Introduction

Sarcocystis neurona, Sarcocystis speeri, and Sarcocystis falcatula are very similar species, transmitted exclusively in the Americas by opossums of the genus Didelphis spp. (Dubey et al. 2015). S. neurona and S. falcatula have already been described in Brazil. Birds of various species are susceptible to S. falcatula, while mammals, predominantly, are susceptible to S. neurona/S. speeri (Dubey et al. 2015). It is worth mentioning that S. neurona and S. speeri are almost identical species according to criteria based on molecular identification (Cesar et al. 2018; Acosta et al. 2018).

Sarcocystis neurona is the main agent associated with equine protozoal myeloencephalitis (EPM), an important neurologic disease of horses in the American continent (Dubey et al. 2015). In addition, neurologic disease caused by S. neurona has been occasionally described in several other terrestrial animals (Dubey et al., 2000), and this parasite has arisen as a significant cause of mortality in marine mammals (Barbosa et al. 2015). Opossums, Didelphis virginiana in North America (Fenger et al. 1995) and Didelphis albiventris in South America (Dubey et al., 2001a), are known definitive hosts of S. neurona, while intermediate hosts include domestic cats, sea otters, armadillos, raccoons, and skunks (Dubey et al., 2001b). Central nervous system (CNS) disease caused by S. neurona has been rarely described in domestic cats (Dubey et al., 1994, 2003).

Therefore, the objective of this study is to describe the clinical, pathological, immunohistochemical, and molecular findings of a fatal case of S. neurona-associated meningoencephalitis in a domestic cat in Southern Brazil.

Material and methods

Animal

A 1.7-year-old domestic cat, with free access to the street, presented a clinical history of anorexia and dyspnea, which lasted 7 days. At clinical examination, hyperthermia and pleural effusion were detected. Serological tests yielded positive results for feline leukemia virus (FeLV) and negative results for feline immunodeficiency virus (FIV). Pleural effusion was submitted for cytological evaluation, and pyogranulomatous inflammation was observed. Thoracic effusion was subjected for coronavirus immunocytochemistry, yielding negative results. Thoracic radiography was performed, and a radiopaque area in the cranial mediastinal region was noted. Due to the clinical suspicion of mediastinal lymphoma, chemotherapeutic treatment was prescribed. Treatment with prednisolone (2 mg/kg, S.I.D., P.O.) was established, associated with vincristine administrations with a 7-day interval (0.75 mg/m2, I.V.). In addition, a single administration of cyclophosphamide (250 mg/kg, P.O.), at the third week after treatment onset, was prescribed. Twenty-one days after the first clinical examination, the cat returned to the hospital presenting neurological signs, such as compulsive walking, head pressing, and vocalization. The cat died and was referred for necropsy. At the necropsy, the animal showed severe dehydration, emaciation, as well as mild cerebellar herniation through the foramen magnum. No mass consistent with lymphoma was evidenced at the mediastinal region or any other organ. In addition, no cavitary effusions or inflammatory lesions compatible com FIP were observed. At necropsy, multiple tissue samples were collected and fixed in a 10% neutral buffered formalin solution. Samples were routinely processed for histopathological examination and stained with hematoxylin and eosin (HE). Fresh frozen samples of brain and cerebellum were collected and posteriorly submitted for DNA extraction and molecular analysis. At the histological examination, lesions were restricted to the encephalon (spinal cord was not available for evaluation). In the gray and white matter and sometimes extending to the leptomeninges (Fig. 1a), there was a marked inflammatory infiltrate consisting of degenerate neutrophils, macrophages, and lymphocytes (Fig. 1b). These areas were often associated with parasitic structures measuring 10–15 μm in diameter (schizonts), which contained numerous elongated structures measuring around 1.5 × 4 μm (merozoites), morphologically consistent with S. neurona (Fig. 1c). Schizonts were observed freely in the midst of inflammation areas, as well as in the soma of neurons and in the cytoplasm of astrocytes. In addition, multifocal moderate to severe perivascular inflammatory infiltrate of lymphocytes, plasma cells, and macrophages was seen in the brain and in the cerebellum. Furthermore, multifocal areas of gliosis associated with gemistocytic astrocytes, few gitter cells, and endothelial cell hypertrophy were seen. Lesions were marked in the telencephalon, diencephalon, cerebellum, and corpus striatum, and moderate in the mesencephalon and in the brainstem.

Fig. 1
figure 1

S. neurona-associated meningoencephalitis in a domestic cat. a Cerebellum, marked diffuse mixed inflammatory infiltrate is observed distending the leptomeninges as well as partially effacing the neuropile or the trilaminar aspect of the cerebellum. HE, bar 300 μm. b Cerebellum, there is moderate inflammatory infiltrate of lymphocytes, plasmocytes and macrophages associated with parasitic schizonts (arrows). HE, bar 100 μm. c Cerebellum, multiple schizonts of S. neurona filled by numerous elongated structures (merozoites) are seen in the neuropile. HE, bar 30 μm. d Cerebellum, marked multifocal anti-S. neurona immunostaing is noted in the neuropile. Numerous schizonts and merozoites are evidenced freely in the neuropile, as well as in the cytoplasm of CNS cells. IHC, chromogen AEC, bar 45 μm

Full size image

Immunohistochemistry

CNS sections were submitted for immunohistochemistry anti-S. neurona (polyclonal antibody non-commercial, 1:200), anti-Toxoplasma gondii (VRMD, Pullman, WA, USA, dilution of 1:1000), and anti-Neospora caninum (VRMD, Pullman, WA, USA, dilution of 1:1000). Antigen retrieval was performed with proteinase K for 1 min for S. neurona and 0.1% trypsin for 10 min for N. caninum and T. gondii. Blocking of nonspecific reactions was performed with 5% skim milk for 15 min. The amplification signal was achieved by using MACH 4 Universal HRP-Polymer (Biocare, Pacheco, CA, USA) for S. neurona and LSAB-HRP Universal kit (Dakocytomation, Carpinteria, CA, USA) for N. caninum and T. gondii. The reactions were visualized with 3-amino-9-ethylcarbazole chromogen (AEC; Sigma, St. Louis, Missouri, USA), and subsequently, all slides were counterstained with Harris’ hematoxylin. Positive control samples consisted of known cases of CNS disease caused by S. neurona, T. gondii, and N. caninum. In the negative control slides, primary antibodies were replaced by PBS.

DNA extraction, PCR amplification, and sequencing

DNA was extracted from frozen samples of the brain of the infected cat with DNeasy Blood & Tissue Kit (Qiagen, Hilden, Germany) as recommended by the manufacturer’s, except for elution of the final product in 50 μL of AE buffer. The DNA samples were tested for the presence of S. neurona by means of multilocus genotyping as described below.

Sarcocystis spp. sequences of 18S small subunit rRNA region, gene coding for cytochrome c oxidase subunit I, and internal transcribed spacer 1, SAG2, SAG3, and SAG4 were nested PCR amplified. Nested PCR directed to 18S (nPCR-18S) was performed using primers 18S9L and 18S1H (Li et al. 2002). DNA amplification of Sarcocystis spp. cytochrome c oxidase subunit I (nPCR-COX1) was performed using primers designed by Gondim et al. (2019). Complete internal transcribed spacer 1 was nested PCR amplified using the primer set described by Soares et al. (2011). Finally, genetic sequences of SAG2, SAG3, and SAG4 were nested PCR amplified using the primers as designed in Monteiro et al. (2013), Valadas et al. (2016), and Gondim et al. (2019). Primers used to amplify genetic sequences of Sarcocystis spp. using nested PCR are depicted in Table 1.

Table 1 Primers for the detection of Sarcocystis spp. genetic sequences in brain tissue sample of an infected cat
Full size table

DNA amplification was done in two rounds, in all cases. The first round of amplifications (external primers) was conducted by adding 4 μL of extracted DNA, 2.5 μL of 10x PCR Buffer (KCL 50 mM; Tris HCl 10 mM; pH 9.0), 1 μL of MgCL2 (50 mM), 0.5 μL of mixed dNTPs (10 mM), 0.35 μL of each primer (10 μM), 0.2 μL of Taq DNA Polymerase Platinum-Invitrogen (5 U/μL), and 16.1 μL of ultrapure autoclaved water to a volume of 25 μL per reaction. The PCR thermal cycling consisted of an initial incubation at 94 °C for 3 min, followed by 40 cycles (94 °C for 30 s, 56 °C for 30 s, 72 °C for 50 s) and a final extension at 72 °C for 5 min. For the second rounds of amplifications (internal primers), 2 μL of template derived from the first reactions was added to 2.5 μL of 10x PCR Buffer (KCL 50 mM; Tris HCl 10 mM; pH 9.0), 1 μL of MgCL2 (50 mM), 0.5 μL of mixed dNTPs (1.25 mM), 1 μL of each primer (10 μM), 0.2 μL of Taq DNA Polymerase Platinum-Invitrogen (5 U/μI), and 16.8 μL of ultrapure autoclaved water to a volume of 25 μL per reaction. The thermal cycling was the same used in the first round.

The nested PCR products were analyzed under ultraviolet transilluminator after electrophoresis in 2% agarose gels and ethidium bromide (0.5 μg/mL) staining for 60 min. The targeted bands were excised from the gel and purified using a commercial purification kit (Illustra GFX PCR DNA and Gel Band Purification, Amersham Biosciencies) according to the manufacturer’s instructions. The purified DNA was bidirectionally sequenced using forward and reverse primers with the Kit ABI PRISM Big Dye Terminator (Applied Biosystem) following the manufacturer’s protocol.

Sequence edition and sequence analysis

The quality of nucleotide sequences, the contig assembly, and sequence edition were assessed using the program Phred-Phrad in Codon Code Aligner, version 4.2.1. The final nucleotide sequences from each locus were analyzed using Blast search tool (https://blast.ncbi.nlm.nih.gov/Blast.cgi). ITS1 and SAG sequences were aligned with homologous sequences available in GenBank using the program Clustal W in BioEdit Sequence Alignment Editor v. 7.0.5.3 (Hall 1999). ITS1 phylogenetic tree was inferred with maximum likelihood method, computing evolutionary distance using K2P + G method. The ITS1 phylogeny was inferred with sequences of S. neurona, S. speeri, S. falcatula, and related organisms retrieved from GenBank after Blast search tool using S. neurona from a cat as query. ITS1 phylogeny was reconstructed using sequences with more than 90% coverage containing at most one degenerate site by using MEGA 7 (Kumar et al. 2016). The software PopART (Population Analysis with Reticulate Trees) (Leigh and Bryant 2015) was used to infer evolutionary relationships and Integer NJ networks inference method among isolates of S. neurona and S. falcatula based on SAG loci. The genetic sequences were deposited in GenBank under the following accession numbers: SAG2 MN175964, SAG3 MN175965, SAG4 MN175966, CO1 MN175967, ITS1 MN172273, and 18S MN169125.

Results

Immunohistochemistry anti-S. neurona was strongly positive (Fig. 1d), and immunohistochemistry anti-T. gondii and anti-N. caninum were negative. Brain tissue samples were positive by all nested PCRs and the results of the multilocus analysis revealed the presence of S. neurona, which was called as S. neurona-07-2015-RS-BR. The complete nucleotide sequence of the nPCR-18S amplicon of S. neurona-07-2015-RS-BR (783 nucleotides, primers excluded) disclosed 100% identity and 100% coverage to S. speeri detected in opossums from Argentina (KT207459), and 99.87% identity and 100% coverage to S. falcatula detected in fatal Sarcocystis infection in the Rainbow Lorikeets (Trichoglossus moluccanus) (isolate Lorikeet #205850, MH626537) (Verma et al. 2018). A G-A substitution differs S. neurona-07-2015-RS-BR and S. falcatula isolate Lorikeet #205850 at nucleotide position 30 (taking MH626537 as reference). Only two single nucleotide polymorphisms differ S. falcatula isolate Lorikeet #205850 from S. speeri among 1618 paired nucleotides (G-A and C-A substitutions at nucleotide positions 30 and 1563, respectively, taking MH626537 as reference). nPCR-COX1 products of S. neurona-07-2015-RS-BR (547 nucleotides, primers excluded) differs from both S. falcatula (MH665257) and S. speeri (KT207461) at two nucleotide positions (A-T and C-A substitutions at nucleotide positions 347 and 497, respectively, taking KT207461 as reference). Curiously, S. falcatula MH665257 and S. speeri KT207461 disclosed 100% identity and 100% coverage among 1002 paired nucleotides. No gap was included in the alignments of 18S and COX1 from S. neurona-07-2015-RS-BR, S. falcatula isolate #205850, and S. speeri.

The almost complete nPCR-ITS1 amplicon was sequenced from S. neurona-07-2015-RS-BR (992 nucleotides). From the Blast search, the ITS1 sequence of S. neurona-07-2015-RS-BR is most similar to S. speeri detected in opossums in Argentina (KT207458), showing 97.89% identity, including 10 gaps. From the sequences annotated as S. falcatula, the most similar to S. neurona-07-2015-RS-BR is S. falcatula isolate Lorikeet #205850 (MH626538). These sequences share 97.08% nucleotide identity, including 16 gaps. ITS1 phylogenetic tree including S. falcatula, S. neurona, S. speeri, and S. falcatula related sequences can be found in Fig. 2. Curiously, the topology of the ITS1 tree shows that S. neurona-07-2015-RS-BR is clearly divergent from S. neurona/S. speeri. The monophyly of S. neurona/S. speeri/S. neurona-07-2015-RS-BR clade is supported with low bootstrap value (< 70, not shown).

Fig. 2
figure 2

Evolutionary relationships among S. neurona, S. speeri, and S. falcatula inferred with ITS1. The tree is rooted with S. lindsayi. Bootstraps values less than 70 are not shown. S. neurona/S. speeri in blue, S. falcatula in red. The list of taxa presented in the tree is depicted in the supplementary table

Full size image

Evolutionary relationships among S. neurona and S. falcatula inferred with SAG loci revealed that S. neurona-07-2015-RS-BR clustered closer to S. neurona (Fig. 3). In addition to S. neurona, the networks in Fig. 3 included S. falcatula obtained from budgerigars that were experimentally infected with opossum’s sporocysts (Gondim et al. 2017, Cesar et al. 2018) and S. falcatula directly detected in intermediate hosts (Konradt et al. 2017, Acosta et al. 2018). At the three SAG loci, S. neurona-07-2015-RS-BR was identical to Sarcocystis sp. genotype II derived from opossum’s sporocysts in the state of Rio Grande do Sul, Brazil, described by Monteiro et al. (2013).

Fig. 3
figure 3

Phylogenetic networks on SAG2, SAG3, and SAG4 genotypes from Sarcocystis spp. Sarcocystis neurona-07-2015-RS-BR is represented by genotype #II in all networks. S. neurona in blue, S. falcatula in red. The size of the circles is proportional to the number of samples. Mutations are shown in hatch marks. The list of taxa presented in the networks is depicted in the supplementary table

Full size image

Discussion

In the present study, the diagnosis of S. neurona-associated meningoencephalitis in a domestic cat was established through the association of the pathological, immunohistochemical, and molecular findings. Data regarding CNS disease caused by S. neurona in domestic cats are scarce. In North America, a recent case report described encephalomyelitis in an 8-year-old cat associated with S. neurona-like infection. The ITS1 sequence showed the most similarity with S. neurona and S. dasypi (Zitzer et al. 2017). The experimental infection of domestic cats by S. neurona, with subsequent formation of cysts in their muscular tissue, has been reported (Dubey et al., 2001b), which suggests that cats may act as one of the intermediate hosts of S. neurona, at least in experimental conditions. Studies conducted in Michigan (Rossano et al. 2002), Missouri (Turay et al. 2002), and Ohio (Stanek et al. 2003) revealed antibodies to S. neurona in feral cats, indicating possible exposure and natural infection by S. neurona in domestic felines with access to the outdoors in general, similarly to what was observed in the present case. However, in some of the mentioned studies, antibody cross-reactivity against other closely Sarcocystis species could not be completely rule out (Rossano et al. 2002).

Rapid reproduction after oral infection of the intermediate hosts is likely responsible for the clinical manifestations (Wünschmann et al. 2009). Possibly, the use of corticosteroids, FeLV infection, and chemotherapy had great importance in the development of the clinical condition in the present case. Similarly, other immunosuppressive factors such as lesions, surgery, and stressful events have been associated with S. neurona infection in horses (Saville et al. 2001, Cooley et al. 2007). It has been suggested that the use of corticosteroids influences the proliferation of the protozoan in the CNS (Cooley et al. 2007), which may explain the aggravation of the clinical condition.

Microscopic lesions observed in the inspected organs were restricted to the brain and cerebellum and were characterized by meningoencephalitis associated with intralesional parasitic structures compatible with S. neurona, confirmed by immunohistochemistry and PCR. Inflammatory lesions were marked in the present case; however, unlike horses, in which S. neurona is rarely observed (Dubey et al. 2015), numerous parasites were noted within areas of inflammation. Toxoplasma gondii and N. caninum, two coccidians candidates that could be also involved in neurological infections, were ruled out because nPCR-ITS1 are capable of detecting both sarocystinae and toxoplasmatinae and genetic sequences of the latter was not found. In addition, immunohistochemistry anti-N. caninum and anti-T. gondii were negative.

To our knowledge, this is the first report on molecular characterization of S. neurona detected in a case of neurologic disease affecting an animal in Brazil. Sarcocystis neurona and S. falcatula are closely related organisms exclusively transmitted within the American continents that use opossums of the genus Didelphis as definitive hosts. Despite morphological and molecular similarities, S. neurona infects mammals predominantly, while S. falcatula exclusively infects birds. Two evolutionarily distant strains have been well recognized among isolates of S. falcatula in the Americas (Valadas et al. 2016, Gondim et al. 2017, Cesar et al. 2018), and to our knowledge, a similar situation has not yet been detected in S. neurona, which appeared to be formed by only one genetic lineage. However, from the results presented here, it is probable that this diversity also occurs within S. neurona. The characterization of the ITS1 locus of the S. neurona-07-2015-RS-BR isolate revealed an ITS1 lineage, which is considerably distant from other organisms classified as S. neurona (including S. speeri). By ITS1-based analysis, S. speeri, found in sporocysts of opossums in Argentina, is almost identical to S. neurona isolates found in several other hosts in North America and the clade that congregates all species of S. neurona and S. speeri with high statistical support excludes S. neurona-07-2015-RS-BR isolate.

Although COX1-based analysis does not distinguish S. speeri and S. falcatula, both species differ from the S. neurona-07-2015-RS-BR isolate by two nucleotide substitutions. It is worth noting that two nucleotide substitutions (considering the segment homologous to nPCR-COX1 fragment) differentiate S. halieti (MH138308) from S. turdusi (KT588511), S. lari (MF596283), and S. lutrae (MG273670), which are all recognized as distinct species (data not shown).

Herein, the divergence of isolate S. neurona-07-2015-RS-BR from S. neurona was further investigated with the SAG loci and the isolate 07-2015-RS-BR reveals, although not identical, very closely related to other samples of S. neurona, which allows concluding that the sarcocystid isolate found in the cat investigated may in fact be classified as S. neurona.

From the analysis of SAG2, SAG3, and SAG4, S. neurona-07-2015-RS-BR was identical to Sarcocystis sp. identified in opossum sporocysts in the state of Rio Grande do Sul, Brazil (Monteiro et al. 2013), indicating that variants of S. neurona different from those occurring elsewhere in the Americas are being transmitted in Brazil. Studies including more samples of S. neurona would be required to test this hypothesis, as well as to assess the impact of this diversity.

Acknowledgments

We thank Dr. Marcele Bettim Bandinelli by performing immunohistochemistry. The authors thank Conselho Nacional de Desenvolvimento Científico e Tecnológico (CNPQ) and Coordenação de Aperfeiçoamento de Pessoal de Nível Superior (CAPES) for the financial support.