Sarcocystis sp. meningoencephalitis in a captive bobcat

Abstract

A geriatric captive bobcat (Lynx rufus) was euthanized due to progressive anorexia and lethargy. Meningoencephalitis with intralesional apicomplexan organisms was identified histologically. With immunohistochemistry, the organisms were immunolabeled by anti–Sarcocystis neurona antibodies. PCR targeting the ITS region of the parasite yielded an amplicon with >99.6% identity to several Sarcocystis dasypi, S. neurona, and S. speeri sequences. Amplification of the 18S region yielded a sequence that was 99.9% similar to sequences of both S. neurona (MN169125) and S. speeri (KX470746). Inflammatory disease of the CNS due to Sarcocystis sp. infection is uncommonly reported in felids and has not been reported previously in bobcats, to our knowledge. Here, we briefly review Sarcocystis-associated CNS disease in other felids, confirm that it can affect bobcats, and highlight the challenges of species-level identification of Sarcocystis sp. in routine diagnostic work.

A 12.6-kg geriatric male captive bobcat (Lynx rufus) owned by a private big cat rescue in east Tennessee was examined by the University of Tennessee Zoo Medicine service (Knoxville, TN, USA) after several days of progressive anorexia and lethargy. Four months before this examination, the bobcat and a female conspecific had been acquired from another big cat rescue in Rosamond, CA, USA, which had recently acquired the animals from a private owner in California. Shortly before transport to Tennessee, the bobcat underwent enucleation due to a punctured globe and was also treated with a course of clindamycin (75 mg PO q12h) for a tooth root abscess. It had no other known significant medical history. The bobcat was reportedly healthy one week before examination, and the female conspecific, which was housed in an adjacent enclosure, had no signs of disease. On physical examination, the bobcat was minimally responsive, and there were no other abnormal findings. It was euthanized via IV pentobarbital injection and submitted for autopsy, which was performed ~66 h after death.

At autopsy, there were no gross findings. Samples of kidney, liver, rostral cerebral hemispheres, heart, and urine were retained frozen. A comprehensive tissue set was collected and preserved in 10% neutral-buffered formalin for ~48 h before sections of the brain, lungs, heart, liver, pancreas, kidney, lymph nodes, bone marrow, and gastrointestinal tract were processed routinely for histologic examination with H&E stain. The brain sections examined were one transverse section of the ventrolateral forebrain containing portions of the thalamus, ventral temporal lobe, parahippocampal gyrus, and hippocampus, and one complete transverse section of the cerebellum and pons.

Histologically, meningoencephalitis was composed of random areas of inflammation throughout both the gray and white matter of all brain structures examined; there was rarefaction of the neuroparenchyma, gliosis, nuclear debris interpreted as necrosis, plus lymphocytes and histiocytes. Near these areas, there were rare aggregates of ovoid, ~2 × 4-μm organisms with a small nucleus, interpreted as apicomplexan parasites (Fig. 1). There were a few aggregates of lymphocytes, plasma cells, and histiocytes in the meninges of the cerebellum. All histologic changes throughout other organs were considered mild and incidental.

Figures 1–3.

Sarcocystis species infection in a bobcat. Figure 1. An area of lymphohistiocytic meningoencephalitis with apicomplexan parasites (arrow and inset) in the cerebral cortex. H&E. Figure 2. Organisms are immunolabeled by anti–Sarcocystis neurona antibody in thalamic white matter. Figure 3. A sarcocyst expands a myocyte in the tongue. H&E.

Sections of the tissue block containing the forebrain sample were submitted to the University of Georgia Veterinary Diagnostic Laboratory (Athens, GA, USA) for immunohistochemistry (IHC) with antibodies against Neospora caninum, Sarcocystis neurona, and Toxoplasma gondii, using internally validated methods (Table 1). The anti–S. neurona antibody highlighted both tightly clustered organisms interpreted as mature schizonts, and individual organisms interpreted as merozoites (Fig. 2). Organisms were not immunolabeled by the anti–T. gondii or anti–N. caninum antibodies.

Table 1.

Antibodies for immunohistochemistry with dilution, source, antigen retrieval and detection methods, and chromogen.

Antibody against:	Host	Dilution	Source	Antigen retrieval	Detection	Chromogen
Toxoplasma gondii	Goat	1:5,000	Invitrogen PAI-41098TOXOPLASMAAGENT	Protease 3	Biotinylated anti-goat and HRP-labeled streptavidin	DAB
Sarcocystis neurona	Rabbit	1:500	David Lindsay, Virginia Tech	Protease 3	Biotinylated anti-rabbit and HRP-labeled streptavidin	DAB
Neospora caninum	Goat	1:300	VMRD 210-70-NC	Proteinase K	Biotinylated anti-goat and HRP-labeled streptavidin	DAB

Positive control for each antibody is tissue with confirmed infection by the pathogen of interest.

Because protozoa were identified in the brain, sections of adrenal gland, tongue, and skeletal muscle from the hindlimb, retained in formalin for ~8 wk, were also examined histologically. The adrenal gland and hindlimb skeletal muscle were within normal limits. Sarcocysts were identified in myocytes in the tongue. These had a thin, basophilic cyst wall and were ~10–12 μm in diameter and 20–30 μm long in tangential sections. They contained tightly packed, ~2 × 4-μm bradyzoites (Fig. 3).

The frozen sample of the rostral cerebral hemispheres and formalin-fixed, paraffin-embedded (FFPE) sections of the tongue were submitted to the University of Tennessee parasitology laboratory (Knoxville, TN, USA) for PCR testing. DNA extraction and amplification from FFPE sections of the tongue was unsuccessful, but DNA was successfully extracted and amplified from the cerebral tissue. Briefly, DNA was extracted from 25 mg of cerebral tissue (DNeasy blood and tissue extraction kit; Qiagen) following the manufacturer’s instructions. A negative water control was used during the extraction. PCR was performed with PCR primers Tg18s48F and Tg18s359R, which target a portion of the 18S rRNA gene in Hammondia, Neospora, Sarcocystis, and T. gondii sp. ⁷ A 310-bp sequence most similar to Sarcocystis spp. but lacking detail to confirm species was generated. To confirm species, a PCR assay was performed targeting a portion of both the 18S rRNA gene and the ITS region of Sarcocystis spp. using both negative extraction and negative PCR controls. Primers 18S9L (GGATAACCTGGTAATTCTATG) and 18S9H (GGCAAATGCTTTCGCAGTAG) were used to amplify a 900-bp region of the 18S rRNA gene. ¹⁷ Cycling conditions were as follows: initial denaturation for 5 min at 95°C followed by 40 cycles of 94°C for 1 min, 56°C for 1 min, and 72°C for 1.5 min, and a final annealing step of 72°C for 7 min. Primers PITSF (ATTACGTCCCTGCCCTTTGT) and PITSR (GCCATTTGCGTTCAGAAATC) were used to amplify a segment of the 18S ITS region using the same cycling conditions but with an annealing temperature of 54°C. ¹⁴ PCR products were visualized on 1.5% agarose gel, and amplicon bands at the expected length were purified (ExoSAP-IT PCR product cleanup reagent; Thermo Fisher) and sequenced with Sanger sequencing at the University of Tennessee, Division of Biological Sequencing (Knoxville, TN, USA). Amplicon sequences were analyzed (Sequencher v.5.4.6; Gene Codes). Sequences were deposited in GenBank (OR800152 [18S], OR803778 [ITS]). Sequences were aligned in BioEdit, and phylogenetic trees were made using the neighbor-joining algorithm with the Kimura 2-parameter model with 500 bootstrap replicates in MegaX (v.10.1.7, https://www.megasoftware.net/; Fig. 4).

Figure 4.

Phylogenetic tree derived from the neighbor-joining algorithm with the Kimura 2-parameter model with 500 bootstrap replicates in MegaX v.10.1.7. “Bobcat ITS” is the sequence from the individual of our report.

PCR successfully amplified both the partial 18S gene and ITS region of the parasite. The 1,069-bp segment of the ITS was 99.6% identical to a S. speeri sequence in GenBank (KT207458), although it was also >99.6% identical with several S. dasypi and S. neurona sequences. The 758-bp 18S sequence was 99.9% similar to sequences of both S. neurona (MN169125) and S. speeri (KX470746). The 18S and ITS sequences generated from the cerebral tissue were not sufficient to definitively identify the protozoal organism to a species level, but we suspect that it is a variant of one of these Sarcocystis sp., or another closely related species.

• S. neurona has been comprehensively reviewed elsewhere. ¹⁰ Briefly, S. neurona causes equine protozoal myeloencephalitis and can cause CNS infections in a variety of other species. ¹⁰ Multiple strains of S. neurona have been identified, and there is abundant genetic and antigenic diversity within the species. ¹⁰ Opossums are the definitive host; proven natural or experimental intermediate hosts include the armadillo, raccoon, skunk, sea otter, and domestic cat. ¹⁰

• S. speeri causes encephalitis in mouse experiments, ¹¹ and natural meningoencephalomyelitis has been reported in an Atlantic spotted dolphin. ¹ Like S. neurona, its definitive host is the opossum, but its natural intermediate host is unknown. ¹¹ Its ITS and 18S sequences are very similar to those of S. neurona, but the 2 species can be differentiated through ultrastructure, and by the ability of S. speeri to form sarcocysts in experimentally infected mice. ¹¹ In one immunohistochemical study, anti–S. neurona antibodies failed to strongly label S. speeri organisms in experimentally infected mice. ¹² In our case, the organisms were labeled with anti–S. neurona antibody, but we do not believe that S. speeri can be ruled out because it is unknown if variations in antibody specificity or host-related or life stage–related changes in protozoal protein expression could lead to antigen–antibody cross-reactivity, which is reported among other species of Sarcocystis. ⁶

The biology and epidemiology of S. dasypi are poorly understood. The sarcocyst stage of S. dasypi was initially described in the nine-banded armadillo and was classified as a distinct species based on morphology. ¹⁶ Later, when it was discovered that opossums fed sarcocyst-laden armadillo flesh developed patent S. neurona infections, it was hypothesized that the organisms initially identified as S. dasypi were actually S. neurona, not a separate species. ⁴ However, this was never proven. At the time of this writing, there are 4 S. dasypi sequences in GenBank. Three are from an unpublished paper, and the method of initial species identification is unknown. The fourth is from an organism identified in its initial publication as “S. dasypi/S. neurona like,” which was maintained in vitro from sarcocysts isolated from armadillo muscle; the method of initial species identification is not reported. ³ Because other publications that report S. dasypi or S. dasypi–like infection use these sequences as the basis of their identification, there is ongoing confusion about the validity of S. dasypi as a separate species.^2,18

Among other felids, Sarcocystis-associated CNS disease has been reported in 4 domestic cats and 1 Canada lynx.^2,8,9,13,15 In 2 of the domestic cats and the lynx, diagnosis was based on light microscopy and IHC alone.^8,9,13 In the other 2 cats, diagnosis also included PCR amplification of selected gene regions. In one of these cases, sequences were 100% consistent with S. neurona, and the authors considered it a new subtype of S. neurona based on amplification of other gene regions. ² In the other case, the amplified sequence was similar to S. dasypi and S. neurona; the authors preferred a diagnosis of S. neurona infection, citing in part the previously described issues with the initial identification and classification of S. dasypi. ² Although S. neurona can form sarcocysts in domestic cats, ¹⁰ sarcocyst formation in muscle tissue was not reported in any of these cases. However, muscle sampling was limited to the heart, ⁹ or the heart and the tongue, ¹³ or muscle sampling was not performed or was not reported.

In domestic cats, exposure to S. neurona appears to be common in some regions of the United States, with up to 40% of cats seropositive for S. neurona antibodies. ¹⁰ However, CNS disease is rarely reported, and the factors that drive its development are unknown. Interferon-gamma (IFNγ) is likely a factor in the prevention of CNS disease given that IFN-γ knockout mice are susceptible to disease, but other immunocompromised animals, including mice and foals with severe combined immunodeficiency, and horses and cats treated with corticosteroids, are not.^2,10 Of the domestic cats reported with naturally occurring CNS disease, 3 were juveniles, 1 of which developed disease shortly after castration.^2,8,9 The other cat was a young, feline leukemia virus (FeLV)-positive adult that was treated with prednisolone for presumptive lymphoma. ¹⁵ The lynx was estimated to be 13-y-old and had chronic pyelonephritis but was FeLV and feline immunodeficiency virus (FIV) negative. ¹³

Based on our searches of the PubMed, Web of Science, and Scopus databases, using combinations of the search terms “Sarcocystis”, “bobcat”, “Felis rufus”, and “Lynx rufus”, Sarcocystis-associated meningoencephalitis has not been reported previously in a bobcat, to our knowledge. Although it is unknown whether the sarcocysts in the bobcat’s tongue were related to the organisms in the brain and could not be identified with PCR, sarcocysts identified as S. neurona and S. dasypi have been identified in skeletal muscle and tongue of wild bobcats. ¹⁸ In those cases, the authors speculated that the bobcats were infected by preying on opossums or by ingesting environmental sporocysts. Uncharacterized Sarcocystis sporozoites are also passed in bobcat feces, ¹⁹ but it is unknown whether those are the same species that encyst in muscle. However, given the complex 2-host life cycle of most Sarcocystis spp., this seems unlikely, and bobcats may serve as definitive hosts for some Sarcocystis spp., and intermediate hosts for others.

The source of infection is unknown in our case. The bobcat was housed in a repurposed circus wagon that was ~1 m off the ground, and the structure had a covered roof but open sides. It was fed raw chicken and beef; these animals are not reported to act as intermediate hosts for S. dasypi, S. neurona, or S. speeri. Opossums were known to reside in the vicinity of the enclosure, but the bobcat’s care staff did not think they had access to the enclosure. Still, contamination of the enclosure by opossum feces or carriage by transport hosts such as cockroaches, as has been reported in other species of Sarcocystis, is possible. ⁵ Predisposing factors for the development of meningoencephalitis in our case are also unknown. The other bobcat housed adjacent to our case likely had similar exposures but remained healthy. The animal had no signs of overt immunosuppression, but it was considered geriatric. It was not tested for FeLV or FIV.

Sarcocystis meningoencephalitis should be a consideration for nonspecific clinical signs in bobcats. Species-level identification of the parasite was not possible in our case using molecular tools. Although ultrastructural examination and experimental propagation and passage of these organisms have been vital in prior species-level descriptions of Sarcocystis spp., these are not practical modalities in routine diagnostic practice, and complete concordance between early descriptions of organisms and current molecular methods has not yet been achieved. Further genetic and antigenic characterization of these parasites will aid in species-level identification of these parasites, as well as elucidation of the pathogenesis of these infections.