Fatal Sarcocystis falcatula Infection in Three Penguins

Sarcocystis falcatula is a well-known cause of fatal pneumonia in some birds, particularly Old World psittacines. Here we describe fatal sarcosystosis due to S. falcatula in 3 penguins (Family Spheniscidae) under managed care, including one African penguin (Spheniscus demersus), and two Southern rockhopper penguins (Eudyptes chrysocome). Randomly distributed foci of necrosis, inflammatory cell infiltrates, edema, and variable numbers of round to elongated protozoal schizonts were observed in sections of lung. Protozoal organisms exhibited strong immunoreactivity for Sarcocystis sp. antigen by immunohistochemistry. Apicomplexan and Sarcocystis genus-specific PCR assays and sequence analysis confirmed S. falcatula as the etiologic agent. These cases of fatal pneumonia attributed to S. falcatula expand the list of aberrant intermediate avian hosts, with particular implications for penguins.

Sarcocystis falcatula is a well-known cause of fatal pneumonia in some birds, particularly Old World psittacines. Here we describe fatal sarcosystosis due to S. falcatula in 3 penguins (Family Spheniscidae) under managed care, including one African penguin (Spheniscus demersus), and two Southern rockhopper penguins (Eudyptes chrysocome). Randomly distributed foci of necrosis, inflammatory cell infiltrates, edema, and variable numbers of round to elongated protozoal schizonts were observed in sections of lung. Protozoal organisms exhibited strong immunoreactivity for Sarcocystis sp. antigen by immunohistochemistry. Apicomplexan and Sarcocystis genus-specific PCR assays and sequence analysis confirmed S. falcatula as the etiologic agent. These cases of fatal pneumonia attributed to S. falcatula expand the list of aberrant intermediate avian hosts, with particular implications for penguins.
Keywords: penguin, protozoal pneumonia, Sarcocystis falcatula, apicomplexa, Spheniscus demersus, Eudyptes chrysocome BACKGROUND Causes of protozoal pneumonia in birds include Sarcocystis falcatula, Toxoplasma gondii, Plasmodium spp., Haemoproteus spp., and Isospora (formerly Atoxoplasma) spp. Such infections are characterized by necrotizing interstitial pneumonia, with air capillaries containing tachyzoites (T. gondii) or meronts/schizonts (S. falcatula and Plasmodium spp.) within endothelial cells and/or macrophages. Malaria and toxoplasmosis are well-described causes of interstitial pneumonia in penguins (1,2). Although molecular evidence of Sarcocystis spp. infection was recently reported in Magellenic penguins in Brazil (3), fatal disease due to pulmonary sarcocystosis has not previously been described in penguins. In this case series, we report the clinical, gross, microscopic, and ultrastructural features of fatal S. falcatula infections in 3 penguins under managed care.

CASE PRESENTATIONS
Case 1 was a 27-year-old, male African penguin (Spheniscus demersus) housed at an aquarium in the United States. The bird was at the end of a molting period and had been depressed and lethargic for a couple days before being found dead. Necropsy revealed congested lungs, and a complete set of formalin-fixed tissues was submitted to the Connecticut Veterinary Medical Diagnostic Laboratory for histopathology. Opossums had been seen on the property where this penguin was housed, but not inside the exhibit.
Case 2 was a 5-year-old, female, Southern rockhopper penguin (Eudyptes chrysocome) from a different zoological institution. Due to renovation of the birds' indoor exhibit, the penguin and its flock were temporarily housed outdoors for 6 weeks in a completely meshed enclosure with access to a temperatureregulated pool. Penguins were given oral itraconazole 1 (15 mg, once daily) for aspergillosis prophylaxis for the duration of the relocation. Two days prior to its death, the bird's appetite declined and it was observed floating in water, rather than exhibiting normal swimming and diving behavior. Physical examination revealed severe dyspnea and generalized weakness. The bird was anesthetized for further diagnostics but died shortly after induction. Post-mortem radiographs revealed increased soft tissue opacity in the lungs. Necropsy revealed dark red, wet lungs that sank in formalin ( Figure 1A), a friable spleen, and edematous pericardial sac. There was a well-demarcated, 1-cmdiameter, white to yellow, raised, coelomic plaque on the inner surface of the ribs and a moderate amount of green fecal staining around the cloacal orifice. Following the death of this bird and the onset of warmer spring temperatures, the remaining flock was moved to an indoor, chilled holding area.
Five days later, a 32-year-old, female, Southern rockhopper penguin (case 3) from the same institution began to exhibit weakness, anorexia and dyspnea, and auscultation revealed harsh lung sounds. The bird was treated with ponazuril 2 (25 mg/kg PO, once), enrofloxacin 3 (15 mg/kg SC, once), meloxicam 4 (0.5 mg/kg IM, once), and furosemide 5 (0.2 mg/kg IM, once) for presumed pulmonary edema, but died 1 h later. Necropsy examination revealed dark red, wet lungs, and an enlarged, nodular spleen. The caudal aspect of the left lung contained a small, focal, tan to brown, firm area.
Cytologic examination of lung tissue imprints from case 3 showed crescent-shaped, 2 × 4-8 µm protozoal zoites ( Figure 1B), toxic heterophils and multinucleated giant cells. A complete set of tissues from cases 2 and 3 were fixed in 10% neutral-buffered formalin and submitted to the University of Georgia Zoo and Exotic Animal Pathology Service for histopathology. Fresh spleen from case 2 was submitted to Athens Veterinary Diagnostic Laboratory for aerobic and anaerobic cultures. Fresh lung and liver from Cases 2 and 3 were submitted to the Southeastern Cooperative Wildlife Disease Study (SCWDS, Athens, GA) for molecular testing.

Histopathology and Immunohistochemistry
Representative sections of all submitted tissues were routinely processed, embedded in paraffin wax, and 4-5-micronthick sections were stained with Hematoxylin and Eosin (H&E) for light microscopy. For case 1, the only relevant immunohistochemical stain available at the Connecticut Veterinary Medical Diagnostic Laboratory was a rabbit polyclonal antibody against T. gondii (BioGenex, San Ramon, CA). Additional stains for Sarcocystis sp. (rabbit polyclonal antibody) and T. gondii (rabbit polyclonal antibody) were performed at the California Animal Health & Food Safety Laboratory System, Davis, CA as previously described (4). For cases 2 and 3, immunohistochemistry was performed at the University of Georgia College of Veterinary Medicine Histology Laboratory using antibodies for S. neurona (rabbit polyclonal antibody, 1:500 dilution for 60 min), Neospora caninum (goat polyclonal antibody 6 , 1:300 dilution for 30 min) and T. gondii (mouse monoclonal antibody 6 , 1:1,0000 dilution for 10 min).
The clinical, histologic, immunohistochemical, and molecular findings for three penguins with fatal S. falcatula infection are summarized in Table 1. All penguins had severe, necrotizing and lymphohistiocytic interstitial pneumonia. Parabronchi were flooded with hemorrhage, edema, and fibrin ( Figure 1C). Air capillaries were obscured by foci of necrosis with fibrin exudation, and accumulations of heterophils, hemorrhage, and protozoal schizonts ( Figure 1D). Air spaces were multifocally lined by hypertrophied epithelial cells ( Figure 1E). In cases 2 and 3, schizonts were numerous and elongate, often conforming to the shape of capillaries ( Figure 1D, inset), while schizonts were fewer and this classic serpentine morphology of S. falcatula was not observed in case 1. In all cases, schizonts occasionally exhibited a "sunburst" arrangement, in which merozoites radiated around a centralized clearing (Figure 1E, inset).
A full set of tissues, including lung, liver, brain, skeletal muscle, and heart was examined for case 1. No extrapulmonary schizonts or sarcocysts were seen, and additional immunohistochemical stains were not pursued. Cases 2 and 3 exhibited mild lymphohistiocytic myocarditis and myositis and schizonts were observed in multiple tissues in these birds ( Table 1). Protozoa stained variably PAS-positive on Periodic acid-Schiff reaction stains, and did not stain with Giemsa. On immunohistochemistry, protozoa exhibited strong immunoreactivity ( Figure 1F) for polyclonal S. neurona containing protozoal schizonts (arrowheads), fibrin, hemorrhage, and low numbers of heterophils. The interstitium is hypercellular due to low numbers of lymphocytes and macrophages. H&E. Bar = 20 µm. Inset: High magnification image highlights the elongate to serpentine appearance of an intracellular schizont (case 3). (E) Lung, case 1. Pneumocytes lining the parabronchus in the top left of the image are hypertrophied (arrowhead) and the air space is filled with edema, fibrin, macrophages, and scant hemorrhage. The interstitium is expanded by inflammatory cell infiltrates, and multifocally obscured by necrosis, hemorrhage, and an accumulation of fibrin and edema. H&E. Bar = 20 µm. Inset: High magnification image with a schizont (arrowhead), which exhibits a "sunburst" pattern, with merozoites radiating around a clear zone. (F) Immunohistochemistry for S. neurona polyclonal antibody on lung tissue (case 2) demonstrates strongly immunoreactive, elongate to serpentine schizonts (arrowheads), free merozoites, and macrophages containing phagocytosed debris. DAB chromogen with hematoxylin counterstain. Bar = 10 µm.
antibodies and variable immunoreactivity for N. caninum and T. gondii antibodies ( Table 1). Cases 2 and 3 had markedly hypercellular spleens with foci of extramedullary granulopoiesis. There were also coalescing foci of coagulative to lytic necrosis, and case 3 had scattered, intraendothelial schizonts (Figure 2A). Under anaerobic conditions, Clostridium perfringens was cultured from the spleen of case 2. Splenic tissue from case 3 was not available for culture.
Low numbers of schizonts were present in foci of splenic necrosis for case 3, so sarcosporidiosis is the most likely explanation for this lesion. However, given that low numbers of gram-positive bacilli in one section of skeletal muscle, septicemia cannot entirely be ruled out in case 3.
All cases had multifocal lymphohistiocytic portal hepatitis and extramedullary granulopoiesis ( Figure 2B); cases 2 and 3 had low numbers of schizonts within sinusoids ( Figure 2B, right inset). All cases also had moderate numbers of sinusoidal macrophages containing fragments of erythrocytes and intracytoplasmic hemosiderin ( Figure 2B, left inset). In addition to fulminant pulmonary sarcocystosis, case 2 had a small, coelomic fungal granuloma. Aspergillosis is suspected but fungal culture and/or PCR would be required for definitive diagnosis and further testing was not pursued. Case 3 had mild air sacculitis and a focal heterophilic granuloma at the caudal aspect of the left lung lobe. This lesion is believed to be related to a prior aspiration event, given the presence of foreign material and lack of microorganisms on special stains for fungi (Gomori Methenamine Silver), bacteria (modified Brown and Brenn Gram), and acid-fast bacilli (Ziehl-Neelsen).

Transmission Electron Microscopy
Formalin-fixed lung tissue from case 3 was trimmed into two, 2-mm-thick pieces and transferred to 2% paraformaldehyde, 2% glutaraldehyde in 0.1 M phosphate buffer, pH 7.25. After overnight fixation, tissue was rinsed in 0.1 M phosphate buffer, post-fixed for 1 h with 1% buffered osmium tetroxide (OsO 4 ), then rinsed in deionized water and dehydrated in an ascending ethanol series before infiltration with propylene oxide and Mollenhauer's Epon-Araldite resin mixture (5). Tissue samples were embedded in fresh resin mixture and allowed to polymerize in a 70 • C oven. One-micron-thick sections were stained with Toluidine Blue O to select areas of interest before 60 nm sections were obtained and placed on grids. Grid sections were stained with uranyl acetate and lead citrate before examination with a JEOL JEM 1011 transmission electron microscope at 80 kV. Images were captured with an XR80M wide-angle multidiscipline mid-mount CCD camera 7 .

Polymerase Chain Reaction (PCR) Analysis and Sequencing
For case 1, formalin-fixed, paraffin-embedded (FFPE) lung tissue was submitted to the Wildlife Conservation Society (Bronx, NY) and DNA was extracted using a QIAamp DNA FFPE Tissue Kit 8 . Extracts were tested by conventional PCR using both pan-apicomplexan and Sarcocystis-specific assays targeting the 18S rRNA and internal transcribed spacer 1 (ITS-1) regions, respectively. A small portion of the 18S rRNA gene was amplified as previously described (9). A portion of the ITS-1 region was amplified using primer P-ITSF and a degenerate primer (ShortITSR; 5 ′ -GGGATTCARTKGYYGAAA-3 ′ ) designed based on publicly available Sarcocystis ITS-1 sequences (10). Amplicons were bi-directionally sequenced commercially, analyzed using Geneious R7 (Auckland, New Zealand) to generate a consensus sequence, trimmed of their primers, and analyzed by BLASTn.
The DNA sequence from the 18S gene was 100% identical to Sarcocystis falcatula in Genbank (MH626537, isolate Lorikeet ID #205850) but also identical to numerous other Sarcocystis isolates in GenBank. The DNA sequence from the ITS-1 region was 100% identical to Sarcocystis falcatula in GenBank (MH626538, isolate Lorikeet ID #205850) and Sarcocystis cf. falcatula in GenBank (AF389339) (11,12). The next closest match for the ITS-1 region was Sarcocystis speeri and Sarcocystis neurona (98-99% identity). 8  For cases 2 and 3, DNA was extracted from fresh lung and liver using a commercial DNA extraction kit 8 at SCWDS. The genus of the organism was determined by screening tissues with primers that amplify a short portion of the 18S rRNA gene of numerous apicomplexan parasites as previously described (13). Amplification products were visualized in 2% agarose gels stained with GelRed 9 . Amplicons were gel-purified using a kit 8 and bidirectionally sequenced at the University of Georgia Genomics Facility (Athens, Georgia). Chromatograms were analyzed using Geneious R7 and the consensus sequence was compared to other sequences in GenBank.
To further characterize the Sarcocystis sp. in cases 2 and 3, partial cytochrome b gene was amplified with primers CYTB-F and CYTB-R, the ITS-1 region was amplified with primers ITS-234F19 and ITS-720R19, and the partial surface antigen 2 gene was amplified with primers SAG2-F1 and SAG2-R1 as previously described (14,15). Amplicons were purified, sequenced, and analyzed as described above. To rule out the possibility of Plasmodium and/or Haemoproteus spp. infection in cases 2 and 3, a nested PCR targeting the mitochondrial cytochrome b (cytb) gene was conducted as previously described using primary primers HaemNFI and HaemNR3 and nested primers HaemF and HaemR2 (16).
For cases 2 and 3, liver and lung samples from both penguins were positive using the Tg18s58F and Tg18s348R PCR protocol and the resulting sequences (302 bp) were identical to each other and 100% similar to numerous Sarcocystis spp. The cytb gene sequences (580 bp) obtained from liver and spleen of cases 2 and 3 were identical and 100% similar to S. falcatula from captive bare-faced ibis (Phimosus infuscatus) from Brazil (KX265018) and Virginia opossum from California, USA (KP871704). The ITS-1 (287 bp) sequences were identical to each other and contained two polymorphic bases. These sequences were 98-99%

DISCUSSION
Death of these penguins was attributed to severe pneumonia caused by S. falcatula. The diagnosis was supported by IHC and confirmed by PCR. The differential diagnosis for apicomplexan pneumonia in penguins includes malaria, toxoplasmosis, and sarcocystosis. Fatal pulmonary infection with S. falcatula has been reported most often in psittacines (17)(18)(19)(20), and the current report is believed to be the first in Spheniscidae. The histopathologic features of pneumonia in these penguins were similar to that described in other avian species infected with S. falcatula, namely necrosis, edema, fibrin deposition, congestion, hemorrhage, heterophilic, and mononuclear inflammatory infiltrates, endothelial cell lysis, and pneumocyte hyperplasia (7,(20)(21)(22).
The life cycle of Sarcocystis spp. involves definitive and intermediate hosts. In North America, the definitive host for S. falcatula and the closely related S. neurona is the Virginia opossum (Didelphis virginiana), which sheds infective sporocysts in feces (8). Infection of intermediate hosts typically occurs through ingestion of food contaminated with opossum feces, but insects can serve as mechanical vectors (17). In case 1, opossums were seen on the premises but not in the bird's enclosure, so ingestion of feces or sporozoite-containing fomite(s) is considered most likely. Contaminated water run-off or insect entry into the enclosure are additional possibilities.
In contrast to case 1 (African penguin), cases 2 and 3 (Southern rockhopper penguins) had disseminated infections with large numbers of schizonts. Potential factors that may have influenced the progression of disease in these animals include the dose of sporozoites ingested, time elapsed since infection, differences in host species susceptibility or pathogen virulence, and concomitant sepsis in case 2.
Acosta et al. (3) recently provided molecular evidence of Sarcocystis spp. infection in a cohort of Magellanic penguins (Spheniscus magellanicus) in Brazil. A Sarcocystis sp. closely related to S. falcatula was isolated from the pectoral muscle of 16 penguins, which were undergoing rehabilitation and died due to other causes. Given the tissue of origin (pectoral muscle), the authors postulated that the genetic material originated from tissue cysts, suggesting that the penguins survived an acute infection and formed sarcocysts. Sarcocysts were not identified in the penguins of the current report.
Cross-reactivity between cyst-forming apicomplexans has been reported for polyclonal antibodies targeting T. gondii, N. caninum, and Sarcocystis sp. (23)(24)(25). Although a previous study has shown a lack of cross-reactivity between a polyclonal S. neurona antibody and S. falcatula schizonts in budgerigars (Melopsittacus undulatus) (26), the S. neurona antibody used in all cases strongly reacted with S. falcatula in the tissue sections, which highlights the variability that can occur between IHC results based on different polyclonal antibodies.
This report expands the list of intermediate hosts for S. falcatula and underscores the importance of excluding opossums, their feces, and potential vectors from penguin enclosures. Additional studies are needed to determine individual and species-specific susceptibilities among penguins and whether or not infection and survival confers protective humoral immunity in individual birds. Prophylactic treatment for surviving penguins may be considered if sarcocystosis is diagnosed in a collection.

DATA AVAILABILITY STATEMENT
All datasets generated for this study are included in the manuscript/supplementary files.

ETHICS STATEMENT
All animals included in this study were treated with the standard of care at each institution, as set forth by the Association of Zoos and Aquariums. An ethical review process was not required for this study, as the methods described herein were part of the routine diagnostic process to determine the cause of death of these penguins. All biological samples were obtained from penguins at postmortem exam, following the natural death of these animals.

AUTHOR CONTRIBUTIONS
SK and RB drafted the manuscript. MH provided clinical care for case 1 and performed the post mortem exam. JT and ZG provided clinical care for cases 2 and 3 and performed the post mortem exams. RB performed the histopathology on case 1. SK and RM performed the histopathology on cases 2 and 3 and TEM on case 3. RO, TS, and AW designed and performed the PCR and sequencing and analyzed the sequencing data for case 1. MY designed and performed the PCR and sequencing and analyzed the sequencing data for cases 2 and 3. All authors assisted with writing and critically reviewing the manuscript.