recombinant plasmid contains the GFP coding region and sequences that allow a stable DNA integration at the β-tubulin loci [7]. A selected clone from this Y-GFP strain was retransfected with the recombinant vector pODC 7 , as previously described [38]. pODC 7 contains the complete open reading frame (ORF) of ODC gene from C. fasciculata [49] cloned into the expression vector pRIBOTEX [50] that integrates at the ribosomal genes loci [38,51].
After transfection, parasite culture was diluted with rich medium and incubated for 72 h at 28 • C to allow its recovery before adding the restrictive culture conditions, consisting in 500 μg/mL Geneticin (G 418) (selection by antibiotic resistance) and semisynthetic medium SDM79 (selection by polyamine autotrophy development). Subsequent growth was followed by cell counting. Samples of T. cruzi cultures before and after transfection were taken to analyze cellular morphology, fluorescence, and ODC activity. Drug selection was completed 50 days after electroporation as indicated by the mock-transfected parasites (transfected with empty p-RIBOTEX) and the appearance of ODC activity. This time was slightly longer than that seen in each of the previous single-transfection cases (the T. cruzi transfection with GFP [7] and with ODC [38], in both cases selected by resistance to 250 μg/mL G418).

Selection of T. cruzi Y-GFP-ODC Clones.
After selection of the double-recombinant Y-GFP-ODC T. cruzi strain, cloned subpopulations were obtained by diluting part of the culture to a final density of 0.1 cells/0.2 mL of SDM79 medium supplemented with 10% FCS and 250 μg/mL of G418 and plated in 96-well plates. They were incubated at 28 • C in humid chamber for 4-5 weeks and analyzed with a fluorescence or confocal microscope. GFP-expressing clones were selected and transferred into 1 mL of the same medium in culture bottles and incubated at 28 • C. Once they reached the plateau the cultures were expanded to 5 mL and ODC activity was assayed (see Section 3).

Preparation of Cell Extracts and ODC Enzymatic Assays.
Cell extracts were obtained and the enzymatic activity was measured by the release of 14 CO 2 from L-[1-14 C]ornithine as previously described [52]. Briefly, parasites were collected at the exponential phase of growth. They were washed and lysated by two cycles of freeze-thawing and sonication, in buffer A (50 mM Hepes, pH 7.4, 1 mM DTT, 0.5 mM EDTA and 0.1 mM pyridoxal 5 -phosphate) supplemented with protease inhibitors and 0.5% Nonidet P40. Cell lysates were centrifuged for 10 min at 10000 ×g. Supernatant fractions were tested for ODC activity by incubation at 37 • C in presence of 2 mM L-[1-14 C] ornithine (5 uCi/mL). The enzymatic reaction was performed for the indicated times and stopped by the addition of 0.5 N perchloric acid. The radioactive CO 2 released during the reaction was trapped on a piece of Whatman 3 MM paper soaked with 2 M KOH and then measured in a scintillation counter. Protein concentration was measured, using BSA as standard, according to Bradford [53].

Kinetic Parameters of Heterologous ODC.
Ornithine decarboxylation was carried out at different times, substrate and inhibitor concentrations. The determination of the kinetic constants apparent Km and V max values for ornithine of the enzyme expressed in Y-GFP-ODC were calculated from Lineweaver-Burk plots. Studies of inactivation by DFMO were carried out as previously described [38,54].

Trypanosoma cruzi Metacyclogenesis.
For in vitro differentiation of T. cruzi epimastigotes, we have used a method previously described [55,56] with some modifications. Briefly, 25 × 10 6 epimastigotes from 15-day-old cultures from Y-GFP and Y-GFP-ODC strain and clone grown in Diamond or SDM79 media, respectively, were harvested by centrifugation at 600 ×g for 15 min at 15 • C. The parasites were then incubated for 2 h at 37 • C in 50 μL of triatomine artificial urine (TAU) medium (190 mM NaCl, 17 mM KCl, 2 mM MgCl 2 , 2 mM CaCl 2 , 8 mM phosphate buffer, pH 6.0). After that, parasites were diluted (1 : 100 dilution) in TAU3AAG medium (TAU supplemented with 10 mM Lproline, 50 mM L-sodium glutamate, 2 mM L-sodium aspartate, and 10 mM D-glucose) and maintained for 24 or 48 h in culture flasks at 28 • C [55,56]. Epimastigotes treated with α-difluoromethylornithine (DFMO) or Putrescine (Put) during metacyclogenesis were obtained by addition of 5 mM DFMO or 0,2 mM Put in both TAU and TAUAAG media at the same conditions as controls.
Mixed parasitic forms corresponding to epimastigotes and metacyclic trypomastigotes (E-MT suspensions) were harvested by centrifugation at 600 ×g for 15 min and used to infect Vero cells monolayers using 24-well plates containing glass coverslips in D-MEM supplemented with 3% FBS and antibiotics (infection medium). 24 h after infection, cells were washed, fixed, and stained with Rhodamine-phalloidin for the fluorescence microscopy analysis (see below). The metacyclogenesis capacity of these suspensions was estimated by counting the percentage of infected cells. Parasites were directly localized as green structures inside cells, whereas host cell extension was visualized in red, as indicate in the following.

Fluorescence Microscopy.
For Rhodamine-phalloidin staining, Vero cells were washed three times with PBS to remove free extracellular parasites, fixed with 3% paraformaldehyde solution in PBS for 10 min at RT, washed with PBS, and blocked with 50 mM NH 4 Cl in PBS. Subsequently, cells were permeabilized with 0.05% saponin in PBS containing 0.5% BSA and then incubated with Rhodamineconjugated phalloidin for 1 h at 37 • C to label F-actin. Cells were washed three times with PBS and mounted with Mowiol before fluorescence microscopy analysis. Confocal images were obtained with an Olympus FV1000 confocal microscope and the FV 10-ASW 1.7 program (Olympus, Japan). Images were processed using Adobe 7.0 software (Adobe Systems).

Growth Kinetics of Y-GFP and Y-GFP-ODC Strain and
Clone. Y-GFP strain cultured in the rich medium BHT, supplemented with G418, maintained the proliferative capacity for long periods of time having a doubling time of 2-3 days and a constant maximal cell density of ∼40 × 10 6 parasites/mL (Figure 1(a)). On the other hand, Y-GFP cultured in Diamond, the rich medium habitually used for maintaining this strain [7,8], grew slowly with a doubling time of approximately 7 days. The maximal cell density (∼37 × 10 6 parasites/mL) was scarcely reached at the first-second passage (Figure 1(a)). To discard that the low proliferation rate was due to low availability of glucose, an important source of energy for T. cruzi [57], we followed the Y-GFP growth in Diamond supplemented with 2 mg/mL glucose. There were no differences in the kinetic characteristics between both conditions. When Y-GFP strain was cultured in the semisynthetic medium SDM79 (containing only traces of polyamines), the kinetics of growth (Figure 1(b)) was similar to that observed in Diamond. After 10-15 days of culture under these three conditions, Y-GFP parasites progressively stopped growth, started decreasing, and finally were lysed. However, when Y-GFP was cultured in SDM79 supplemented with 1 mM putrescine the culture could recover the proliferation rate (Figure 1(b)). The double-transfected Y-GFP-ODC clone grew continuously in SDM79, even better than Y-GFP in BHT, with a doubling time of 2-3 days and a maximal cell density of ∼80 × 10 6 parasites/mL (Figure 1(c)). Similar results were obtained for the Y-GFP-ODC strain (not shown). The continuous proliferation of the double-transfected Y-GFP-ODC strain and clone in absence of polyamines indicates that the original auxotrophy for diamines of T. cruzi was reversed after trans-fection with the heterologous ODC gene, as it was previously described for monotransfected strains [38].
α-difluoromethylornithine (DFMO), the specific and irreversible inhibitor of ODC, altered the growth rate of Y-GFP-ODC clone, particularly the maximal cell density (lower than 55 × 10 6 parasites/mL) (Figure 1(c)). The partial effect of DFMO on the proliferative rate could be explained in different ways [58,59] that will be discussed later.
The three cultures were analyzed by confocal microscopy. There were ∼80% of parasites expressing GFP in Y-GFP strain, while the Y-GFP-ODC double-transfected strain and the selected clone presented values of ∼95%, similar to those described by Pires et al. [7]. The cell morphology of the three cultures was similar (Figure 2).

ODC Enzymatic Activity in Double-Transfected T. cruzi
Epimastigotes. The ODC activity was measured at 72 h and 10, 20, 30, 40, and 50 days after transfection of Y-GFP T. cruzi epimastigotes with the heterologous ODC. We have found that double-transfected Y-GFP-ODC cells expressed high levels of ODC activity after 50 days following transfection (Figure 3(a)). It was different from previous reports, where a transient expression (between 48 h and 7 days) was detectable in ODC single-transfected parasites; that expression decreased markedly during the following weeks and recovered at 30-40 days after transfection [38]. This difference could be related to the cell machinery readaptation to transcribe-translate a new foreign and active gene, doing this already with GFP and the neomycin phosphotransferase (neomycin resistance).
The ODC specific activity in the double-transfected Y-GFP-ODC was around 39 ± 4 nmol/h/mg prot. Once the Y-GFP-ODC clone was selected, it showed an ODC activity with little variation during the course of the experiment reaching values of 590 nmol/h/mg prot (Figure 3(b)). The ODC activity in the Y-GFP strain was, as expected, negligible in all assays.

Properties of the Heterologous ODC Expressed in the
Selected Y-GFP-ODC Clone. The kinetic parameters of ODC expressed in the Y-GFP-ODC clone were calculated from a Lineweaver-Burk plot (Figure 4(a)). The estimated V max value was 476.2 nmoles/h/mg of protein, and the apparent Km for ornithine was 0.51 ± 0.16 mM. While the Km value was similar to those obtained previously for other transfected T. cruzi strains [38] and for C. fasciculata [60], the V max was higher than those obtained in previous assays of singletransfection with ODC gene in DFMO-sensitive strains. Instead, the V max value obtained in Y-GFP-ODC clone was similar to that observed for ODC-transfected strains after reaching DFMO resistance [38,58,59].
The assays of in vivo ODC turnover (determined by stopping protein synthesis with 50 μg/mL cycloheximide and measuring the remaining enzymatic activity at different times) showed that the ODC expressed in Y-GFP-ODC clone was very stable, with a half-life longer than 6 h (data not shown), remarkably longer than the half-life of the ODC expressed in Crithidia fasciculata, its natural cellular context [60]. This notorious change in stability has already been observed in experiments of single transfection in other strains of T. cruzi [38].
When Y-GFP-ODC clone was cultured in presence of DFMO, the inhibitor was effectively internalized into the cell causing a notable ODC activity inhibition (Figure 4(b)). Additional in vitro assays showed that DFMO inhibits the ODC reaction in a dose-response way, with an IC 50 lower than 0.13 mM (Figure 4(c)).
Based on these observations, the best explanation of why DFMO caused only a moderated effect on Y-GFP-ODC clone proliferation (Figure 1(c)) seems to be the high V max of ODC activity that correlates with a high number of active ODC molecules, making Y-GFP-ODC clone resistant to DFMO.

T. cruzi Metacyclogenesis in Y-GFP and Y-GFP-ODC Parasites.
We next compared the capacity of Y-GFP-ODC (strain, and clone) and the auxotrophic Y-GFP to differentiate from epimastigotes (E) to metacyclic trypomastigotes (MT) forms using an in vitro protocol previously described [55,56]. Y-GFP, Y-GFP-ODC strain and Y-GFP-ODC clone were incubated with TAU medium for 2 h and subsequently with TAU3AAG medium for 48 hrs to induce T. cruzi differentiation (TAU treatment). The MT generated for each cell culture was estimated using the E-MT suspensions to infect Vero cell monolayers during 24 h and comparing with their controls, in Diamond or SDM79 media for Y-GFP or Y-GFP-ODC, respectively.
After the infection time, the cells were washed, fixed and incubated with Rhodamine-phalloidin to localize actin cytoskeleton of host cell by fluorescence microscopy, process that allows the clear observation of cellular limits, whereas T. cruzi parasites were observed directly by the presence of GFP protein. Figure 5(a) shows, at two differential amplification levels, confocal images of Vero cells infected by MT generated by epimastigotes of Y-GFP or Y-GFP-ODC strain under TAU treatment. Quantification studies indicated that the percentage of infected cells increased when comparing parasites obtained from Y-GFP-ODC metacyclogenesis with the Y-GFP control ( Figure 5(b), black bars); even more, in some cases it was possible to observe more than one Y-GFP-ODC parasite per cell. Although the magnitude of this increment was not as high as to be statistically significant, the tendency was constant. Supporting this result, we observed that the presence of DFMO in the metacyclogenesis assay (see details in Materials and Methods) impaired the infection of Y-GFP-ODC strain and clone, while it has no evident effect over Y-GFP ( Figure 5(b), dashed bars) because the inhibitor target is not present in this strain. Conversely, the percentage of infected cells was increased when Put was added to Y-GFP parasites during the differentiation process ( Figure 5(b), grey bars) showing similar values to those found in Y-GFP-ODC parasites. Those results indicate that polyamines supply favours the E to MT conversion.
There is an important number of morphological and biochemical events that occur during metacyclogenesis process. Although the molecular causes of those changes are largely unknown, these experiments clearly show the participation of polyamines in that process. Recently published results showed that spermidine can regulate the transcription of cell cycle and autophagy genes in numerous organisms [13]. Our results show that the availability of polyamines during T. cruzi differentiation favours the global process, increasing the infective capacity of E-MT suspensions. As our results were obtained by indirect infection of Vero cells, the major infective capacity, reached in the presence of endogenous or exogenous source of polyamines, could be explained in different ways. Polyamines could, for example, activate the expression of genes that participate in differentiation from E to MT or those that code to membrane molecules necessary for host cell interaction and invasion, improving the quantity or quality of the process, respectively.

Concluding Remarks
Polyamines availability shows to be an important requirement for T. cruzi proliferation, invasion of mammalian host cells, intracellular multiplication, and infection progression, but the way that they act is still uncertain [1,44,[61][62][63]. Unlike other trypanosomatid parasites, T. cruzi does not synthesize polyamines by an endogenous ODC and relies on the polyamine availability from the parasite external milieu [38,41]. However, some evidence suggests that polyamines are not as free and available as it could be predicted by polyamine abundance in the host; polyamines could be conjugated or stored in reservoirs so they are not homogeneously distributed and equally available around all host cells and fluids or under different treatments [37,46]. The polyamine content in the host relies on diet and, specially, in its ODC enzyme, whose activity is highly regulated at multiple levels [64]. Different treatments that alter the host ODC activity modulate, in turn, polyamine intracellular content-ergo its availability for T. cruzi intracellular stages-as much as polyamine concentration in extracellular fluids, the natural medium for bloodstream stages. In this context, the host ODC as well as the parasite uptake system would be potential targets, as by self or as a complementary target increasing the effect of antitrypanosomal therapies. As there has been some discrepancies about the possible presence of arginine decarboxylase (ADC) activity under specific metabolic conditions [41,42,[65][66][67][68] as well as there are clear evidence about the presence of ADC in at least some kind of cells that could be infected by T. cruzi [27][28][29], this other possible way to obtain polyamines should be taken into account when the role of polyamines among the T. cruzi biological cycle is studied. In any case, the growth curves and the results obtained in metacyclogenesis assays support the idea that polyamines are engaged in parasite proliferation and in differentiation process. Therefore, this double transfected T. cruzi becomes a powerful and easy-to-follow tool to study the role of polyamines in Chagas disease pathology. In particular, the expression of GFP gene in this double-mutant (or in new mutants for other genes), which allows a simple visualization of T. cruzi cellular limits, represents an advantage compared to the detection techniques frequently used (such as T. cruzi nuclear and kinetoplast DNA Hoechst staining or indirect immunofluorescence) to localize and characterize the host cell parasitophorous vacuole (TcPV) as well as the transit and escape of T. cruzi from the TcPV. Further experiments of infection in cell monolayers and animal models will be done to confirm and extend the knowledge about the role of polyamines in host cell invasion and infection progression.