Molecular diversity of Sarcocystis spp. in opossums (Didelphis spp.) from Southeastern and Midwestern Brazil

Abstract South American opossums (Didelphis spp.) are definitive hosts of Sarcocystis neurona, Sarcocystis speeri, Sarcocystis lindsayi and Sarcocystis falcatula. In Brazil, diverse studies have demonstrated a high frequency of Sarcocystis falcatula-like in sporocysts derived from opossums, and high genetic diversity has been observed in surface antigen-encoding genes (SAGs). In this study, genetic diversity of Sarcocystis spp. derived from Didelphis albiventris and Didelphis aurita from the cities of Campo Grande and São Paulo, was accessed by sequencing SAG2, SAG3, SAG4, the first internal transcribed spacer (ITS-1) and cytochrome c oxidase subunit I (cox1). Molecular identification was performed for 16 DNA samples obtained from sporocyst or culture-derived merozoites. The ITS-1, cox1, and SAG3 fragments were cloned, whereas SAG2 and SAG4 were sequenced directly from PCR products. Four alleles variants were found for SAG2, 13 for SAG3 and seven for SAG4, from which four, 13 and four, respectively, were novel. Twenty-seven allele variants were found for ITS-1, all phylogenetically related to S. falcatula-like previously described in Brazil. Sarcocystis sp. phylogenetically related to Sarcocystis rileyi was evidenced by cox1 in three opossums. Further studies are needed to clarify the role of Didelphis spp. as definitive hosts of Sarcocystis spp. other than that previous described.

Several genetic markers have been used to molecularly characterize Sarcocystis spp.shed by opossums in the Americas.Genome annotation of the North American S. neurona SO SN1 (Blazejewski et al., 2015) and S. neurona SN3 provided important insights into the molecular biology of the parasite.In Brazil, various studies have demonstrated the presence of Sarcocystis spp. in sporocysts derived from opossums (Casagrande et al., 2009;Monteiro et al., 2013;Gallo et al., 2018;Valadas et al., 2016;Gondim et al., 2017Gondim et al., , 2019;;Cesar et al., 2018).The majority of the Sarcocystis identified in the country have been classified as Sarcocystis falcatula-like, due to genetic characteristics and/or experimental infectivity to budgerigars (Gondim et al., 2017(Gondim et al., , 2019;;Acosta et al., 2018;Cesar et al., 2018).Extensive variability has been observed in surface antigen-encoding genes (SAGs) of Sarcocystis spp.derived from opossums in Brazil (Monteiro et al., 2013;Valadas et al., 2016;Gondim et al., 2017Gondim et al., , 2019;;Cesar et al., 2018).It has been suggested that the diversity of Sarcocystis species in the intestine of opossums could enable allele exchange through sexual recombination, contributing to their allelic variability (Monteiro et al., 2013).
Considering the widespread occurrence of Sarcocystis spp. in opossums in Brazil and the wide genetic variation observed in previous studies, this study sought to assess the genetic diversity of Sarcocystis spp. in D. albiventris and D. aurita sampled in the cities of Campo Grande (midwestern) and São Paulo (southeastern), Brazil.The detection and molecular characterization of these agents in the opossums of these regions contribute to increasing the knowledge related to the genetic diversity of Sarcocystis spp. in Brazil.
In vitro growth of Sarcocystis spp.
Sporocysts of Sarcocystis spp.were processed as previously described (Gondim et al., 2019).The antibiotic/ antimycotic solution containing sporocysts was adjusted to a minimum of 1 × 10 4 sporocysts/mL and 1 mL of the resulting solution was treated with 2.5% sodium hypochlorite for 30 min.Simultaneously, ~0.5 mL glass beads (400-600 μm in diameter, Sigma-Aldrich) were treated with 2.5% sodium hypochlorite.Sporocysts and glass beads were washed thrice with Iscove's modified Dulbecco medium (IMDM) (Invitrogen/Gibco, Carlsbad, USA) to remove hypochlorite.The sporocysts were resuspended in 700 μL of IMDM and added to a microtube containing glass beads.The mixture was vortexed at maximum speed for 3 min.A drop of the solution was observed under a light microscope (Olympus BX-51, 400x magnification) to examine the released sporozoites.The solution was filtered using a sterile 5 μm-pore size filter and subsequently inoculated into T25 flasks containing confluent monolayers of Vero cells (BCRJ: 0245).Cultures were maintained in IMDM supplemented with 1% penicillin/streptomycin/ amphotericin B and 5% inactivated bovine calf serum, in a 37 °C incubator with 5% CO 2 (Nuare, NU-4750E, Plymouth, MN, USA).Cultures without parasite propagation were discarded 60 d post-infection.

DNA extraction
DNA was extracted from culture-derived merozoites using the DNeasy blood and tissue kit (Qiagen, Valencia, California, USA), and from sporocysts using the QIAamp DNA stool mini kit (Qiagen, Valencia, California, USA), according to the manufacturer's instructions.The DNA concentration and quality were determined using a NanoDrop 2000c spectrophotometer (Thermo Scientific, San Jose, CA, USA).
Molecular detection of Sarcocystis spp.based on the ITS-1, cox1, SAG2, SAG3 and SAG4 gene fragments Conventional PCR assays were performed to amplify Sarcocystis spp.DNA of the loci encoding cox1, SAG2, and ITS-1.Nested and hemi-nested PCR assays were used to amplify loci encoding SAG3 and SAG4, respectively.Primers that amplify most of the open reading frames from SAG2 and SAG3 were used to maximize the likelihood of detecting polymorphisms at these loci.The primer sequences and cycling conditions are shown in Supplementary Table S1.
The first round of PCR was conducted in a 25-μL total reaction volume containing ~200 ng of target DNA, 0.2 mM mixed deoxynucleotide triphosphates, 3.0 mM MgCl 2 , 1.25 U Taq Platinum DNA Polymerase (Life Technologies, Carlsbad, CA, USA), 0.4 μM of each primer, 2.5 μL of 10X reaction buffer, and sterile ultra-pure water.For SAG2, SAG3 and SAG4, 2 μL of the product derived from the first amplification was used as a template in an additional round of conventional PCR, nested PCR, and hemi-nested PCR, respectively.A conventional thermocycler device (T100 Thermal Cycler, Bio-Rad, Hercules, CA, USA) was used to conduct the PCR.Ultrapure water and DNA from S. neurona merozoites obtained from the intestines of an opossum (Didelphis virginiana) (Lindsay et al., 2004) were used as negative and positive controls, respectively.PCR products were separated by electrophoresis on 1% agarose gels containing 0.5 µg/mL ethidium bromide.The gels were imaged under ultraviolet light (ChemiDoc MP Imaging System, Bio-Rad, Hercules, CA, USA) using the Image Lab Software v4.1 (Bio-Rad, Hercules, CA, USA).

Cloning and sequencing
The ITS-1, cox1, and SAG3 amplicons were cloned using the pGEM-T vector (Promega, Madison, WI, USA).Ligation reaction products were used to transform One Shot Mach1 T1 Escherichia coli cells (Invitrogen Cat # C8620-03) (109-1010 CFU/ng DNA) by thermal shocking.Plasmid DNA was extracted using the Wizard Plus SV Minipreps DNA Purification System (Promega, Madison, WI, USA).All the procedures were performed following the manufacturer's recommendations.
Up to three clones from each sample were selected for sequencing in an automatic sequencer (ABI Prism 310 Genetic Analyzer, Applied Biosystem/Perkin Elmer) (Sanger et al., 1977), using primers M13 F and M13 R (Lau et al., 2010) that flank the cloning site of the pGEM-T vector.Additionally, primers ITS-720R19 (Valadas et al., 2016) and cox1-275F22 (Gondim et al., 2019) were used to sequence an internal region of ITS-1 and cox1 fragments to increase the Phred quality of the consensus sequence.The amplicons obtained from the SAG2-and SAG4-based PCR assays were purified using EXOSAP-IT (Applied Biosystems) and sequenced with the same primer pairs used in the PCR assays.Parasitol 2023; 32(1): e014222   4/13 Sarcocystis spp. in opossums in Brazil Sequence editing Consensus sequences were obtained through analysis of electropherograms with Phred-base calling and the Phrap-assembly tool available in the suite Codoncode aligner v.4.2.1.(Codoncode Corp. Dedham, MA, USA) with a Phred quality score of ≥ 20 (99% accuracy of the base call).Consensus sequences were submitted to BLASTn (http://www.ncbi.nlm.nih.gov) to determine sequence identity by comparison with the sequences available in the GenBank database.The significance of the alignments was determined based on the E-value analysis.A BLAST search, using the newly generated sequences as a query, allowed the retrieval of homologous fragments from Sarcocystis spp. with similar lengths.

Braz J Vet
Identification of genetic relationship of Sarcocystis spp.
Genetic diversity was assessed with sequences from all five molecular markers used in this study.The sequences were aligned and used to calculate nucleotide diversity (π), polymorphism level (allele diversity [ad]), number of alleles (a), average number of nucleotide differences (K), and number of variable sites (v), using DnaSP v5.10 (Librado & Rozas, 2009).The alleles were then aligned with sequences available in the GenBank database using ClustalW (Thompson et al., 1994), and adjusted using Bioedit v. 7.0.5.3.(Hall, 1999).Phylogenetic reconstructions using the maximum likelihood (ML) method were performed on MEGA-X software (Tamura et al., 2011).The model for evolutionary distances was also calculated using the MEGA-X software and was applied to identify the most appropriate model for nucleotide substitution.The robustness of the ML tree was statistically evaluated using bootstrap analysis with 1.000 replicates.Representative genomic DNA sequences for each Sarcocystis spp.allele observed in this study were deposited in GenBank under the following accession numbers: OL830303-OL830329 (ITS-1), OL780806-OL780816 and OL780818-OL780820 (cox1), OL809965-OL809968 (SAG2), OP321544-OP321555 and OP811212 (SAG3), and OL862280-OL862286 (SAG4).

Results
Sarcocystis spp.sporocysts were observed in the intestinal scrapings of 18 out of 37 opossums: 13 D. albiventris from Campo Grande, and five D. aurita from São Paulo.From the 18 sporocysts suspensions, 12 presented more than 1 × 10 4 sporocysts/mL and were subjected to in vitro culture.Replication of Sarcocystis spp. was observed in six cultures: four inoculated with D. albiventris-derived sporozoites and two inoculated with D. aurita-derived sporozoites.Therefore, from six samples, DNA was obtained from culture-derived merozoites, and from 12 samples, DNA was obtained from sporocysts.
Sequencing could not be performed for all samples due to unsuccessful amplification or low DNA concentration (faint bands) in some PCR amplicons.At least one molecular marker was sequenced from 16 samples, and sequences from all five molecular markers were obtained from 8 samples.From two sporocyst samples none of the molecular markers were successfully sequenced.After cloning, 29 sequences were obtained from ITS-1, 39 from cox1, and 13 from SAG3.Additionally, nine sequences were obtained from SAG2 and 15 from SAG4.After genetic diversity assessment using DnaSP v5.10, 27 allele variants were obtained for ITS-1, 14 for cox1, four for SAG2, 13 for SAG3, and seven for SAG4.SAG3 was the molecular marker with the highest allele diversity (1.00 ± 0.03), nucleotide diversity per site (0.055 ± 0.00), and number of nucleotide differences between all sequences (55.4) (Table 1).Samples were named with alphanumeric code (Dal00-CG or Dau00-SP), and allele variants from each locus were designated with Arabic numerals (Table 2).
After BLASTn search for ITS-1, all 27 alleles disclosed identity ranging from 98.98-100% with a sequence of S. falcatula from a rainbow lorikeet (Trichoglossus moluccanus) from the United States (MH626538), with a query cover value of 100% (Supplementary Table S2).The phylogenetic reconstruction of ITS-1 exhibited three clades (Figure 1).The first clade was formed by the 27 alleles from this study and sequences from S. falcatula detected in rainbow lorikeet (Trichoglossus moluccanus, MH626538), brown boobies (Sula leucogaster, MW822665, MW822670), and ducks (Anas sp., OL323100), and sequences of Sarcocystis sp.detected in Magellanic penguins (Spheniscus magellanicus, MG493471) and opossums (D. virginiana, AY082647, and D. aurita/D.marsupialis, MK803362).The second clade was formed by sequences from S. neurona (MN172273, AF252407, AF081944, and AF204230) and S. speeri (KT207458).A third clade was formed exclusively with sequences from S. falcatula from the United States (AF098244, AF098242, AY082639, and AY082638).In the BLASTn analysis of cox1, alleles 8, 9, and 11 disclosed identities ranging from 96.40-96.81%with S. rileyi from a common eider (Somateria mollissima) from Norway (KJ396582), with a query cover value of 100% (Supplementary Table S3).The remaining alleles showed identities ranging from 99.69-100% with sequences from S. falcatula (MH665257) and S. speeri (KT207461), with a query cover value of 100%.The phylogeny for cox1 positioned the alleles from this study into two well-separated clades (Figure 2).Although the SAG2 and SAG3 nucleotide sequences from this study contained most of the open reading frame of the genes, the majority of the sequences available in GenBank were substantially shorter.Therefore, two phylogenetic reconstructions were performed for these loci: one with short fragments and other composed exclusively of longer sequences.The SAG2 phylogeny based on 211 bp fragment showed four clades with high statistical support (Figure 3).All alleles obtained in this study were positioned in clades along with sequences of Sarcocystis sp.detected in sporocysts shed by opossums (Didelphis spp.) and S. falcatula detected in experimentally and naturally infected birds.The SAG2 phylogenetic tree based on 645 bp (Supplementary Figure S1) fragment positioned the alleles from this study in the same clade as S. falcatula (GQ851953).The other clade was formed exclusively with sequences from S. neurona (BLASTn information on Supplementary Table S4).
For SAG3, five alleles (1, 2, 4, 5, and 10) were related to S. falcatula (GQ851956), with nucleotide identities ranging from 90.4-99.9%(Supplementary Table S5).The remaining eight alleles (3, 6, 7, 8, 9, 11, 12, and 13) were monophyletic and closely related to a sister clade exclusively formed by the S. neurona genetic sequences (Supplementary Figure S1).The alleles related to S. neurona were typically larger than the rest and showed the presence of "AT" repetitions (approximately 90 nucleotides long), which were absent in the S. falcatula SAG3-related alleles.In the phylogenetic reconstruction based on 317 bp fragment, the alleles split into six highly supported clades (Figure 3).All SAG3 sequences in this study were phylogenetically related to Sarcocystis sp. and S. falcatula, either shed by opossums and/or detected in birds.One of the six SAG3 clades was formed exclusively the S. neurona genetic sequences.The tree was inferred by using the maximum likelihood (ML) method with the evolutionary model Tamura 3-parameter with gamma distributed rates.The numbers at the nodes correspond to bootstrap values higher than 50% accessed with 1000 replicates.Sequences from Toxoplasma gondii, Neospora caninum and Hammondia heydorni were used as an outgroup (hidden).The sequences from the present study are marked in bold.
Braz J Vet Parasitol 2023; 32(1): e014222 8/13 Sarcocystis spp. in opossums in Brazil Phylogenetic reconstruction of SAG4 failed to fully resolve branching in some samples.As with other SAG phylogenies, S. neurona-derived sequences were grouped into a highly statistically supported branch.The majority of SAG4 alleles were closely related to sequences annotated as Sarcocystis sp. or S. falcatula, either detected in Didelphis spp. or birds.Allele 3 was the most divergent taxon and could not be associated with any known Sarcocystis-derived sequences (Figure 3) (BLASTn information on Supplementary Table S6).
None of the molecular markers used in this study demonstrated host specificity for different Sarcocystis lineages or alleles, as the distribution of the alleles among D. aurita and D. albiventris was random.Alleles derived from both species clustered together on several occasions.

Discussion
Previous studies have shown that S. falcatula constitutes a heterogeneous population (Marsh et al., 1999;Dubey et al., 2001d;Valadas et al., 2016;Gondim et al., 2017Gondim et al., , 2019;;Cesar et al., 2018;Llano et al., 2022).Thus far, two lineages of S. falcatula have been described in the Americas, based on ITS-1.One was described only in North America, while the other was described in both North and South America.In this study, the ITS-1-based phylogeny separated the two S. falcatula lineages into two well-supported clades.All 27 alleles obtained in this study belonged to the same lineage and positioned in a clade along with sequences from S. falcatula-like detected in naturally infected birds from Brazil (MG493471, MK803362, MW822665, MW822670, OL323100, MH626538,) and North America (AY082647).The second clade comprised S. falcatula sequences detected only in birds from North America.Most Sarcocystis found in Brazil were identified as S. falcatula-like, as they were phylogenetically related to S. falcatula and infective for birds, but had molecular differences compared with S. falcatula described in North America.Several Sarcocystis, primarily found in opossums and described as Sarcocystis spp., were after observed in natural and experimental infections in birds (Valadas et al., 2016;Gondim et al., 2017Gondim et al., , 2019;;Konradt et al., 2017;Acosta et al., 2018;Cesar et al., 2018;Gallo et al., 2018;Llano et al., 2022).
All the cox1 alleles from this study showed BLASTn identity (> 96%) with a wide variety of Sarcocystis species.Phylogenetic reconstruction positioned the alleles in two well-separated clades with high statistical support (Figure 2).The fragment from cox1 was too conserved to allow differentiation between the alleles from this study  2022) also reported the failure of cox1 to differentiate Sarcocystis species obtained from the muscle tissue of wild birds, even though the phylogeny based on ITS-1 demonstrated the presence of different species of Sarcocystis.Herein, while only Sarcocystis falcatula-like was detected in ITS-1, phylogeny based on cox1 demonstrated the presence of two species of Sarcocystis: Sarcocystis falcatula-like sequences from clade B and Sarcocystis sp.phylogenetically related to S. rileyi, positioned in clade A. The evolutionary divergence revealed between clades B and C suggests that alleles present in the later branch belong to some species not yet described in opossums.
Mixed Sarcocystis infections may occur in the intestines of opossums; therefore, more than one Sarcocystis species may be detected in DNA obtained from sporocysts (Dubey et al., 2000).Allele 11 from cox1 was identical to allele 2 from bases 1-193 and 768-972, and to alleles 8 and 9 from bases 194-767.These findings suggest that allele 11 might be a chimera, that is, an artifact sequence formed by the incorrect junction of two biological sequences.Chimeras may occur during PCR using mixed templates and do not represent sequences that exist in nature (Kalle et al., 2014).The presence of chimeras could lead to the misinterpretation of the results.Therefore, the scientific community should be aware of the need to avoid polluting public databases with artifact sequences.
Twenty-one new SAG alleles were found in comparison with those previously described: four at SAG2, 13 at SAG3 and four at SAG4.This result reinforces that Sarcocystis spp.from opossums in Brazil exhibit a diverse portfolio of surface antigen-encoding genes (Monteiro et al., 2013, Valadas et al., 2016;Cesar et al., 2018).The amplification of the entire open reading frame of SAG2 and SAG3 demonstrated that the use of primers that amplify short fragments of SAGs may be underestimating the variation in SAG alleles from Sarcocystis shed by opossums in Brazil.Most of the SAG3 alleles from this study presented insertions of "AT" repetitions of ~90 nucleotides in the region of the intron, resembling a micro-satellite pattern.Alleles presenting "AT" repetitions disclosed identity with S. neurona (96.3-96.7%),but phylogeny positioned them in a separate clade."AT" repeats have been previously described in sequences from S. neurona obtained from sea others (Enhydra lutris nereis -GQ851954 and GQ851955) and whitenosed coati (Nasua narica molaris-MF154006), and in Sarcocystis sp.types VIII, IX, and X obtained from opossums in Brazil (Valadas et al., 2016).Sequences identical to those observed in Sarcocystis sp.types VIII, IX, and X have also been found in S. falcatula infecting budgerigars (M.undulattus) (Gondim et al., 2017;Cesar et al., 2018).Interestingly, S. falcatula that present "AT" repeats were more similar to S. neurona than to the other S. falcatula.Among the three SAGs evaluated in this study, SAG3 had the highest number of alleles, similar to that previously observed by Valadas et al. (2016).The phylogeny based on SAG4 failed to fully resolve branching in some samples.However, it was possible to notice that none of the SAG4 alleles from this study was phylogenetically related to S. neurona.All the SAG4 alleles in this study were closely related to S. falcatula.
The diversity of South American fauna acting as definitive hosts for Sarcocystis spp. is higher than that of North America.A single species of opossum D. virginiana is found in North America, while, in South America, five species, grossly divided into white-eared opossums (D. albiventris, D. pernigra, and D. imperfecta) and black-eared opossums (D. aurita and D. marsupialis), have been described (Cerqueira, 1985;Lemos & Cerqueira, 2002).In this study, host specificity for different Sarcocystis species was not observed, as the distribution of the alleles among D. aurita and D. albiventris was random.This suggests that both species of opossum might present the same susceptibility as definitive hosts for different Sarcocystis species.
Although D. albiventris is considered the definitive host for S. neurona in South America, only on one occasion S. neurona was isolated from D. albiventris in Brazil (Dubey et al., 2001e).Hammerschmitt et al. (2020) reported S. neurona causing meningoencephalitis in domestic cats.The authors observed significant molecular differences related to S. neurona detected elsewhere in the Americas.At three SAG loci (2, 3, and 4), the parasite found in the cat was identical to Sarcocystis sp.genotype II derived from opossums (D. albiventris and D. aurita) in the state of Rio Grande do Sul (Monteiro et al., 2013).None of the molecular markers used in this study demonstrated the presence of S. neurona in the sampled opossums.The results observed here, along with numerous other studies performed in Brazil, imply that Sarcocystis spp. that use opossums as definitive hosts in Brazil are different from those found in North America.Furthermore, the frequency of infection caused by S. falcatula-related organisms in opossums from Brazil is much higher than that of S. neurona infection (Monteiro et al., 2013;Valadas et al., 2016;Acosta et al., 2018;Cesar et al., 2018;Gondim et al., 2017Gondim et al., , 2019)).

Conclusions
This study reported the occurrence, genetic diversity, and phylogenetic relationships of Sarcocystis spp.derived from the wild-caught opossums D. albiventris and D. aurita sampled in midwestern and southeastern Brazil.S. falcatula-like was detected by ITS-1, whereas S. falcatula-like and Sarcocystis spp.were detected by cox1.The assessment of phylogenetic inferences based on SAG2, SAG3 and SAG4 revealed the genetic richness of Sarcocystis spp.occurring among opossums and highlighted the presence of several genotypes infecting these animals in the country.

Figure 1 .
Figure1.Phylogenetic relationships within the Sarcocystis genus based on 965 bp fragment of ITS-1.The tree was inferred by using the maximum likelihood (ML) method with the evolutionary model Kimura 2-parameter with gamma distributed rates.The numbers at the nodes correspond to bootstrap values higher than 50% accessed with 1000 replicates.A sequence from Sarcocystis lindsayi was used as an outgroup.The sequences from the present study are marked in bold.

Figure 2 .
Figure 2. Phylogenetic relationships within the Sarcocystis genus based on 916 pb fragment of cox1.The tree was inferred by using the maximum likelihood (ML) method with the evolutionary model Tamura 3-parameter with gamma distributed rates.The numbers at the nodes correspond to bootstrap values higher than 50% accessed with 1000 replicates.Sequences from Toxoplasma gondii, Neospora caninum and Hammondia heydorni were used as an outgroup (hidden).The sequences from the present study are marked in bold.

Figure 3 .
Figure3.Phylogenetic relationships within the Sarcocystis genus based on SAG2, SAG3 and SAG4.The trees were inferred by using the maximum likelihood (ML) method with the evolutionary models Jukes-Cantor (SAG2 and SAG4) and Kimura 2-parameter (SAG3) with uniform rates.The numbers at the nodes correspond to bootstrap values higher than 50% accessed with 1000 replicates.The sequences from the present study are marked in bold.

Table 1 .
Polymorphism and genetic diversity of Sarcocystis spp.detected in wild-caught opossums (Didelphis albiventris and Didelphis aurita) captured in Campo Grande state of Mato Grosso do Sul, and São Paulo, state of São Paulo, Midwestern, and Southeastern Brazil, respectively.The number of nucleotides in all the sequences was standardized based on the length of the shorter sequence. *