Old Methods, New Insights: Reviewing Concepts on the Ecology of Trypanosomatids and Bodo sp. by Improving Conventional Diagnostic Tools

Mixed infections by different Trypanosoma species or genotypes are a common and puzzling phenomenon. Therefore, it is critical to refine the diagnostic techniques and to understand to what extent these methods detect trypanosomes. We aimed to develop an accessible strategy to enhance the sensitivity of the hemoculture, as well as to understand the limitations of the hemoculture and the blood clot as a source of parasitic DNA. We investigated trypanosomatid infections in 472 bats by molecular characterization (18S rDNA gene) of the DNA obtained from the blood clot and, innovatively, from three hemoculture sample types: the amplified flagellates (“isolate”), the pellet of the culture harvested in its very initial growth stage (“first aliquot”), and the pellet of non-grown cultures with failure of amplification (“sediment”). We compared (a) the characterization of the flagellates obtained by first aliquots and isolates; and (b) the performance of the hemoculture and blood clot for trypanosomatid detection. We observed: (i) a putative new species of Bodo in Artibeus lituratus; (ii) the potential of Trypanosoma cruzi selection in the hemoculture; (iii) that the first aliquots and sediments overcome the selective pressure of the hemoculture; and (iv) that the blood clot technique performs better than the hemoculture. However, combining these methods enhances the detection of single and mixed infections.


Introduction
The class Kinetoplastea comprises flagellated protists characterized by the presence of the kinetoplast, a highly condensed organelle consisting of mitochondrial DNA [1]. From the most basal to the derived evolutionary lineage, this class includes the orders Prokinetoplastida, Neobodonida, Parabodonida, Eubodonida, and Trypanosomatida [2][3][4]. Members of Prokinetoplastida and Trypanosomatida are parasites (but see [5]). Neobodonida and Parabodonida are mainly composed of free-living beings, with a few exceptions of parasitic species. Organisms of the order Eubodonida-uniquely represented by the genus Bodo (family Bodonidae)-are known, to date, to be free-living heterotrophs, found globally in aquatic environments [3,6,7].
The level of knowledge about kinetoplastids presents huge discrepancies. Due to their impact on the economy and human health, most attention is given to a few members of the family Trypanosomatidae [8]-the only representative of the order Trypanosomatida. Trypanosomatids include monoxenous genera, mainly found in the digestive tract of insects, and heteroxenous genera [4,9]. The genus Trypanosoma comprises nearly 500 species of

Bat Sampling
Bats were captured using ground-level mist nets (9 × 3 m, 20 mm mesh) and taken to the field laboratory. The animals were anaesthetized via intramuscular injection (ketamine chloridrathe 10% and acepromazine 1%) and submitted to a careful asepsis using bactericidal soap, iodized alcohol, and 70% alcohol, prior to blood collection via intracardiac puncture in an aseptic environment with a camp stove.

Hemoculture and Blood Clot Collection
We evaluated the trypanosomatid infection in bats by molecular characterization of the DNA obtained from three HC sample types (isolate, sediment, and first aliquot) and from the BC (Figure 2).

Bat Sampling
Bats were captured using ground-level mist nets (9 × 3 m, 20 mm mesh) and taken to the field laboratory. The animals were anaesthetized via intramuscular injection (ketamine chloridrathe 10% and acepromazine 1%) and submitted to a careful asepsis using bactericidal soap, iodized alcohol, and 70% alcohol, prior to blood collection via intracardiac puncture in an aseptic environment with a camp stove.

Hemoculture and Blood Clot Collection
We evaluated the trypanosomatid infection in bats by molecular characterization of the DNA obtained from three HC sample types (isolate, sediment, and first aliquot) and from the BC (Figure 2).  For the HC, bat blood samples were cultured in two tubes containing NNN/LIT (Novy-MacNeal-Nicolle/liver infusion tryptose overlay) and NNN/Schneider's Insect medium overlay (0.2-0.4 mL in each tube), respectively. The tubes were incubated at 28 • C and analyzed at the Laboratory of Trypanosomatid Biology every two weeks for up to four months. The cultures with sustainable growth were amplified until the stationary phase, cryopreserved, and deposited in the Coleção de Trypanosoma de Mamíferos Silvestres, Domésticos e Vetores, Coltryp/Fiocruz. These samples are called "isolates". The liquid phase of the cultures that showed failure of growth was centrifuged at 1180× g for 15 min, and the supernatant was discarded. These pellets are called "sediments". To evaluate a putative selection during parasite growth, we compared the culture before and after amplification. For this, we collected a portion of the liquid phase of positive hemocultures (n = 12) once the flagellates were detected (i.e., in their very initial growth stage), centrifuged it at 1180× g for 15 min, and discarded the supernatant. These pellets are called "first aliquot". The obtention of the HC samples is illustrated in Figure 2.
The BC was obtained by centrifugation of the blood samples at 1180× g for 15 min, followed by separation of the supernatant using sterile tips. The samples were stored at −20 • C with absolute ethanol (1:1).

Molecular Diagnosis
The technique of DNA extraction was performed according to the sample type ( Table 2). The DNA samples were resuspended in Tris-EDTA buffer (10 mM Tris-HCl pH 7.4; 1 mM EDTA pH 8.0), and the concentration and purity were quantified (OD260/OD280 ratio) using the BioPhotometer ® (Eppendorf, Hamburg, Germany). The nested-PCR targeting the 18S small subunit of the ribosomal gene (SSU rDNA) was performed using GoTaq ® Green Master Mix (Promega, Madison, WI, EUA) and the external primers TRY927F (5 GAAACAAGAAACACGGGAG 3 ) and TRY927R (5 CTACTGGGCAGC TTGGA 3 ), at a final volume of 25 µL [41]. The concentrations of the chemical components were adjusted according to the sample type (Table 2). We used 2 to 5 µL of the DNA samples to achieve a maximum of 200 µg of DNA per reaction. For the second round, the PCR products, diluted at 1:10 in ultrapure sterile water, were used as templates, following the same protocol as in the first round with the internal primers SSU561F (5 TGGGATAACAAAGGAGCA 3 ) and SSU561R (5 CTGAGACTGTAACCTCAAAGC 3 ). The amplification was performed on a Swift™ MiniPro Thermal Cycler (Esco Scientific, Singapore) with the following cycling conditions: 94 • C/3 min; 30 cycles at 94 • C/30 s, 55 • C/60 s, and 72 • C/90 s; and 72 • C/10 min. To confirm the results of some isolate samples (Table S1), we performed a PCR of another region of the SSU rDNA with the primers 609F (5 CACCCGCGGTAATTCCAGC 3 ) and 706R (5 CTGAGACTGTAACCTCAA 3 ) submitted to 30 cycles as follows: 1 min at 94 • C, 2 min at 48 • C, and 2 min at 72 • C (with an initial cycle of 3 min at 94 • C and a final cycle of 10 min at 72 • C) [42]. Ultrapure sterile water and T. cruzi DNA from positive hemoculture were used, respectively, as negative and positive controls in all reactions.
The products derived from the second-round reaction were visualized on 1.5-2.0% agarose gels stained with GelRed (Biotium, Fremont, CA, USA). The PCR from the blood clot sample of one bat (Voucher LBT 7097) resulted in double DNA bands (ca. 550 and 600 bp). The bands were collected one by one with sterile scalpel blades, and were separately analyzed. The DNA of each band was purified with Illustra GFX PCR DNA and Gel Band Purification Kit (GE Healthcare Life Sciences, Little Chalfont, Buckinghamshire, UK), according to the manufacturer's protocol.
The DNA of the PCR products derived from the isolate samples was purified using the Illustra GFX PCR DNA and Gel Band Purification Kit, according to the manufacturer's protocol. The PCR products derived from the first aliquot, the sediment, and the blood clot were directly sequenced, since the column-based purification results in loss of DNA (according to the manufacturer's protocol), and these sample types show low DNA concentration.
Both forward and reverse fragment strands of the purified DNA samples and the non-purified PCR products were sequenced at the Oswaldo Cruz Foundation Sequencing Platform facility (PDTIS/FIOCRUZ, Rio de Janeiro, Brazil). The samples were subjected to fluorescent dye-terminator cycle sequencing reactions with the ABI 3730 BigDye Terminator  Table 2.
The chromatograms of both strands were inspected and manually edited using Seq-Man Lasergene v.7.0 (DNASTAR Inc., Madison, WI, EUA). The consensus sequences were compared against the GenBank database with the BLASTn algorithm [43], considering 99.0% as the identity and coverage cut-off. Due to the high similarity among the MOTUs T. spp-Neobats [19], the identity cut-off for this group was set to 99.5%.

Phylogenetic Analysis
The sequence obtained from the BC sample of LBT 10097 was aligned to other kinetoplastid sequences retrieved from the GenBank database using the algorithm L-INS-I, available in MAFFT v.7.0 software (Japan) [44]. The alignment was inspected and manually edited on MEGA7 [45]. Maximum likelihood (ML) estimation and Bayesian inference (BI) were performed. For each analysis, the best base substitution models were chosen according to the corrected Akaike information criterion in jModelTest-2.1.10 [46]. ML reconstruction was performed using the IQ-Tree software (Vienna, Austria) [47]. For branch support, ultrafast bootstrapping [48] was performed with 5000 replicates with 1000 maximum interactions and 0.99 minimum correlation coefficients, and the SH-aLRT branch test was performed with 5000 replicates to validate the ultrafast bootstrapping result. The heuristic search method used was the program's default, and the algorithm to obtain the final tree was Neighbor Joining. Bayesian inference was performed in the MrBayes program [49,50], using the Bayesian Markov Chain Monte Carlo method to assign trypanosomatids prior to information. Four independent runs were performed for 20 million, with sampling every 2000 generations and 25% burn-in from each run. All the programs were available on the Phylosuite v1.2.2 platform (China) [51] and the reconstruction trees were visualized in FigTree v.1.4.3 software.
A pairwise distance matrix (PDM) was performed for the Eubodonida order to evaluate genetic divergence between different sequences. The Tamura-Nei parameter model, plus gamma distribution among sites (TrN + G), was used. The analysis was performed using MEGA7 software [45].

Statistical Analysis
The performances of the HC and the BC methods were compared by McNemar's test [52] using the online OMNI Calculator, available in https://www.omnicalculator.com/ statistics/mcnemars-test (accessed on 26 October 2022).
The Cohen's kappa coefficient was calculated to determine the strength of agreement between the HC results and the BC results, which was interpreted as no agreement (0-0.20), minimal agreement (0.21-0.39), weak agreement (0.40-0.59), moderate agreement (0.60-0.79), strong agreement (0.80-0.90), and almost perfect agreement (>0.90) [53]. Two evaluations were performed: (1) the agreement between negative and positive results, independently of the characterized species (n = 472); and (2) the agreement between the characterized species of both HC and BC positive results (n = 31). Data analyses were performed using the "pacman" and "vcd" packages, implemented in the R platform [54,55]. A p-value greater than 5% confidence (p < 0.05) was considered significant for all the analyses mentioned above.

Performance of the First Aliquot and the Sediment Samples
The first aliquot and the sediment samples were demonstrated to be suitable samples for the identification of cultivable and non-cultivable Trypanosoma species undetected by the conventional isolation methods in axenic media. We observed two dissimilar profiles resulting from the comparison between the first aliquot and the isolate of 12 bats: the identification of T. dionisii and T. cruzi, respectively, in the first aliquot and isolate sample. This result revealed mixed infections, and suggested positive selection of T. cruzi over T. dionisii in the axenic media (Table 3). The use of the sediments allowed for the identification of TcI, T. dionisii, T. sp. Neobat 1, and T. sp. Neobat 4 ( Table 4 and Table S1).

Performance of the HC and BC Techniques
The results of the characterization of kinetoplastids, detected by the HC samples (isolates and sediments) and the BC samples of 472 bats, are exhibited in Table 4 and  Table S1.
The molecular diagnosis of the parasitic DNA extracted from the BC was demonstrated once more to be a valuable strategy for Trypanosoma diagnosis. This technique performed better than the HC (χ 2 = 26.305; p-value < 0.0001); it showed 78% of sensitivity (116 positive bats detected through the BC from 149 positive bats detected through the BC and/or the HC) and detected 14 taxa. Meanwhile, the HC showed 41% sensitivity (61/149) and detected 10 taxa.
Nevertheless, the BC method is unable to replace the HC in terms of Trypanosoma detection. Instead, the combination of the BC method with the HC considerably enhanced the sensitivity and revealed mixed infections. It was found that 21% of negative bats by the HC were, in fact, positive by the BC. On the other hand, 8% of bats considered negative by the BC were found to be positive by the HC. The Kappa coefficient values for negative and positive results between the characterized trypanosomes were, respectively, 0.22 and 0.26 (p values < 0.05), indicating minimal agreement between these methods.

Identification of a New MOTU of the Genus Bodo
We detected a new MOTU of the order Eubodonida in one Artibeus lituratus (Stenodermatinae, Phyllostomidae) captured at the Fiocruz Atlantic Forest Biological Station. The sequence obtained from the BC sample of LBT 10097 (GenBank accession no. OP104266) showed 100% coverage and 91.85% identity (E-value = 0.0) with Bodo saltans. A phylogenetic tree based on SSU rDNA, inferred through ML and Bayesian analyses, showed LBT 10097 to be basal to one of the clades constituting the order Eubodonida (Figure 3). The genetic distances of the SSU rDNA locus among representatives of the order Eubodonida ranged from 5.3 to 19.5% (Table S2).
(Stenodermatinae, Phyllostomidae) captured at the Fiocruz Atlantic Forest Biological Station. The sequence obtained from the BC sample of LBT 10097 (GenBank accession no. OP104266) showed 100% coverage and 91.85% identity (E-value = 0.0) with Bodo saltans. A phylogenetic tree based on SSU rDNA, inferred through ML and Bayesian analyses, showed LBT 10097 to be basal to one of the clades constituting the order Eubodonida ( Figure 3). The genetic distances of the SSU rDNA locus among representatives of the order Eubodonida ranged from 5.3 to 19.5% (Table S2).

Discussion
Mixed infections by different species or genotypes of trypanosomes are a common and puzzling phenomenon [16,34,37]. Thus, it is critical to refine the diagnostic techniques and to understand to what extent these methods detect trypanosome infections. Isolation in culture media for the identification and characterization of these parasites is widely

Discussion
Mixed infections by different species or genotypes of trypanosomes are a common and puzzling phenomenon [16,34,37]. Thus, it is critical to refine the diagnostic techniques and to understand to what extent these methods detect trypanosome infections. Isolation in culture media for the identification and characterization of these parasites is widely performed, and the BC was recently recognized as an important source of Trypanosoma DNA [19]. However, the limitations of both techniques are unknown.
Our results disclose four main aspects: (1) a new MOTU of Bodo sp. in the bat species A. lituratus; (2) the potential of T. cruzi selection in the HC; (3) the advantages of employing the first aliquot and sediment samples for the diagnosis of trypanosomatids; and (4) the combination of the HC and the BC techniques, which enhances the detection of single and mixed trypanosomatid infection.
The very low identity (91.85%) of the MOTU LBT 10097 with its closest related species, B. saltans, evidences the fact that the diversity of the eubodonids is highly underestimated [3]. Moreover, the large genetic distance of the extant representatives of the order (Table S2) brings into question whether it is a novel species.
We ruled out external contamination, since bats were submitted to careful asepsis throughout blood collection, and all procedures in the field and laboratory were performed using disposable and autoclaved materials. Although the idea is mind-blowing, we cannot confirm infection by this undescribed organism, since this bat could have ingested contaminated water, and meal-derived DNA fragments can overcome degradation and enter the bloodstream, as observed in humans [56]. Moreover, even if the DNA originated from living flagellates, it is impossible to ascertain whether the infection would be sustainable. Despite this, we cannot rule out infection by Bodo in mammals, and this deserves future studies. Within the class Kinetoplastea, the eubodonids descend from orders that partially or fully include parasitic species [3]. Thus, Eubodonida could also present exceptions in their status of mode of life. Further, other free-living protozoans may cause infection in mammals, as observed in opportunistic amoebae [57].
The culturing in axenic media apparently favors T. cruzi over T. dionisii. Assuming that PCR and Sanger sequencing are biased towards the most abundant genotype [58][59][60], T. dionisii was initially the major population. This fact notwithstanding, TcI outgrew T. dionisii after the exponential growth phase. Additionally, the ubiquity of T. cruzi isolation in the HC from bats concomitantly infected with other cultivable species is intriguing (Table 4). To our knowledge, this is the first observation of T. cruzi selection over other trypanosome species by the conventional culturing method in axenic media.
Considering the low sensitivity of the HC and its selective pressure, any association between a particular trypanosome species or T. cruzi genotype and a given host species, an epizootiological scenario, or a clinical picture that was based on the encounter of a single species or genotype in the HC becomes questionable. Additionally, as the HC potentially disregards other coexistent parasites, the mixed infections and their outcomes on the course of the parasitosis are overviewed, which is the possible cause of the old-fashioned "one parasite, one disease" approach. With the improvement of the sensitivity and analytical power of diagnosis techniques, these concepts, such as the specificity and the ecology of many trypanosomatids, are being revisited [19,22,[60][61][62][63][64].
Herein, we developed an accessible strategy to improve the sensitivity of the hemoculture. Using samples in the initial growth stage of the flagellates in the HC (first aliquot) enables us to overcome the selective pressure of the culture medium. Additionally, the employment of the pellet of non-grown cultures (sediment) allows for the detection of flagellates which are incapable of growing in the axenic medium. Thus, we strongly suggest including these types of samples, which are cheap and easy to obtain, into the laboratory's routine. Furthermore, we show that even cultivable species may fail to grow, as demonstrated by the presence of T. dionisii and T. cruzi DTU TcI in the sediments. This might be explained by an initially low flagellate load in the inoculum, or variation of the growth pattern, according to TcI and T. dionisii heterogeneity [65,66].
The molecular diagnosis protocol of each sample type was adjusted according to its distinct nature (Table 2). Since parasitic DNA in the blood clot, the first aliquot, and mainly in the sediment was low, we employed extraction methods which provided high yields of good-quality DNA [19,67]; meanwhile, the DNA from the isolate samples could be extracted with a non-toxic commercial kit. The PCR protocol of the sediments differed from the other sample types for the same reason. By modifying the PCR protocol, we doubled our chances of detecting parasitic DNA.
Our results reject our hypothesis regarding the BC method as the gold standard to detect Trypanosoma infection. Instead, they show that combining the HC and the BC techniques increases the capacity for diagnosing both single and mixed trypanosomatid infections. In light of the high sensitivity of the BC method to detect trypanosomatid subpatent infections (negative HC) [19], it was surprising that half of the bats with patent infections (positive HC) showed negative results using the BC method. The DNA extraction is performed by taking 50 µL of the blood clot, which is a subsample of a sample of blood. Moreover, the hemoflagellate cells and the DNA molecules are in suspension in the blood and, thus, heterogeneously distributed in the BC. Hence, the 50 µL subsample may not be representative of the diversity present in the individual host.
Nevertheless, the BC technique displays some advantages over the HC. It shows higher sensitivity; it detects higher Trypanosoma richness, including non-cultivable species and new MOTUs [17,19]; and it requires a smaller amount of blood. In this sense, we suggest prioritizing the blood samples for BC collection in epizootiological investigations of hemoflagellates in animals with low blood volume.
The overwhelming majority of T. sp. Neobat 4 in Carollia suggests the occurrence of ecological constraints restricting T. sp. Neobat 4 distribution. Thus, investigations evaluating whether Carollia spp. share exclusive traits that might expose them to T. sp. Neobat 4 infection are needed. Furthermore, the identification of T. sp. Neobat 4 in bats belonging to distinct genera (Anoura caudifer, Glossophaga soricina, and Platyrrhinus recifinus) (Table S1) disproves the previous hypothesis about its specificity to Carollia spp. [17].
This work shows that the limitations of the techniques can bias our conclusions about the ecology of the kinetoplastids. The improvement of the analytical power of diagnostic tools will broaden our knowledge of parasite-host interactions and the diversity within this group. Accordingly, we detected a putative new species of Bodo sp. in bats. This finding will help to build knowledge on the diversity of eubodonids, and might challenge what is currently known about the mode of life of this group. We also showed positive selection of T. cruzi in the HC, which could be overcome using samples in the very initial growth stage of the HC (first aliquot). Together with the sediments, these sample types enhanced the detection of trypanosomatids. Finally, the combined usage of both BC and HC techniques enhanced the sensitivity of the diagnosis of the kinetoplastids, and revealed mixed infections.

Data Availability Statement:
The data presented in this study are openly available in the GenBank database (https://www.ncbi.nlm.nih.gov/genbank/) (accessed on 27 October 2022). The SSU rDNA accession numbers are provided in Table S1.