DNA content analysis allows discrimination between Trypanosoma cruzi and Trypanosoma rangeli

Trypanosoma cruzi, a human protozoan parasite, is the causative agent of Chagas disease. Currently the species is divided into six taxonomic groups. The genome of the CL Brener clone has been estimated to be 106.4–110.7 Mb, and DNA content analyses revealed that it is a diploid hybrid clone. Trypanosoma rangeli is a hemoflagellate that has the same reservoirs and vectors as T. cruzi; however, it is non-pathogenic to vertebrate hosts. The haploid genome of T. rangeli was previously estimated to be 24 Mb. The parasitic strains of T. rangeli are divided into KP1(+) and KP1(−). Thus, the objective of this study was to investigate the DNA content in different strains of T. cruzi and T. rangeli by flow cytometry. All T. cruzi and T. rangeli strains yielded cell cycle profiles with clearly identifiable G1-0 (2n) and G2-M (4n) peaks. T. cruzi and T. rangeli genome sizes were estimated using the clone CL Brener and the Leishmania major CC1 as reference cell lines because their genome sequences have been previously determined. The DNA content of T. cruzi strains ranged from 87,41 to 108,16 Mb, and the DNA content of T. rangeli strains ranged from 63,25 Mb to 68,66 Mb. No differences in DNA content were observed between KP1(+) and KP1(−) T. rangeli strains. Cultures containing mixtures of the epimastigote forms of T. cruzi and T. rangeli strains resulted in cell cycle profiles with distinct G1 peaks for strains of each species. These results demonstrate that DNA content analysis by flow cytometry is a reliable technique for discrimination between T. cruzi and T. rangeli isolated from different hosts.


Introduction
Members of the genus Trypanosoma are protozoan parasites found worldwide and are capable of infecting humans, domestic and wild animals, and insects. Trypanosoma cruzi is the causative agent of Chagas disease, a chronic and debilitating disease that affects approximately 8 million people, mainly in Latin America [1]. Trypanosoma rangeli is also a protozoan parasite, which occurs in sympatry with T. cruzi. Despite T. rangeli infecting humans, it is considered nonpathogenic to the vertebrate hosts. However, T. rangeli infection can elicit the production of antibodies that cross-react with T. cruzi antigens. This may lead to the misdiagnosis of Chagas disease leading to a socioepidemiological impact and has not been considered by health authorities [2,3]. T. cruzi population is divided in six genetic groups, viz., TcI-TcVI [4]. T. cruzi is considered diploid, but some parasitic strains are aneuploids because of a variation in the number of chromosomal bands or distribution of genetic markers, as determined by microsatellite (MS) typing [5,6]. Sequencing of the clone CL Brener of T. cruzi revealed a haploid genome estimated to be 55 Mb [7]. Moreover, flow cytometric analysis using the clone CL Brener as the reference cell line demonstrated a variation in the nuclear genome size between T. cruzi groups, ranging from 80.64 Mb to 153.58 Mb [6,8].
T. rangeli also possesses a high intraspecific variability, and analysis of kinetoplast DNA (kDNA) allowed the determination of two main genetic lineages in the parasite, viz., KP1(+) and KP1(−) based on the presence or absence of KP1 minicircles in the parasitic kDNA [9]. The division of T. rangeli into two main groups has been confirmed by several techniques, including RAPD [10], molecular karyotype [11], and terminal restriction fragment analyses [12]. However, analysis of other genetic markers, such as mini-exon, SSU rDNA, and CatLlike genes, detected increased variability allowing the division of the taxon into five groups, viz., TrA-TrE [13]. Recently, the investigation of single nucleotide polymorphisms (SNPs) and MS typing revealed a subdivision of the KP1(−) group, making a total of three T. rangeli groups [14]. Compared to other trypanosomatids, T. rangeli has the smallest genome sequenced thus far, with its haploid complement estimated to be 24 Mb [3].
Considering the limitations of serological methods for differential diagnosis of infections caused by T. cruzi and T. rangeli, several molecular methods have been developed for the differential diagnostic between these parasitic strains. Souto and colleagues demonstrated that analysis of the divergent D7a domain of rDNA permits simultaneous identification of T. cruzi and T. rangeli [15]. Ferreira and colleagues performed a comparative genome sequence analysis to identify molecular markers, which can specifically identify and distinguish between T. cruzi and T. rangeli [16]. Furthermore, DNA sequencing analysis of KP1(+) and KP1(−) strains of T. rangeli revealed the occurrence of a high frequency of nucleotide substitutions, which were named group specific substitutions (GSP) [16].
Despite several attempts for developing techniques for differential discrimination between T. cruzi and T. rangeli, investigating the DNA content of the parasitic strains has been explored limitedly. Thus, in this study, we investigated the DNA content of T. cruzi and T. rangeli by flow cytometry and demonstrated this approach to be a reliable alternative for discrimination between these species.

Culture conditions
Epimastigote forms of T. cruzi and T. rangeli were cultured in liver infusion tryptose (LIT) medium [18] supplemented with 10% fetal bovine serum. T. rangeli strains were cultured in LIT supplemented with 3% (v/v) human urine [19]. Promastigote forms of L. major were cultured in M199 supplemented as previously described [20].The cultures were incubated for four days at 28˚C in a biochemical oxygen demand incubator (BOD).

DNA extraction
DNA extraction of T. cruzi and T. rangeli epimastigote forms and L. major promastigote forms was performed with the Wizard 1 Genomic DNA Purification kit (Promega, Madison, Wisconsin, USA) according to the manufacturer's instructions. Approximately 1 × 10 7 parasite forms obtained from the exponential phase of growth curves were used for DNA extraction.
Genetic characterization of the T. cruzi strains was performed by PCR-restriction fragment length polymorphism (PCR-RFLP) using the TcSC5D (Sterol C-5 Desaturase from T. cruzi) gene as a target [22]. The primers TcSC5D-fwd (5 0 -GGACGTGGCGTTTGATTTAT-3 0 ) and TcSC5D-rev (5 0 -TCCCATCTTCTTCGTTGACT-3 0 ) amplify an 832-bp fragment that contains polymorphisms associated with restriction sites for endonucleases HpaI and/or SphI. In this fragment, HpaI sites are found in homozygosity in TcI (generating fragments of 177 and 655bp) and TcII (231 and 601bp) strains and SphI sites are also found in homozygosity in TcIII (337 and 495bp) strains. TcIV has no restriction sites for HpaI or SphI in the 832-bp fragment. HpaI and SphI sites are found in heterozygosity in the 832-bp fragments of TcV and TcVI (231, 337, 495, and 601bp) strains [22]. In order to discriminate between TcV and TcVI, we used primers Tc-Mec-kinase26-Fw (5 0 -TTTTTGCATGTCATTTTGG-3 0 ) and Tc-Mec-kinase662-Rv (5 0 -AGCGGTCTTGTAATGAGCAC-3 0 ) that amplify a fragment of 637bp T. cruzi mevalonate kinase (TcMK) gene [22]. XhoI digestion of the 637-bp fragment discriminates TcV (digestion) from TcVI (no digestion). The PCR protocols were conducted as described [22]. Briefly, the final volume of the reaction mixture was 25 μL, containing 10 pmol of each primer, PCR buffer (Invitrogen), 1.6 mM MgCl 2 , 50-100 ng genomic DNA, 200 mM dNTPs, and 1 unit Taq DNA polymerase (Invitrogen). PCR conditions were as follows: 94˚C for 5 min; 35 cycles of 94˚C for 30 s; 55˚C for 30 s; 72˚C for 30 s; and final extension at 72˚C for 5 min for amplification of TcSC5D and at 94˚C for 4.5 min, 35 cycles at 94˚C for 30 s, followed by 30 s at 58˚C, and 72˚C for 30 s and final extension of 5 min at 72˚C for amplification of TcMK [22]. The amplified products were observed on 1.2% agarose gel stained with ethidium bromide. Aliquots of 20 μL of the amplified products were digested with 1U of the enzyme HpaI (NEB R105) at 55˚C for 1 h and with 1U of the enzyme SphI (NEB R0182) at 37˚C for 1 h (TcSC5D fragment) or with one unit of XhoI (NEB R0146S) endonuclease (TcMK fragment). The resulting digestion fragments were observed on 1.5% agarose gel and stained with ethidium bromide.

Flow cytometry analysis of T. cruzi and T. rangeli
To estimate the DNA content, aliquots with approximately 1 × 10 7 epimastigote forms of T. cruzi and T. rangeli strains were centrifuged at 2000 ×g for 5 min at 4˚C. The supernatant was discarded, and the cells were washed with ice cold 1× PBS. The pellet was suspended in 500 μL 1× PBS and 4.5 mL of 70% ethanol for fixation. The fixed cells were maintained at 4˚C until flow cytometry DNA content analysis.
Propidium iodide (PI) staining was performed according to the method previously described [6]. Fixed cells were centrifuged at 2000 ×g for 5 min at 4˚C. The supernatant was discarded, and the cells washed once with ice cold 1× PBS. The pellet was resuspended in 1 mL of a solution containing PI (100 μg/mL), Triton X-100 [0.1% (v/v)], and DNase free RNase A (200 μg/mL) and incubated for 15 min at 37˚C in dark. Data acquisition was performed in a BD FACSCanto II Flow Cytometer (Becton, Dickinson and Company, USA). A minimum of 50,000 events were counted for each sample, at least in duplicates. Data analysis was performed using the FlowJo software (Tree Star Inc., Oregon, USA) and expressed as PI mean fluorescence intensity (MFI) in G0/G1 (2n) using FlowJo Cell Cycle identification tool. In brief, the parasitic strains were identified by the forward scatter (FSC) and side scatter (SSC) parameters, and debris and aggregates were excluded by using pulse area vs. pulse width and pulse area vs. pulse height.
Finally, PI fluorescence at 617 nm was evaluated by histograms. T. cruzi clone CL Brener was used as the reference for DNA content determination of all other parasitic strains analyzed because its genome size is known (106.4-110.7 Mb) [7] and its DNA content has been previously profiled by flow cytometry [6]. We also use the L. major CC1 clone as a reference because genome sequencing of three L. major strains revealed little variation in genome size [23]. In our analysis, we considered a value of 65.6Mb for a L. major diploid cell.
From fluorescence values at G1 peaks and genome sizes of reference strains, we determined linear regression curves that were used to estimate genomes of other T. cruzi and T. rangeli strains from their μ values. Furthermore, for estimating the genome sizes of T. cruzi and T. rangeli strains, the amount of T. rangeli kDNA was considered to be equivalent to the amount of T. cruzi kDNA [6,8,24]. For the statistical analysis of estimated genome size data (EGS), the Levene's and Welch's tests for homogeneity evaluation and the Komolgorov-Smirnov's and Shapiro-Wilk's W tests were used to evaluate the normality of the distribution. The tests assumed non-parametric distribution and, therefore, the independent samples Mann-Whitney U test was used. The tests which p-value was less than 0.05 were considered significant. All analyzes were performed with the IBM1 SPSS1 Statistics version 20.0 program.

Genetic identification of T. cruzi and T. rangeli strains
In order to ensure absence of cross-contamination in T. cruzi and T. rangeli cultures, we used a duplex-PCR protocol that allows the amplification of a fragment of 100bp from T. cruzi and a fragment of 170bp from T. rangeli. Accordingly, a 170-bp fragment was detected from all strains of T. rangeli and a 100-bp fragment was detected from all T. cruzi strains (Fig 1). Therefore, we conclude that there is no contamination of T. rangeli cultures with T. cruzi and vice-versa.

Genetic characterization of T. cruzi and T. rangeli strains
Next, we characterize the main genetic groups of parasitic strains studied. Primers S35, S36, and KP1L were used to characterize kDNA minicircles in T. rangeli strains. As expected, all T. rangeli samples allowed the amplification of fragments between 300-400 bp (corresponding to KP3 minicircles) and a faint 760-bp fragment corresponding to the KP2 minicircles. Additionally, only P02, P07, P19, and Cas4 T. rangeli strains presented fragments in the size of approximately 165 bp, confirming their genotype as KP1(+) (Fig 2A).
PCR-RFLP of the TcSC5D gene was performed to determine the genetic groups of the T. cruzi strains. An 832-bp fragment was detected in all T. cruzi strains analyzed ( Fig 2B). Next, each fragment was subjected to digestion with HpaI and SphI. Thus, strain Dm28c was classified as TcI (655 bp and 177 bp bands); strains RN1, JG, and Hel were classified as TcII (601 bp and 231 bp bands); strain 3663 was classified as TcIII (495 bp and 337 bp); and the results obtained for clone CL Brener (fragments of 601 bp, 495 bp, 337 bp, and 231 bp) are compatible with TcV or TcVI (Fig 2C). In order to confirm CL Brener genotype, we conducted a PCR-RFLP analysis of mevalonate kinase amplification product incubated with XhoI, determining its genotype as TcVI (S1 Fig).  Fig 3 represents the strategy used to analyze the DNA content of T. cruzi and T. rangeli strains using the T. cruzi CL Brener clone and the L. major strain CC1 as reference cell lines. In the parameter SSC-A and FSC-A (Fig 3A and 3D) it is possible to visualize the cellular populations in terms of complexity and size. Cells labeled with PI are shown in PI-A em PI-W parameters (Fig 3B and 3E). Fig 3C and 3F show histograms with distinct peaks in the different species showing cells in the G1-0 and G2-M phase of the cell cycle.

Flow cytometry analysis of epimastigote forms of T. cruzi and T. rangeli
In this study, we included the clone CL Brener (TcVI) of T. cruzi and the CC1 strain of L. major as the reference cell lines in all DNA content analyses to estimate the genome size of the T. cruzi and T. rangeli strains studied. All T. cruzi and T. rangeli strains provided with clearly identifiable cell cycle profiles with G1-0 (2n) and G2-M (4n) peaks (Fig 4A and 4B).  5).
The mean genome size for T. cruzi strains was 100.39±9.88 Mb, which was significantly higher than mean genome size of T. rangeli strains (65.71±3.37 Mb; p < 0.001). Mixtures of epimastigote forms of the clone CL-Brener of T. cruzi and T. rangeli strains resulted in histograms, which allowed discrimination between the G1 peaks of the two species (Fig 6A and 6B).

Discussion
In this study, we have presented a novel strategy for discrimination between the main Trypanosoma spp. that infect humans in South America and Central America. We used PI-stained culture forms of two reference strains of trypanosomatids, T. cruzi CL Brener clone and L. major CC1 strain, that had their nuclear genome size determined by DNA sequencing [7,23]. From these two points, we determine estimate genome sizes of T. rangeli and other T. cruzi strains.
The duplex-PCR subtelomeric protocol [21] excluded the possibility of cross-contamination between T. cruzi and T. rangeli cultures, allowing the use of the parasite forms for DNA content analysis. Next, we confirmed the genetic characterization of the parasites. T. rangeli genetic variability has been studied by several techniques and, according to the molecular marker used, parasite populations can be divided from two to five groups [17]. In our study, we investigated the organization of T. rangeli KP minicircles, because parasites from each group are associated with distinct vector species and the results of KP1 classification were Discrimination between T. cruzi and T. rangeli by DNA content analysis confirmed by several other molecular markers [11,12,17,25]. Accordingly, we observed the fragments amplified from the expected KP2 and KP3 kDNA minicircles in all T. rangeli strains and detected the presence of KP1 minicircles in the previously characterized strains [17,26]. The variation observed in the intensity of the KP2 fragments is consistent with the results previously described for other T. rangeli strains [9]. In the same way, all T. cruzi strains were associated with specific groups after PCR-RFLP analysis of polymorphisms in the SC5D and/or MK genes [22].
We investigated the DNA content of T. rangeli by flow cytometry and conducted a comparative analysis with the DNA content of T. cruzi. Considering the characteristics of flow cytometry, such as efficiency of PI incorporation by DNA and fluorescence compensation [27], the use of at least one reference cell line is mandatory for comparative analysis and consequent estimation of the parasitic genome size.
DNA content analysis of T. cruzi strains revealed that the genome size ranged from 85.2 Mb (Dm28c strain, TcI) to 113.0 Mb (Hel strain, TcII). These values are within the range of genome sizes determined for T. cruzi strains in previous studies that demonstrated a significant Discrimination between T. cruzi and T. rangeli by DNA content analysis variability in the parasitic DNA content [6,8]. Additionally, the genome size of Dm28c was estimated to be 85.2 Mb, which is very similar to the value estimated for other Dm28c strain which had its absolute DNA content previously determined [8]. This result demonstrates the accuracy of our estimates.  however, DNA sequencing has been performed, and the haploid genome size has been estimated to be 24 Mb [3]. Considering the differences between genome size estimation by flow cytometry and DNA sequencing, underrepresented sequences may account for part of the differences found between these values. Other possibility is a significant variation in the kDNA content between the two species. We consider in our analyses that the amount of kDNA is constant between parasite species and strains, as previously proposed for T. cruzi estimates [6]. However, the actual contribution of kDNA and nuclear DNA to the parasite's total DNA content will require further investigation.
T. rangeli strains belonging to both KP1(+) and KP1(−) groups were used for determination of DNA content analysis. Previous studies on molecular karyotyping of these strains demonstrated differences in both the number and size of chromosomal bands between KP1(+) and KP1(−) groups of T. rangeli [11]. However, DNA content analysis between the strains of each group did not reveal significant differences between the two genetic groups of T. rangeli. Furthermore, cultures containing mixtures of the epimastigote forms of T. cruzi and T. rangeli strains resulted in cell cycle profiles with distinct G1 peaks for strains of each species. Accordingly, DNA content analysis by flow cytometry may be useful for species discrimination and for ploidy investigation in these parasites.
Ferreira and colleagues found a high frequency (1/56) of nucleotide substitutions when analyzing two loci of strains representing the main genetic groups of the parasite. This high intraspecific frequency of nucleotide substitutions is uncommon and has been called GSP to distinguish it from SNP, which occurs at a lower frequency in other organisms. In addition, sequencing the clones of P07 and Cas4 strains revealed the occurrence of heterozygosity in two loci [16], which may indicate the occurrence of hybridization in this parasite. In T. cruzi, TcIII Discrimination between T. cruzi and T. rangeli by DNA content analysis and TcIV may be considered as homozygous hybrids, whereas TcV and TcVI are considered as heterozygous hybrids [28,29]. Furthermore, T. cruzi "natural hybrids" are diploids, whereas hybrid lineages obtained under experimental conditions possess DNA content that is compatible to that of tetraploid cells [6]. T. rangeli P07 and Cas4 strains, isolated from Didelphis albiventris and Rhodnius prolixus, respectively, have the DNA content of a diploid cell. Recent studies suggested that the population structure of both T. cruzi and T. rangeli is primarily clonal [14,30]. However, the main reproduction mechanism of this parasite remains controversial. Taken together, these results demonstrate that DNA content analysis by flow cytometry is a useful and reliable technique for discrimination between T. cruzi and T. rangeli isolated from different sources and may be applied for parasite identification if used in association with other approaches, such as PCR analysis. Further studies may reveal the extension of heterozygous loci in T. rangeli and other genetic and biological aspects of this parasite.