THE HOST INFECTIVITY
SPECTRUM OF SARCOCYSTIS NEURONA
AND SARCOCYSTIS FALCA TULA
TIMOTHY JOSEPH CUTLER
A THESIS PRESENTED TO THE GRADUATE SCHOOL
OF THE UNIVERSITY OF FLORIDA IN PARTIAL FULFILLMENT
OF THE REQUIREMENTS FOR THE DEGREE OF
MASTER OF SCIENCE
UNIVERSITY OF FLORIDA
To the horses
To those who have touched
and shaped my life
I would like to acknowledge firstly the many animals who contributed to this
project, particularly the horses, for whom it is dedicated. Secondly, I recognize all who
make animal care their life goal and passion; the work described herein was inspired by
your efforts and dedication, and aims to relieve some of the frustration of dealing with this
My long time mentor and friend, Dr. Rob MacKay, in his inimitable style has
played a large role in molding me into the clinician and researcher that I am. I am
immeasurably proud to have had the chance to work so closely with him and to have been
entrusted with these works. Dr. Ellis Greiner was a sustaining force during the wildlife
phases of my work and helped me overcome bizarre critter handling phobias. I am pleased
to have experienced the kindness and concern for which he is so famous among both
veterinary and graduate students. Dr. John Dame caused me to know considerably more
about molecular biology than I ever would want and rounded out my graduate work
experience! His calm and quiet ways were in ultimate contrast to the rodeo-style
undertakings at our horse facilities in Floridian summers. Dr. Pam Ginn appeared to be
always available and enthusiastic despite her full schedule: the necropsy protocols would
have been impossible without her dedicated work and teaching.
We wish to acknowledge the contributions of numerous student assistants without
whom it would not have been possible to achieve these lofty goals so expeditiously. Crissy
Arellano's dedication and horse sense were a tremendous asset throughout handling our
sometimes uncooperative patients. Erin LeRay has been stalwart, reliable and
uncomplaining and it has been a pleasure to watch her become an excellent horsewoman,
assistant and also a good friend. Both will be superb equine veterinarians and make us all
proud. Kristen Munsterman, Carey Gunia, Ellen Siedlecki, Michelle McCann, Robert
Porter, Shanna Hill, Dave Kern, Rick Sutliff, Miryam Hofstetter, and Roeluf Irasquain all
were valuable additions to our large crew and kept smiling throughout (I think!). It is my
sincere hope that they learned as much from me as I did from them. Karen Gillis was a
great resource and tireless in her efforts with cell culture and organization. Dr. Susan
Tanhauser was pivotal in all molecular aspects of this work, as well as an able assistant
with the horses. Dr. Rick Alleman willingly examined many CSF samples containing
various floating detritus but, I believe, has yet to see a merozoite! Dr. Siobhan Ellison was
an endless source of ideas, tissue culture discoveries and, most importantly, relief of
Orthopedic and other surgical problems in my horses were solved with the ever-
willing assistance of Dr. Jim Bryant, who also supported me very significantly during my
graduate program, for which I am tremendously thankful. I am grateful to Jay Gilbraith
and his hardworking team for the daily care of all the groups of horses that have been in
EPM projects; not all of them were angels, but all were cared for alike.
We are particularly grateful to Storm Roberts and WKTK 98.5 FM for so
generously donating public service announcements and airtime for the opossum collection
project. He made the discussion of roadkill on primetime local radio appear quite a normal
event. The 'Possum Training Institute won't be forgotten easily!
We thank the officers and technicians at Alachua County Animal Control for
assistance in collecting opossums, and for going beyond their duties to help us. We are
also grateful for samples collected and sent to us by Dr. Deb Anderson and the Wildlife
Care Center in Fort Lauderdale and from many farms in Marion County.
We acknowledge the very generous financial support from Equine Biodiagnostics
Inc. and Neogen, Inc. for testing literally thousands of western blot samples on these
horses. I am particularly grateful to Dr. Jennifer Morrow and Marcy Smith, respectively,
for answering difficult questions, for rerunning samples that were suspect and for
tolerating the interminable codes that I used to keep them confused about horse identities.
Most of this work was supported with funding from the State of Florida's Pari-
Mutuel Wagering Trust Fund for which we are grateful. The remainder of funding
through Dr. Courtney's Office is also gratefully acknowledged. Sally O'Connell and
Debbie Couch are to be commended for never allowing me to fall out of their well-policed
safety nets. I know they battled often on my behalf and saved me much effort.
We also acknowledge the gift of S. neurona merozoites (UCD-1 isolate) from Dr.
Antoinette Marsh, University of California, Davis and immunohistochemical staining of
CNS slides by Dr. Brad Barr, California Veterinary Diagnostic Laboratory System.
Much is due to Kristen Snyder, for helping me through the difficult times with
laughter you know the difference you have made. Writing this tome would have been a
very different proposition without you around.
A general apology is also due to those who will never again traverse this nation's
highways without noticing the presence, species and location coordinates of those less
successful travelers, particularly the Virginia opossum.
Finally, to those who might have read this and not found your name please know
that so many people have played important roles in this period of my life, and the space on
these pages is insufficient to recognize all the kind people and deeds. If I did my job well
all along, you already know how I feel.
TABLE OF CONTENTS
A C K N O W L E D G M E N T S .............................................................................................. iii
L IST O F T A B L E S ....................................................... ................ x
LIST OF FIGURES ............................................................. ...... ........xii
A B S T R A C T ...................................... .................................. ............... xiii
1 IN TR O D U C TIO N ................................ .................. .......... .... .. .............. 1
2 ARE SARCOCYSTIS NEURONA AND S. FALCA TULA
SYNONYMOUS? A HORSE INFECTION CHALLENGE....................... 11
Introduction ................................................................ ....... ......... 11
M materials and M ethods .................................................. ....... ............. 12
O possum Infection ...................................................... .................... ...... 12
Horse Infection.... .............. ............ .............. 13
Necropsy of Horses .................................... ..................... .......... 15
Polymerase Chain Reaction ............................. ................. 15
Western Blot Analysis ............ ................................ 16
Viability of Sporocysts ............ ............................................ .... .......... 17
R e su lts ........... ................... ................. ..................... . ..... ...... 1 7
Opossum Infection ................... ................ ......... ..... .............. 17
Horse Infection................. ................. ......... 18
Gross Pathology and Histopathology ................................... ................ 18
V ability of Sporocysts ........... ................................. ......... .............. 19
Discussion .............. ......... ....................... 19
3 EQUINE PROTOZOAL MYELOENCEPHALITIS: NASOGASTRIC
ADMINISTRATION OF SARCOCYSTIS NEURONA SPOROCYSTS
CAUSES SEROCONVERSION AND DISEASE IN HORSES ................... 27
In tro du ctio n .............................................................................. 2 7
M materials and M ethods ................... ...... ........................... ............ 28
Opossum Collection and Sporocyst Recovery.................................28
Inoculum Preparation ............................................................. ... ........... 29
Single-Dose Horse Challenge ...... ................. .................30
E nvironm ental Sentinels.......................................................... .............. 31
Multiple-Dose Horse Challenge ............ ...... ......................... 32
Western Blot Analysis ......................... ............. .............. 32
Gross and Histopathological Examination.................. ..................................... 33
R results ..................... ....................... .. .................. ............... 34
Opossum Collection and Inoculum Preparation..................... .............. 34
Single-D ose H orse Challenge .................................. ............. ................. 35
W western B lot R results ......................................................... .............. 36
Multiple-Dose Horse Challenge ............. ............................... .............. 36
CSF Analysis..... ....................................................... ....... .............. 37
D iscussion.................................................................... .......... 38
4 CHALLENGE OF PUTATIVE INTERMEDIATE HOSTS OF S.
NEURONA WITH CHARACTERIZED SPOROCYSTS .......................... 50
Introduction .............. ..... .... ... ...... ........ ................ .......... 50
Materials and Methods .............. ................................52
R esu lts ........... ....... ................ ... ...... ..................................... . 54
Gross Necropsy and Histopathology: Non-Psittacidae .............. ............... 54
Gross Necropsy and Histopathology: Psittacidae ...................................... 55
Gross Necropsy and Histopathology: Mice.............................................. 56
D iscussion............................. ...................... 57
5 DIDELPHIS VIRGINIANA: DEFINITIVE HOST OF MULTIPLE
SARCOCYSTIS SP. ................................... ................................... 74
Introduction ............................................................... .. .... ........ 74
M materials and M ethods ....................................................... .. ............... 76
O possum C collection ............................................................ ..... ..... 76
Density Purification of Sporocysts ....................................... ...............77
Excystation of Sporocysts .................................... ............................ ....... 78
Polym erase Chain R action ................................... ..................................... 78
Restriction Endonuclease Digestion ............. ......................................... 79
R results .............. ............................. ........... ............................ 80
Opossum Collection ................... ......... ............... 80
C categorization of Isolates ........... ............................................ ........ ...... 80
6 CON CLU SION S ........................................... ................. ..... .........95
A STANDARDIZED NEUROLOGIC EXAMINATION FORM............ 104
B NEUROLOGIC EXAMINATION OF THE HORSE ......................... 105
C O N LIN E R E SO U R CE S ................................................ .................. 108
LITERATURE CITED.............................. . ............ .................... 109
BIOGRAPH ICAL SKETCH ................................................. ............................ 115
LIST OF TABLES
2-1. Summary of horse experiment. Horses in the challenge group were administered
sporocysts characterized as Sarcocystisfalcatula in water by nasogastric tube. 24
3-1. Summary of single-dose horse challenge. .......................................................... 46
3-2. Summary of multiple-dose horse challenge......................................................... 47
3-3. Summary of western blot tests from laboratory 1 for horses in the single-dose
challenge, demonstrating consistency of test results. .............. .. .................47
3-4. Summary of western blot tests from laboratory 1 and laboratory 2 for horses
in the m ultiple-dose challenge .................................................. ... .............. 47
3-5. Results of western blot testing of horse #817 from laboratory 1 ............................ 48
4-1. Administration of sporocysts to brown-headed cowbirds, boat-tailed grackles,
Bobwhite quail, European starlings and redwing blackbirds. .......................... 68
4-2. Administration of sporocysts to budgerigars and summary of results..................... 69
4-3. Administration of sporocysts to NIH Swiss mice........... ........... ............. 69
4-4. Results of histopathological examination of budgerigars................. ................ 70
5-1. Number of opossums collected and shedding by month for 1997, 1998
and com bined ............................................................................. 9 1
5-2. Number of opossums collected and shedding by county in Florida for 1997
and 1998 ...................................................................... ............... 92
5-3. Numbers of isolates by Sarcocystis sp. or type with descriptive statistics ................ 92
5-4. Numbers of opossums shedding sporocysts by weight (where weight available), as a
percentage of all collected opossums of that weight, for 1997 and 1998........... 92
5-5. Numbers of opossums shedding sporocysts by weight and gender (where
weight and gender available), as a percentage of all collected opossums
per category, for 1997 and 1998 .............................................. ............. 93
5-6. Numbers of opossums shedding sporocysts by gender (where gender available), as a
percentage of all collected opossums of that gender for 1997 and 1998. .......... 93
B-1. Grades of neurologic disease ....................................................... ....... ... 107
LIST OF FIGURES
1-1. Photograph of the Virginia opossum (Didelphis virginiana). ................................. 9
1-2. Schematic of the life cycle of Sarcocystis neurona showing current state of
know ledge .......... ... ...... ............................. ....... .... .. .......... 10
2-1. Schematic drawing of horse pasture, showing location of some isolation stalls........25
2-2. Photograph of the pasture fence, showing adaptations to prevent ingress of wildlife.26
3-1. Densitometric comparison of 17 kD protein on western blot of CSF of challenged
horses compared to reference standard. Each horse was challenged on day 0..49
4-1. Photograph of brown-headed cowbirds (Molothrus ater). .............. .............. 71
4-2. Photograph of a boat-tailed grackle (Cassidix mexicanus). .................................. 72
4-3. Photograph of European starling (Sturnus vulgaris)........................................ 72
4-4. Photograph of redwing blackbird (Agelaiusphoeniceus).............. .................... 72
4-5. Photograph of Bobwhite quail (Colinus virginianus). ..................... ........... 73
4-6. Photograph of budgerigars (Melopsittacus undulatus) .................. .............. 73
5-1. Poster used to inform the public and horse farm personnel about opossum collection
program. ............................................. 94
Abstract of Thesis Presented to the Graduate School
of the University of Florida in Partial Fulfillment of the
Requirements for the Degree of Master of Science
THE HOST INFECTIVITY
SPECTRUM OF SARCOCYSTIS NEURONA
AND SARCOCYSTIS FALCA TULA
Timothy Joseph Cutler
Chairman: John B. Dame
Major Department: Department of Veterinary Medicine
Equine protozoal myeloencephalitis has been called the most important unsolved
infectious disease of horses. Its rise to this position has taken just 35 years and has
stunned the equine industries of the Americas. Seroprevalence has dramatically increased
to more than 50% but disease remains less common (<1%). Several milestones have
passed in investigating the disease during its emergence. The suggestion of synonymy
between S. neurona and S. falcatula has been disproved but caused increased scrutiny of
both parasites and development of different animal models. Our primary aim was to clarify
the biology of S. neurona and, in consequence, understand its relationships with other
opossum-borne Sarcocystis sp. Over 400 opossums were collected and screened for
sporocyst shedding to obtain isolates for host challenge. Overall, 20% (87/419) of
opossums were infected, but prevalence was highest in the 2nd quarter and lowest in the 3rd
and 4h quarters. Using published molecular markers, isolates were classified into 4 types
by DNA sequence. Together with published data, at least 3 are individual species. Sixteen
isolates were S. neurona, 13 S. falcatula, 6 Type-1085, 3 Type-2079 and 40 remain
uncategorized. Type-2079 isolates were shed with highest intensity. Type-1085 isolates
were almost all collected in South Florida and were also shed with high intensity.
Sarcocystis neurona isolates were frequently from urban areas and were intermediate in
intensity while S. falcatula was typically shed in low numbers. Horses and potential
intermediate hosts were challenged with different isolates. All S. neurona-challenged
horses seroconverted but only high doses (6 daily doses of 5 x 105 sporocyst each) caused
unequivocal neurologic disease (Grade II or worse). Weakness was more pronounced than
ataxia. The protozoan long known as S. falcatula has a benign sylvatic cycle, kills
psittaciforms at low doses but is harmless to horses. Type-1085, previously-undescribed,
behaves similarly to S. falcatula but has distinct DNA sequences. Type-2079 caused
equivocal or no disease and remains least understood. The complexity of opossum-borne
Sarcocystis biology, and the diversity of hosts that may be infected, is greater than
previously suggested. This horse model of EPM will be vital for pharmaceutical
evaluation and understanding pathogenesis of the natural disease, but requires further
development. In our opinion, it is the greatest contribution of this work. Some potential
hosts for the Sarcocystis isolates identified have been eliminated, but now 3 distinct types
have unknown intermediate hosts. The descriptive statistics should contribute to solving
the natural biology and sylvatic cycles of opossum-borne Sarcocystis sp.
Equine protozoal myeloencephalitis (EPM) is a potentially debilitating disease of
horses in the Americas (MacKay, 1997a). The disease has been recognized in the USA
since the 1960s, but may have been recognized since the 1950s in South America (Macruz
et al., 1975). Equine protozoal myeloencephalitis has now been reported in Canada,
Central and South America (Clark et al., 1981; De Barros et al., 1986: Granstrom et al.,
1992; MacKay et al., 1992). When first recognized, the disease was called Segmental
Myelitis or Focal Myelitis because lesions were found in multiple sections of the spinal
cord with clinical signs referable to the affected spinal cord segment (Rooney et al., 1970).
The agent was initially thought to be Toxoplasma gondii, (Beech and Dodd, 1974; Cusick
et al., 1974), but was reported to be a Sarcocystis species in 1980 (Simpson and
Mayhew). In 1991, Dube et al. successfully isolated and propagated in cell culture a
protozoan parasite from the spinal cord of a neurologic horse which had been diagnosed
with equine protozoal myeloencephalitis: the organism was named S. neurona.
Clinical signs of EPM are highly variable among horses. The spectrum covers
chronic progressive (usually mild) disease through to devastating peracute presentations.
Paradoxically, although horses with chronic disease suffer only quality-of-life problems,
they have the poorest chance of recovering to normal because of the duration of the
disease. Presumably, by the time of diagnosis some lesions have progressed from sites of
inflammation to sites of cell death and are irreparable. The most common presentation is a
mild gait abnormality, such as dragging a limb or failing to advance the limb in normal
cadence. Muscle atrophy locally is also common. The subtlety of disease and the broad
range of differential diagnoses make identification of the cause problematic, and as a
result, most of these cases are diagnosed after a considerable intermission. Clinical signs
in horses with acute disease may include recumbency, severe ataxia, paraplegia or
tetraplegia, depression or dementia, vestibular disease or self-inflicted injuries. Because
such cases show dramatic contrast to the normal behavior of the animal, veterinary
attention is sought quickly. Early and aggressive therapy in those cases may have equally
rapid results and return the animal to normal.
The differential diagnosis of EPM ultimately includes all causes of neurologic or
gait abnormalities of the horse. Differential diagnoses for specific cases will depend on the
time since onset, severity of condition and systems involved (e.g. locomotion versus
dysphagia). Clearly, to be convinced of the diagnosis in a disease known to vary so much
in its clinical signs, additional support would be very helpful. The first specific commercial
diagnostic test became available in 1993 in the form of a western blot for S. neurona
antibody (Granstrom, 1993; Granstrom et al., 1993). Two assumptions were necessary to
use the test reliably. First, infected horses with EPM would mount a humoral immune
response against the organism. Second, healthy horses in the general population would be
unlikely to have produced specific or cross-reactive antibodies against S. neurona (or S.
cruzi in the case of the older FIAX test). The second assumption could be met by having
a short-lived immune response or rare interactions between the horse and the organism.
This test was a great advance and very useful in the diagnosis of EPM for several years.
Since that time it has become increasingly clear that these two assumptions are no
longer valid. Sarcocystis neurona infection is both common and widespread in the United
States, with seroprevalence of -50% reported in surveys from Ohio, Oregon, and
Pennsylvania (Bentz et al., 1997; Blythe et al., 1997; Saville et al., 1997). Additionally, it
appears that the humoral response is either long-lived or that horses are re-challenged
sufficiently frequently to maintain specific or cross-reactive antibodies. Older horses are
more likely to be seropositive than young horses (Saville et al., 1997), which is compatible
with both assumptions, but it may suggest that the exposure is not dramatically higher
now than it was 5 or 10 years ago.
As EPM emerged as a greater threat to the horse industries of the Americas,
particularly the USA, a greater effort was mounted to understand the biology of the
organism S. neurona. These efforts would be essential groundwork for developing the
tools and pharmaceuticals necessary to control and eliminate EPM. The monetary
implications of this disease are enormous. The horse industries contribute $112 billion to
the US economy annually (Anonymous, 1996). It is estimated that therapy of EPM costs
$4 million per year in the state of Ohio alone (http://prevmed.vetmed.ohio-state.edu/epm-
home.html). The greatest losses are undoubtedly the majority of horses which are mildly
affected, but lose athletic ability as a consequence. The uncertain outcome with treatment
together with the expense and prolonged course, means that many horses are never treated
and may be unable to continue their working life whether as pleasure horse or racehorse.
Despite the frequent occurrence of EPM, the life cycle of the causative agent still
was unknown in 1994. Protozoa belonging to the genus Sarcocystis are obligatorily
heteroxenous and typically host-specific. Isolates from horses affected with EPM were
available (Dubey and Miller, 1986; Davis et al., 1991a, Davis et al., 1991b, Bowman et al.,
1992, Granstrom, 1992) and provided a ready source of DNA. In 1995, Fenger et al.
demonstrated 99.8% concordance between the DNA sequences of 18S rRNA genes
amplified from sporocysts shed by opossums (Didelphis virginiana) (Figure 1.1) and S.
neurona (isolate SN5) merozoites. This was the first piece of evidence to solve the life
cycle (Figure 1.2), and made a subsequent breakthrough in the intermediate host a real
Others in my laboratory hypothesized that S. neurona might be a previously known
species. Over 200 belong to the genus Sarcocystis but less than a third of those have an
identified definitive host (Levine and Tadros, 1980; Levine, 1986). The intermediate host
is more often known because sarcocysts are identified during routine necropsy or at other
examinations. Thus it was suspected that S. neurona had been described in its natural
intermediate, and possibly definitive, host. Two known Sarcocystis sp. use the opossum as
definitive host: S. falcatula, an organism that may use either brown headed cowbirds
(Molothrus ater) or boat-tailed grackles (Cassidix mexicanus) as intermediate hosts and S.
rileyi that typically cycles between the striped skunk (Mephitis mephitis, definitive host)
and any one of numerous species of Anatidae including the pintail (Anas acuta), green-
winged teal (A. carolinensis) and blue-winged teal (A. discors). Sarcocystisfalcatula is
unusual in that it can experimentally infect a variety of intermediate hosts spanning 3
orders of birds, whereas most Sarcocystis sp. use only a single species as definitive or
intermediate host. That biological behavior made S. falcatula an even more attractive
candidate to be the agent of EPM. The host spectrum of S. rileyi is not as well understood
but may be almost equally promiscuous: in addition to the skunk, the opossum (D.
virginiana, Duszynski and Box, 1978), the dog (Canisfamiliaris), and domestic cat (Felis
catus) (Golubkov, 1979) have been reported to act as definitive hosts, although the latter
2 are in some question (Levine and Tadros, 1980; Levine, 1986). Sarcocystisfalcatula
was selected for initial evaluation because of its greater host range and the presence of
large numbers of opossums and brown-headed cowbirds in Florida (Bull and Farrand,
1977). In contrast, striped skunks are relatively uncommon in Florida.
Generic primers were used to amplify a 742-bp segment of the 18S rRNA gene
from S. falcatula sarcocysts recovered from naturally infected brown-headed cowbirds.
There was >99.5% homogeneity between that sequence and the published S. neurona
(SN5 isolate) sequence (Dame et al., 1995). In fact, the S. falcatula sequence was
identical both to that reported by Fenger et al. (1995) from sporocysts isolated from 2
opossums and to another isolate of S. neurona (UCD-1) (Marsh et al., 1996a). On the
basis of these findings, it was suggested that the 2 species (S. neurona and S. falcatula)
were synonymous (Dame et al., 1995).
Box and Duszynski (1978) have demonstrated that S. falcatula can complete its
life cycle not only in its natural intermediate hosts, but also in other birds including English
sparrows (Passer domesticus), canaries (Serinus canarius) and pigeons (Columba livia).
In stark contrast to other birds, psittacines are highly susceptible to infection by S.
falcatula. The common budgerigar (Melopsittacus undulatus, Australian grass parakeet)
can develop a fatal pneumonitis after ingesting as few as 25 sporocysts per gram
bodyweight (Smith et al., 1989). Higher doses result in earlier death: 3,000 sporocysts per
gram bodyweight is fatal in 8 days. Other Psittaciforms are also susceptible to the
organism (Clubb and Frenkel, 1992, Hillyer et al., 1991). Overall the ability of this
protozoan to infect many different hosts is unusual within the genus Sarcocystis (Tadros
and Laarman, 1982). Efforts to infect rats, cats, a dog and a ferret (as definitive hosts)
were unsuccessful (Tadros and Laarman, 1982).
On the basis of these collective findings, the next logical step was to challenge
naive horses with S. falcatula. This experiment is described in chapter two. After two
rounds of challenge of the subject horses without seroconversion or disease, further
probing of the genomes revealed that, despite virtually identical 18S rRNA, S. neurona
and S. falcatula were distinct species.
Once the hypothesis of synonymy was rejected, further development of a horse
model reached an impasse. Sarcocystis neurona sporocysts could not be generated as a
single population in the laboratory because no intermediate host had been identified and
therefore the sarcocyst stage of the parasite was not available for challenge of naive
opossums. The only alternative was wild-acquired opossums that were naturally-infected
with S. neurona. The molecular tools developed to differentiate S. neurona from S.
falcatula could be used to characterize isolates, and if a small percentage of S. falcatula
were present it would be unlikely to confound results, given the results in chapter two.
An essential first step to further investigation of any host was recovery of
Sarcocystis sp. sporocysts. Several methods have been described to recover sporocysts
from the gastrointestinal epithelium of the opossum. A common step is the identification
of sporocyst shedding by examining slides prepared from material floated in Sheather's
solution. In one method (Box and Duszynski, 1978), infected gut is then stripped of fecal
material, washed with saline and the mucosa scraped from the intestinal wall into a beaker.
Scrapings are diluted in 10 volumes of digestive fluid (pepsin 0.65% w/v, NaCl 0.86%
w/v, HC1 1%, v/v in water) and stirred for 60 minutes at 38 C. The mixture is centrifuged
and washed twice in water. Sediment is resuspended in 10 volumes of 0.4% (w/v) trypsin
1-300 in Ringer's solution, pH 7.4 7.8, and digested for a further 60 90 minutes at 38
C. The final material is stored at 4 C in several times its volume of antibiotic solution
(penicillin 100 U/mL, streptomycin 100 [tg/mL, mycostatin 500 U/mL in sterile water or
phosphate buffered saline). In an alternative procedure, the mucosa is not subjected to
digestion but rather is aggressively scraped with a microscope slide and the recovered
tissue placed in an equal volume of 5.25% NaOCl (commercial bleach) for 30 minutes and
stirred several times to agitate the material. The bleach causes tissue clumps to denature
and aggregate as well as reduce the bacterial burden. Bleach is washed out by 2-3 cycles
of centrifugation and is resuspended in Hank's Buffered Saline Solution (HBSS; Gibco
BRL, Gaithersburg, Maryland) and refrigerated as above. We selected the latter method
because of convenience in handling large numbers of opossums, efficiency of sporocyst
recovery appears to be similar, and because tissues may be processed more expediently.
An additional quirk was reported for the first time in 1996. Marsh and colleagues
(1996b) recovered a Neospora sp. from a horse exhibiting classic clinical signs and
histopathology of equine protozoal myeloencephalitis. Because the disease name does not
specify an individual parasite species, no nomenclature change was necessary, but an
additional parasite was implicated. In the report by Marsh et al., the horse had a low titer
of S. neurona antibodies when tested, although the possibility of cross-reactive antibodies
could not be excluded. Subsequently, an additional clinical case of EPM caused by a
Neospora sp. has been reported (Hamir et al., 1998). In this second case the horse had
both S. neurona and Neospora sp. antibodies. The S. neurona antibodies are most likely
present as a consequence of the high seroprevalence already reported.
Despite intense investigation the literature on EPM was confusing. Few data were
unequivocal and further investigation was hampered by lack of clear direction. We felt
strongly that a better understanding of the biology of S. neurona was critical to further
progress. The purpose of the studies reported herein, therefore, are threefold: firstly, to
create a model of EPM in the horse; secondly, to examine selected species as possible
intermediate hosts, and; thirdly, during collection of materials for the first and second
objectives to record descriptive data on infected opossums. It is a prerequisite of
pharmaceutical trials and vaccine development that a characterized method is available
that can reliably induce EPM in horses. Discovery of the intermediate host would permit
both easier access to sporocysts for the horse model and development of strategies to
control the sylvatic cycle of S. neurona. Understanding the risk factors for infecting
opossums will permit some preventative measures until the intermediate hosts are
Figure 1-1. Photograph of the Virginia opossum (Didelphis virginiana).
ARE SARCOCYSTIS NEURONA AND S. FALCATULA SYNONYMOUS? A HORSE
Equine protozoal myeloencephalitis (EPM) is a potentially life-threatening
neurologic disease of horses which appears to be increasing in incidence (MacKay,
1997a). The disease has been recognized throughout the USA, as well as in Canada and
Central and South America (De Barros et al., 1986; Granstrom et al., 1992; MacKay et
al., 1992). Since the 1960s when the disease was first recognized, and throughout its
further description, much debate has surrounded the clinical diagnosis of the disease and
the identification of its etiologic agent. The agent initially was thought to be Toxoplasma
gondii, (Cusick et al., 1974), but subsequently was identified as a Sarcocystis sp.
(Simpson and Mayhew, 1980). Dubey et al. (1991) successfully isolated and propagated
in cell culture a protozoan parasite from the spinal cord of a horse diagnosed with EPM
and named it S. neurona. A western blot for S. neurona antibody was developed and
made commercially available (Granstrom, 1993). Since then, it has become increasingly
clear that S. neurona infection is common and widespread in the United States, with
seroprevalence of -50% reported in surveys from Ohio, Oregon, and Pennsylvania (Bentz
et al., 1997; Blythe et al., 1997; Saville et al., 1997). Despite the frequent occurrence of
EPM, the life cycle of the causative agent still is incompletely understood.
Because there was 99.8% concordance between the DNA sequences of 18S rRNA
genes amplified from opossum sporocysts and S. neurona (isolate SN5) merozoites,
Fenger et al. (1995) suggested that the opossum (D. virginiana) was the likely definitive
host for S. neurona. Work from our laboratory showed that there was >99.5%
homogeneity between a 742-bp segment of the 18S rRNA gene from S. falcatula, an
organism known to parasitize opossums (Box et al., 1984), and S. neurona (SN5 isolate)
(Dame et al., 1995). The S. falcatula sequence was identical both to that reported by
Fenger et al. (1995) for sporocysts isolated from 2 opossums and to another isolate of S.
neurona (UCD-1) (Marsh et al., 1996a). On the basis of these findings, it was suggested
that the 2 species (S. neurona and S. falcatula) were synonymous (Dame et al., 1995).
The objective of this study was to extend these findings by attempting to induce EPM in
horses challenged with an authenticated population of S. falcatula sporocysts.
Materials and Methods
Brown-headed cowbirds (Molothrus ater) were trapped and killed. The external
surfaces of thigh and breast muscles were examined for the presence of sarcocysts.
Brown-headed cowbird muscles containing sarcocysts were refrigerated at 4 C and were
fed to opossums within 4 hr of collection. Muscle sections were fixed in 10% neutral-
buffered formalin for later histologic examination. Sarcocysts were identified as S.
falcatula based upon sarcocyst wall morphology and host species (Box et al., 1984).
Seven hand-raised opossums (5- to 6-mo-old) were purchased and housed in
individual cages. Opossums were not shedding sporocysts according to results of fecal
flotation with Sheathers' sugar. Muscles from infected brown-headed cowbirds were fed
to opossums (1 bird/opossum). Two of the opossums were not fed brown-headed
cowbird muscle and were kept as controls. Feces were collected and examined daily.
Once sporocysts were detected, feces were collected and examined twice weekly.
Opossums were killed (Beuthanasia-D Special, Schering Plough Animal Health,
Kenilworth, New Jersey) at 42-73 days post-feeding and their gastrointestinal tracts
harvested. Sporocysts were recovered and processed as follows: small intestinal mucosal
scrapings were collected and placed in an equal volume of 5.25% NaOCl (commercial
bleach) on ice for 30 min, and were stirred every 10 min. Tissue aggregates were
removed by passage of the suspension through a gauze mesh, and then sporocysts were
washed in deionized water by 3 cycles of centrifugation at 800 g for 10 min each. The
sporocysts were resuspended in phosphate-buffered saline solution containing 100 U/mL
penicillin G and 100 [tg/mL streptomycin and stored at 4 C.
Ten horses were identified which were negative for both S. neurona antibody (as
detected by western blot analyses) and Sarcocystis sp. DNA in blood (as detected by PCR
amplification using specific primers; Dame et al., 1995). A group of seronegative horses
for this study was identified on Prince Edward Island, Canada, which is outside the known
geographical range of the opossum (Gardner, 1982). Horses were transported to Florida,
vaccinated against eastern and western equine encephalomyelitides and tetanus, dewormed
(Eqvalan; Merck AgVet, Rahway, New Jersey) and kept isolated for 14 days prior to
beginning the experiment. To minimize exposure of the horses to naturally-shed
sporocysts of S. neurona or S. falcatula, horses were fed only a complete pelleted ration
(Purina Horse Chow 100; Purina Mills, Inc., St Louis, Missouri). The pelleting process
requires heating beyond 65 C, a temperature which destroys S. gigantea sporocysts
(McKenna and Charleston, 1992). Horses were kept in a 4,600 m2 pasture that was
modified to prevent opossum access (Figure 2-1). Specifically, the pasture was
surrounded by a 1.2 m high, 5 cm by 10 cm welded-wire-mesh fence, that was buried to
15 cm below ground-surface, and had a pulsed-electric wire on the outer upper edge
(Figure 2-2). No tree limb or other structure overhung the fences or enclosure.
Horses were assigned randomly (5/group) into challenge and control groups.
Challenged horses were relocated to isolation stalls and approximately 106 sporocysts in 1
L of water were administered by nasogastric tube (Table 2-1). Horses were kept in stalls
for 7 days after challenge to ensure that pasture contamination did not occur due to
passage of unexcysted sporocysts in feces. During the 3-mo study period, blood was
drawn at least twice weekly and cerebrospinal fluid (CSF) samples were collected every
14 days. Blood for PCR was drawn every 2-3 days up to 30 days post-challenge.
Complete neurologic examinations were performed every 14 days on each horse and
recorded on videotape. Atlanto-occipital cisternal puncture was performed under short-
term injectable anesthesia with xylazine (1.1 mg/kg bodyweight; Xylaject, Phoenix
Pharmaceuticals, St Joseph, Missouri) and ketamine (2.2 mg/kg bodyweight; Ketaject,
Fort Dodge Labs Inc, Fort Dodge, Iowa) to acquire CSF samples. Horses were examined
daily and heart and respiratory rates and rectal temperature were recorded.
At the conclusion of the initial observation period (12 wk), the 4 challenged horses
that remained seronegative were reinoculated with 106 sporocysts and followed for an
additional 8 weeks.
Necropsy of Horses
The horse that developed disease was subjected to a full necropsy after 7 days of
daily intravenous dexamethasone (0.1 mg/kg) administration. Gross examination of
organs included central nervous system, lung, heart, liver, kidneys, spleen, gastrointestinal
tract and mesenteric lymph node and musculature. Representative sections were placed in
10% neutral-buffered formalin for fixation. Sections from the central nervous system were
taken as follows: 2 sections each from cervical, thoracic and lumbar spinal cord; the
occipital lobe, diencephalon at thalamus, mesencephalon at rostral colliculus,
metencephalon at pons at cerebellar peduncles, caudal medulla, and cerebellum. After
fixation, histological sections were cut at 6 |tm intervals, stained with hemotoxylin and
eosin and examined under light microscopy by a specialty pathologist. Sections with
suspect abnormalities were subjected to immunohistochemical staining (laboratory of Dr.
Brad Barr, University of California, Davis) to identify S. neurona or S. falcatula.
Polymerase Chain Reaction
Ribonucleic acid was extracted from 0.25 mL of whole blood using the Triazol
reagent kit according to manufacturer's instructions (BRL GIBCO, Life Sciences,
Gaithersburg, Maryland). RNA was resuspended in 50 [tl H20, and RT-PCR was
performed on 5 [tl aliquots using the Access RT-PCR System (Promega, Madison,
Wisconsin) employing primers JD351 (5'-CAGCCAGTCCGCCCTTTGT-3') and JD352
(5'-CATGCTGCAGTATTCAAGGCAAC-3'). These primers were designed to
specifically amplify a 160-bp segment of the 18S rRNA from S. neurona and S. falcatula.
RT-PCR was performed in a final volume of 50 [tl of AMV/Tfl reaction buffer containing
1 [tM primers, 0.2 mM dNTPs, 1 mM MgSO4, 5 units AMV reverse transcriptase and 5
units Tfl DNA polymerase. After an initial 45-min incubation at 48 C, the PCR mixtures
were subjected to 40 of the following thermocycles: 93 C for 2 min, 60 C for 30 seconds,
then 68 C for 1 min. After the last cycle, the reaction mixture was kept at 68 C for 7 min
and then cooled to 4 C and held. A positive control consisting of RNA extracted from
uninfected horse blood containing 103 S. neurona (isolate UCD-1) merozoites/mL was
included in each reaction series. A negative control was included which contained all
components of the PCR mixture except template. The possibility of contamination by
extraneous amplicons was minimized by performing reaction setup, sample extraction, and
amplification/analysis in 3 separate laboratories. Reaction tubes were closed in the room in
which the reaction mixtures were prepared and only opened after amplification. PCR
products were analyzed by electrophoresis using a 1% agarose gel (NuSieve 3:1, FMC,
Western Blot Analysis
Serum and CSF samples were submitted to a commercial diagnostic laboratory
(Equine Biodiagnostics Inc., Lexington, Kentucky) and analyzed as previously described
(Granstrom et al., 1993). Test results are reported by the laboratory as either negative,
suspect-positive or positive. It is our observation that samples classified as "suspect-
positive" frequently are negative when retested; therefore, we report as positive only those
samples so classified by the diagnostic laboratory. All other samples are considered S.
Viability of Sporocysts
The pool of sporocysts used for challenge of experimental horses was examined
for excystation. Aliquots of 10,000 sporocysts each were pelleted by centrifugation and
resuspended in undiluted equine bile (pH 7.0) containing 2% w/v trypsin (Sigma T-8642,
Sigma Chemical Company, St. Louis, Missouri). Sporocysts were then incubated for 4 hr
at 37 C for maximal excystation. Counts were performed on a hemocytometer at 200 x
magnification by evaluating 5 fields of 100 sporocysts each and counting the sporocyst
wall fragments. To test whether sporocysts from the same lot given to horses were
biologically active, 500 sporocysts were administered orally to each of 5 brown-headed
cowbirds. Birds were maintained in a 2 m by 2 m by 2 m wire enclosure and fed crushed
corn. After 90 days, cowbirds were killed and examined in an identical manner to the
naturally-infected cowbirds described previously.
Beginning 8-10 days after infected muscles were fed, sporocysts were found in the
feces of all challenged opossums. In contrast, the unchallenged opossums remained
negative. Opossums shed sporocysts until they were killed (range 42-73 days after
challenge). Maximal shedding occurred on day 72 (range 26-112) post-feeding.
Control horses remained negative for S. neurona antibody and Sarcocystis sp.
DNA in both blood and CSF. One of the experimental horses (#56) had S. neurona
antibody detected on western blot. When additional samples were analyzed, it was
discovered that horse 56 had seroconverted prior to the administration of any sporocysts,
i.e., was naturally exposed and infected before challenge. Sarcocystis neurona antibody
was frequently, but not invariably, detected in serum and CSF from that horse for the
remainder of the project. In addition, beginning on day 28, horse 56 showed progressive
clinical signs typical of EPM, including moderate to severe ataxia and weakness of the
limbs, generalized muscular atrophy, and obvious weight loss. The remaining S. falcatula-
challenged horses did not develop clinical signs or have S. neurona antibody or DNA
detected in blood or CSF. After the second challenge of these horses with S. falcatula
sporocysts, there still was no clinical or other evidence of infection or disease. All positive
controls for the PCR process yielded appropriately-sized DNA fragments, whereas no
negative control amplified this band on any occasion.
The nucleated cell count in CSF remained within reference range (<8 nucleated
cells/|tl) at all examination times. Repeated CSF collection did not appear to induce an
inflammatory response in the CNS.
Gross Pathology and Histopathology
Only horse 56 was killed because it was apparent that, at most, only equivocal
evidence of neurologic disease was present in other horses. The only abnormal finding on
gross examination was thin condition. Histopathological examination of lung, kidney,
skeletal muscle, myocardium, tongue, esophagus, pancreas, stomach, colon, duodenum,
ileum and cecum revealed no lesions. There was moderate diffuse degeneration of the
liver parenchyma. Rarely, kidney tubule cells contained mineralized material and scattered
glomeruli had thickened basement membranes. Mesenteric lymph nodes showed severe
depletion of cells throughout. A similar, less severe, pattern was present in the spleen. All
brain sections were normal except some vacuoles contained axonal debris in white matter
of the medulla. In the spinal cord, lesions were present at all levels (cervical, thoracic and
lumbar) but were most pronounced in the mid-cervical area (C3 and C4). Changes
primarily comprised vacuolar degeneration of white matter (predominantly lateral and
ventral) and degenerating nerve cell bodies in gray matter, with occasional mild
hemorrhages in the lumbar spinal cord. Immunohistochemical staining (for S. neurona
and for S. falcatula) was negative on all sections examined.
Viability of Sporocysts
All brown-headed cowbirds (N =5) given sporocysts from the same pool used for
horse challenge had sarcocysts present when examined 90 days after infection. Each
experimentally-infected bird had a large number of sarcocysts in all skeletal muscles,
particularly the breast and thigh muscles. Excystation of sporocysts was consistently 30-
We believe this experiment casts considerable doubt over the previous hypothesis
that S. neurona and S. falcatula are synonymous (Dame et al., 1995). Challenge of horses
with 106 S. falcatula sporocysts resulted in neither seroconversion against S. neurona nor
clinical signs of EPM. The results from the second challenge of 4 horses from the same
experimental group were likewise negative. Control experiments demonstrated that the
sporocysts were viable and infectious.
The hypothesis of synonymy was based on finding virtual sequence identity of the
18S rRNA gene from S. falcatula bradyzoites and S. neurona merozoites. Typically,
divergence of the 18S rRNA gene is slow and closely related species might have identical
18S rRNA gene sequences (Hillis and Dixon, 1991). Subsequent analyses in our
laboratory, however, have identified several DNA sequence differences elsewhere in the
genome which discriminate between isolates of S. falcatula and S. neurona (Tanhauser et
al., 1999). Further, the sarcocysts and sporocysts used in the various challenge
experiments in this report have been examined by these additional criteria and verified to
be solely S. falcatula (Tanhauser et al., 1999).
The failure of any control horse to seroconvert against S. neurona indicates that
when substantive measures are taken to block exposure of horses to wildlife and
sporocysts, infection of horses with S. neurona can be prevented. Although 1
experimental horse did develop clinical EPM, the presence of S. neurona antibody on the
day of challenge with S. falcatula sporocysts indicates strongly that this was a natural
infection acquired prior to introduction to the protected enclosure. Sarcocystis sp. rRNA
was not amplified by RT-PCR from this horse's blood, which suggests either that the agent
may already have completed its blood-borne phase, or that the parasitemia was below the
limit of detection of our RT-PCR test. In contrast, it was consistently possible to amplify
RNA from 103 S. neurona merozoites (UCD-1) grown in cell culture added to normal
horse blood. Horizontal transmission between horses would not be expected because
Sarcocystis species are obligatorily heteroxenous (Dubey et al., 1989). Histopathological
findings on horse 56 were compatible with EPM, although no parasite was seen
microscopically. Immunohistochemical staining of spinal cord did not identify any S.
neurona or S. falcatula antigen. Together these findings are still most compatible with a
natural occurrence of EPM.
Seroprevalence studies conducted in Ohio, Pennsylvania, and Oregon suggest that
the prevalence of S. neurona infection may be approximately 50% throughout much of the
USA (Bentz et al., 1997; Blythe et al., 1997; Saville et al., 1997). We found that all 29
horses kept in college pastures close to our own were seropositive, suggesting that
prevalence may approach 100% in the area surrounding our experimental enclosure. A
further testament to the protective efficacy of our enclosure was the finding that, after the
second challenge experiment was complete, when experimental horses were transferred to
another pasture, all seroconverted within 30 days. The fact that horses remained
seronegative despite unfettered exposure to birds and insects from outside the enclosure
additionally indicates that these species did not act efficiently as transport hosts for S.
neurona sporocysts under these conditions.
Our results are somewhat in conflict with those reported by Fenger et al. (1997a).
In the previous study, pooled sporocysts obtained from 10 wild opossums behaved like
infective S. neurona in that they induced both seroconversion to S. neurona and
neurologic signs in most subject horses. The same inoculum also was S. falcatula-like in
that it killed budgerigars (Melopsitticus undulatus) and caused sarcocyst development in
sparrows (Passer domesticus). These results can be interpreted to mean either that the 2
species (S. neurona and S. falcatula) are the same, as previously suggested, (Dame et al.,
1995) or that they are different and both were present in the pooled inoculum used for
horse and bird challenges. On the basis of the study we report here, we believe that the
latter interpretation is correct, and that there is compelling evidence that S. falcatula
sporocysts are harmless to horses, whereas S. neurona sporocysts cause the disease we
recognize clinically as EPM. Clearly, both species use the opossum as definitive host, but
then complete their life cycles in different intermediate hosts. It is possible that other as
yet unidentified Sarcocystis sp. also use the opossum as a definitive host, and thus
epidemiological work aimed at defining risk factors cannot merely identify sporocysts in
opossum feces, but will need to characterize the species with DNA-based or other tests.
The opossum has been shown to support transient infection with S. rileyi although the
striped skunk (Mephitis mephitis) is the recognized definitive host in the wild (Duszynski
and Box, 1978; Levine, 1986).
Currently, there is no evidence to indicate whether EPM is more likely to result
from single or multiple episodes of ingestion of infective sporocysts. We chose our
challenge dose to be larger than the single LD50 oral challenge doses for S. cruzi (200,000
sporocysts; Fayer and Dubey, 1986) and S. tenella (100,000 sporocysts; Leek et al., 1977)
previously given to cattle and sheep, respectively. The dose of 106 sporocysts also was
considerably greater than the total number of sporocysts shed in a single day by any
opossum during our experimental S. falcatula experiment. Fenger et al., (1997a) reported
that seroconversion and neurologic signs were induced when large numbers of opossum-
shed sporocysts (1.2-4 x 107) were given to foals. Seroconversion occurred after doses of
106 and 2 x 106 sporocysts. There was a trend toward earlier appearance of antibody and
disease with larger doses, although these results are not strictly comparable to our
experiment as previously discussed. We believe strongly that a single dose of 106
sporocysts should be sufficient to cause infection and perhaps even disease in a species
that naturally can function as host.
The results of this work are important. We present strong biological evidence that
S. neurona and S. falcatula are not synonymous. The data suggest that S. falcatula does
not infect horses. This difference reopens the fundamental question of which vertebrate
species act as intermediate host for S. neurona. This is practically important because
pharmacological studies and vaccine trials will require highly-characterized sporocysts
produced in the laboratory under controlled conditions.
Table 2-1. Summary of horse experiment. Horses in the challenge group were
administered sporocysts characterized as Sarcocystisfalcatula in water by nasogastric
Horse ID# Breed Age (years) Gender Group Sporocysts Neurologic
(x 106) disease
50 Std 5 g Control 0 No
51 Std 3 f Control 0 No
52 Std 3 f Challenged 1.0 No
53 Std 6 f Control 0 No
54 Std 3 f Challenged 1.0 No
55 Std 4 g Challenged 1.0 No
56 Std 4 f Challenged 1.0 Yes
57 Std 3 g Challenged 1.0 No
58 Std 3 f Control 0 No
59 TB 3 g Control 0 No
Std = Standardbred, TB = thoroughbred; f = filly, g = gelding.
Figure 2-1. Schematic drawing of horse pasture, showing location of some isolation stalls.
147.91' 1 3i stalls
1 49 17' '7 74 building
Figure 2-2. Photograph of the pasture fence, showing adaptations to prevent ingress of
, !/ ^
EQUINE PROTOZOAL MYELOENCEPHALITIS: NASOGASTRIC
ADMINISTRATION OF SARCOCYSTISNEURONA SPOROCYSTS CAUSES
SEROCONVERSION AND DISEASE IN HORSES
It has previously been demonstrated that the administration of characterized S.
falcatula sporocysts to horses does not induce seroconversion against S. neurona nor
does it induce disease (Cutler et al., 1999). In a similar horse-challenge experiment using
an uncharacterized mixture of sporocysts from wild-caught opossums, disease and
seroconversion occurred in challenged horses and disease also occurred in challenged
budgerigars (Fenger et al., 1997a). This apparent paradox is resolved if a mixture of
Sarcocystis sp. sporocysts was present in the inoculum for the latter experiment. Ideally, a
new horse challenge is indicated utilizing biologically-purified S. neurona sporocysts to
fulfill Koch's postulates. The intermediate host of S. neurona is not known, however, and
therefore it is not possible to generate a single population of S. neurona sporocysts by
infecting naive laboratory animals. The closest alternative is to collect wild opossums,
screen for infection, and then characterize sporocyst isolates using molecular markers.
This technique does not exclude the presence of other species at low concentrations.
However, the risk of confounding the results by accidentally including S. falcatula is
minimal in light of the negative data obtained previously when horses were challenged
with S. falcatula sporocysts.
Once a highly characterized method for the experimental induction of EPM exists
in horses, further progress can be made in understanding the pathogenesis of the natural
disease. Refinements in challenge dose and administration protocol would permit better
characterization of the risks of developing EPM and identify horse populations placed at
greatest risk by athletic, or other, activity. Ultimately, repeatable and dependable induction
of EPM in horses is an essential prerequisite to further progress in understanding and
eliminating this disease.
Materials and Methods
Opossum Collection and Sporocyst Recovery
Opossums were acquired either as roadkill or were trapped and euthanized.
Gender and weight were recorded and a unique identification number was assigned. Some
samples were provided to my laboratory as intestinal tracts only, (without other
information). Necropsies were performed as soon as possible after collection. The
gastrointestinal tract was isolated and removed intact through an incision of the midline
ventral abdominal wall. The mesentery was stripped off and the intestinal contents milked
into a clean, labeled container. A section of distal ileum was incised, flattened and the
mucosal surface exposed. Excess ingesta were washed off the mucosa with deionized
water and then mucosa was collected by scraping with a microscope slide at a 45 degree
angle. A 2-g sample of feces and a 1-mL sample of intestinal-mucosal scrapings were
taken from the original samples and each was homogenized in 10 mL deionized water.
Samples were passed through cheesecloth into labeled 15 mL tubes and were centrifuged
for 10 minutes at 800 g. Pellets were re-suspended in saturated Sheather's sugar solution
and transferred onto coverslips by centrifugation for 10 minutes at 800 g. Coverslips were
scanned at 100 x magnification to determine if sporocysts were present. Sporocysts were
frequently noted to have a pink hue if examined within 12 hours of removing the intestinal
tract. When sporocysts were identified, the remaining gastrointestinal tract mucosa was
recovered. Mucosal scrapings were mixed in to an equal volume of 5.25% NaOCl
(commercial bleach) on ice for 30 minutes and stirred every 10 minutes. Tissue aggregates
were removed by pouring the sporocyst suspension through a gauze mesh. Bleach was
removed by 2-3 washes in deionized water. The sporocysts were collected by
centrifugation at 800 g for 10 minutes between washes and finally stored at 4 C in
phosphate-buffered saline (PBS) with 10,000 U/mL, 100 [tg/mL streptomycin and 50
[tg/mL gentamicin to prevent bacterial overgrowth.
An aliquot of 5,000-10,000 sporocysts from each isolate was pelleted at 800 g and
re-suspended in equine bile (pH 7.0) with 2% w/v purified trypsin (Sigma T-8642; Sigma
Chemical Company, St Louis, Missouri) followed by incubation at 37 C for up to 4 hours.
Excystation was confirmed by examination of a slide at 100 x total magnification.
Estimated percentage excystation was made by examining 10 fields and identifying
excysted sporocyst cases and motile sporozoites. DNA from the sporocysts was extracted
using a non-ionic detergent (NP-40; Sigma Chemical Company, St. Louis, Missouri)
method and PCR was performed as previously described (Tanhauser et al., 1999) using 6
different primer-pairs, in separate reactions, for each sporocyst isolate. According to
published protocol, sporocysts were characterized as either neurona-like, falcatula-like,
dissimilar to both or, rarely, a mixture (Tanhauser et al., 1999). An aliquot of sporocysts
from any isolate to be used for horse challenge was re-examined to confirm concentration
and ability of the sporocysts to excyst prior to use. All doses for each horse were
prepared in advance and each vial was labeled and re-counted prior to beginning the
experiment. All counts were performed twice.
Single-Dose Horse Challenge
Five horses were identified which had no serum S. neurona antibodies (as detected
by western blot analyses). Horses were vaccinated against eastern and western equine
encephalitis and tetanus, dewormed (Eqvalan; Merck AgVet, Rahway, New Jersey) and
kept in an isolation pasture for a minimum of 21 days before beginning the experiment to
permit acclimation. To ensure minimal chance of accidental exposure of the horses to
naturally-shed sporocysts of S. neurona or S. falcatula, only feed which had been heat-
treated beyond 65 C, a process which appears to destroy sporocysts (McKenna and
Charleston, 1992), was fed. Heat-extrusion of pelleted feeds (Purina Horse Chow 100 and
Purina Pure Pride 300; Purina Feed Mills, St Louis, Missouri) occurs at temperatures in
excess of 140 C and is followed by air-drying and immediate bagging, eliminating any
chance of recontamination. No hay was fed. The isolation pasture has previously been
described (Cutler et al., 1999) and used for previous horse challenge experiments. As in
earlier experiments, no structures were permitted to overhang the enclosure as possible
access. An area of the pasture was sub-divided to allow separation of horses into smaller
groups for handling and feeding. Two horses were randomly selected for challenge and
were housed in individual isolation stalls. On day 2, 5 x 105 sporocysts of the
characterized S. neurona isolate pool were administered by nasogastric tube in 10 mL
PBS followed by 1 L deionized water. Challenged horses were kept in the isolation facility
for 8 days thereafter to ensure pasture contamination did not occur by passage of
sporocysts in equine feces. During the 4-month study period, whole blood for serum
preparation was drawn at least three times weekly and CSF samples were collected and
neurologic examinations performed every 14 days on each horse and recorded on
videotape. Atlanto-occipital cisternal puncture was performed under short-term injectable
anesthesia with xylaxine (Xylaject; Phoenix Pharmaceuticals, St. Joseph, Missouri; 1.1
mg/kg bodyweight) and ketamine (Ketaject; Fort Dodge Laboratories Inc.; Fort Dodge,
Iowa; 2.2 mg/kg bodyweight) to acquire the CSF samples. Evaluation of CSF was
performed within 1 hour of collection for WBC concentration and cytospin slides were
prepared for differential cell counts. Horses were examined daily to monitor for presence
of intercurrent disease and for evidence of neurologic clinical signs.
Contemporary environmental sentinels were necessary to reflect any breach of
biosecurity in the isolation pasture. Sentinels were handled identically to challenged-
horses except they never left the isolation pasture. One young horse (#801) was separated
from the others at feed time and was permitted a typical quantity (approx. 2%
bodyweight) of hay in addition to the pelleted diet for a 4-week period prior to beginning
the experiments. This protocol deviation was necessary for the dietary management of
this horse and, in consequence, allowed potential evaluation of the risk of feeding hay on
the serostatus of this individual. After 4 weeks, horse #801 was handled and fed
identically to the others, although it was still separated at feed times. No hay was fed to
any horse at any other time. The "hay-challenge" horse was euthanized and necropsied at
the end of the first horse challenge. The 2 environmental sentinels were maintained in the
pasture until the end of the second horse challenge. One of the challenge horses for the
multiple-dose experiment was introduced into the pasture before the end of the experiment
and served as an additional environmental sentinel during its quarantine period.
Multiple-Dose Horse Challenge
Two additional horses were acquired, acclimated and handled as previously
described. After the quarantine period, both horses were assigned as subjects and were
transferred to an isolation stall in a closed building. On days 2 7, horses received 5 x 105
sporocysts daily in 10 mL PBS followed by 1 L deionized water by nasogastric tube. The
inoculum consisted of a single opossum isolate (#2009). On days 9-11 horses were given
500 g psyllium mucilloid (approximately 1 g/kg) in 4 L water by nasogastric tube to assist
in flushing unattached sporocysts out of the gastrointestinal tract before return to the
isolation pasture. This experiment began 6 weeks after the completion of the single-dose
experiment, lasted 50 days, and used the same 2 environmental sentinel horses.
Western Blot Analysis
Serum was prepared by centrifugation of clotted blood at 800 g for 10 minutes.
Serum and CSF were aliquoted, and those aliquots not submitted for analysis immediately
were frozen at -80 C for archival purposes. All submitted samples were coded non-
sequentially and accession sheets did not include specific details on the animal sampled.
At least one portion of every sample was submitted to the commercial diagnostic
laboratory Equine Biodiagnostics Inc., Lexington, Kentucky and analyzed as described
(Granstrom, 1993). In the multiple-dose challenge, portions of some samples were also
submitted to the commercial diagnostic laboratory Neogen Inc., Lexington, Kentucky and
were analyzed according to their standard protocol. All results not designated "positive"
by the laboratory are considered "negative" for purposes of comparison. One laboratory
(Neogen Inc., Lexington, Kentucky) reported a densitometric comparison of the 17 kD
protein band in each CSF sample to the same band in a series of standards, expressed as a
unit-less number between 0 and 100.
Gross and Histopathological Examination
Challenged horses and the "hay-challenged" horse were euthanized using an
overdose of sodium pentobarbitone intravenously (Beuthanasia-D Special; Schering
Plough Animal Health, Kenilworth, New Jersey). Gross examination of organs included
lung, heart, liver, kidneys, spleen, pancreas, mesenteric lymph node, multiple sites within
the gastrointestinal tract and multiple sites (epaxial, appendicular and axial) of skeletal
muscle. Representative sections of each were placed in 10% neutral-buffered formalin for
fixation. The spinal cord was recovered cleanly in 3 parts- namely, cervical, thoracic and
lumbosacral sections. Each piece of spinal cord, with dura intact, was placed into a sterile
disposable jar of 1000 mL normal saline containing penicillin 100 U/mL and streptomycin
100 [tg/mL, transferred to a sterilely-draped laminar flow hood and was sectioned at 5-
mm intervals. Particular attention was paid to segments of spinal cord considered suspect
on neurologic exam of the living animal. Areas appearing discolored were bisected and
one half removed and placed aseptically into a petri dish containing culture media (RPMI-
1640; Gibco-BRL, Gaithersburg, Maryland) while the remaining spinal cord was fixed in
10% neutral-buffered formalin. Sections from the brain were taken as follows: the
occipital lobe, diencephalon at thalamus, mesencephalon at rostral colliculus,
metencephalon at pons at cerebellar peduncles, caudal medulla, and cerebellum. After
fixation, histological sections of all tissues were made at 6-[tm intervals and stained with
hemotoxylin and eosin and were examined under light microscopy by a specialty
pathologist. Sections were also sent to the laboratory of Dr. Brad Barr, University of
California, Davis for immunohistochemical staining to identify S. neurona or S. falcatula
Opossum Collection and Inoculum Preparation
One hundred seventy three road-kill opossums were collected in 1997 and 66 were
collected in 1998 in time to characterize before horse challenge. Forty-five isolates were
obtained from these animals, of which 14 were identified as being S. neurona alone on at
least 4 tests and were considered for use as inocula. Some isolates appeared to be
neurona-like with some molecular tools, but falcatula-like with others: these isolates were
not used. For the single-dose challenge, the inoculum consisted of 4 isolates in a 3:1:1:1
ratio (isolate #1112, #1013, #1067 and #1071). Additional aliquots of the inoculum were
reserved for further host challenge. A mix of sporocysts had the potential added
advantage of controlling for varied infectivity among isolates. In the multiple-dose
challenge a single high count isolate (#2009) was used to provide the large stocks required
without consuming too many isolates. Isolate #2009 also contained a small percentage of
sporocysts identified as S. falcatula. However, in light of previous negative data from
horses challenged with biologically-purified S. falcatula it was considered unlikely that
low level contamination by this parasite would confound the results of this trial (Cutler et
Single-Dose Horse Challenge
No sentinel horse ever produced S. neurona antibodies. Both challenged horses
produced serum antibodies (day 19 and day 26) and CSF antibodies (day 40) [Table 3-1].
Challenged horses showed subtle neurologic signs from day 26 until euthanasia but these
deficits never progressed to be unequivocal. The hay-challenge horse (#801) showed
variable subtle clinical neurologic signs that were somewhat ephemeral, but predominantly
hindlimb and never exceeded grade I/V (Appendix B). Neither of the sentinel horses
showed neurologic signs (subtle or otherwise). The hay-challenge horse produced
antibodies in serum (day 30) and CSF (weak positive on day 30, strong positive on day
60). All antibodies persisted to the time of euthanasia without apparent diminution. No
therapy was necessary during the experiment for intercurrent disease. All horses showed
clinical signs of conjunctivitis and tracheitis, believed to be caused by a Streptococcus sp.,
which resolved without therapy. No gross abnormality was noted at necropsy except that
a large amount of sand was present in the gastrointestinal tract of each horse.
Western Blot Results
Results are available only from laboratory 1 (EBI) for the single-dose challenge.
Table 3-3 summarizes the results giving number of consecutive negative and then positive
tests and the number of consecutive days in each category. Data are presented for both
challenged horses, both environmental sentinels, the hay-challenge horse and a horse that
was in its quarantine period prior to being challenged in the subsequent multiple-dose trial.
In total, there was only a single discordant result in one of the challenged horses (#817).
Before the multiple-dose horse challenge began, all 4 remaining horses seroconverted
while still in the pasture. Results from laboratory 2 are available from the time of
seroconversion onwards. The summarized data are presented in Table 3-4. The results
for horse #817 from laboratory 2 are consistent. The full results for horse #817 from
laboratory 1 are presented in Table 3-5 because they are discordant. Some samples were
resubmitted to the same diagnostic laboratory in 1999 for comparative purposes. These
results are also presented in Table 3-5.
With the exception of horse #817 results from laboratory 1, the consistency of
within-horse results from either laboratory are excellent, with only a single negative result
being reported in any horse after first seroconversion. The agreement between
laboratories for horse #816 is complete.
Multiple-Dose Horse Challenge
Both challenged horses became unequivocally neurologic by day 13 (Table 3-2).
No definitive signs of cranial nerve dysfunction were observed. The predominant clinical
sign in each horse was weakness and ataxia of the pelvic limbs. Horse #817 was grade 2
or worse in pelvic limbs from day 13 until euthanized, and at least grade 1 in both thoracic
limbs. Horse #816 showed similar limb deficits. Diminution of the cervicofacial and
panniculus responses was noted in horse #816 on 4 of 6 examinations. Muscle atrophy
was a minor component of the presentation in both horses. Weight loss was not a
prominent feature of the disease in these horses. Neither of the environmental sentinel
horses developed evidence of clinical neurologic disease at any time during the trial.
The sequential densitometric comparisons of 17 kD protein to reference standard
protein for both challenged horses are shown in Figure 3-1.
At necropsy no gross abnormality was noted except that a large amount of sand
was present in the gastrointestinal tract. Both horses had mild myocarditis, mild lymphoid
hyperplasia in lymph nodes, and moderate lymphoplasmacytic to eosinophilic enterocolitis.
Histopathological lesions within the CNS consisted of neuroaxonal degeneration with
spongiosis, gliosis and occasional spheroids. Lymphoid cuffing was occasionally present
around vessels, but was mild or minimal. In horse #816, the lesions were most severe in
the cervical spinal cord at C6 and C7, consisting of neuroaxonal degeneration. Acute
severe meningeal hemorrhage was present multifocally throughout the cervical region. In
horse #817, lesions were similar but milder and most pronounced in the cervical region.
No meningeal hemorrhage was present.
Concentration of nucleated cells in the CSF was normal at all time-points (<8 cells
/ [IL). The highest cell count recorded was 5, in horse #799 on day 12 post challenge.
Minor contamination of CSF with blood occurred on three occasions: the RBC counts
were 137, 700 and 770 / [tL. On each occasion, RBC concentration returned to normal
before the next CSF aspiration.
This paper describes the first successful induction of equine protozoal
myeloencephalitis using molecularly-characterized S. neurona. Challenged horses either
seroconverted (single-dose study) or developed disease (multiple-dose study), whereas
contemporaneous environmental sentinels did not. Previous publications on this subject
(Dame et al., 1995; Fenger et al., 1997a; Cutler et al., 1999) clearly underline the great
importance of correctly and fully identifying the Sarcocystis species being used.
Otherwise, induction of EPM underscores the fact that opossums shed the infectious
agent, but does nothing to confirm the agent's identity.
Data reported elsewhere (chapter four) is relevant to the single-dose horse
challenge. Five species of birds were challenged in separate groups with 3 of the 4 S.
neurona isolates that constituted the inoculum for the single-dose challenge. The fact that
no bird developed sarcocysts when administered characterized S. neurona sporocysts
further supports the biological difference between S. neurona and S. falcatula, as has been
previously suggested (Tanhauser et al., 1999). Bird species challenged included the
brown-headed cowbird (Molothrus ater) and the boat-tailed grackle (Cassidix mexicanus)
which are known intermediate hosts of S. falcatula. In a previous bird challenge using S.
falcatula sporocysts, sarcocysts were very numerous and easily recognized in all muscles
at 90 days post-feeding (Dame et al., 1995).
It is unfortunate that it is still not possible to replicate the entire life cycle of S.
neurona in the laboratory in order to produce a single population of sporocysts:
nonetheless, these results do clearly demonstrate that opossum-shed S. neurona alone (as
identified by 6 molecular tools) is harmless to birds, but can induce EPM in horses.
Eventually, the ability to generate sporocysts in the laboratory will remove a large burden
of collection and identification of isolates and will be a necessary component of fulfilling
First evidence of seroconversion at days 19 and 26 is in agreement with previously
published data (Fenger et al., 1997a) where seroconversion occurred as early as 19 days
after challenge. Fenger reported development of clinical neurologic signs at day 28 (n=2)
or day 42 (n=2) whereas in this study definitive (grade II/IV or worse) were noted earlier,
by day 13. It is extremely difficult to design a truly objective neurologic examination for
horses: therefore, we report only those signs that were repeatedly identified by a blinded
examiner, using standard forms (Appendix A) and a standardized examination
methodology (Appendix B). Clinical neurologic signs were somewhat progressive, and in
one horse (#816) showed some improvement before euthanasia (at day 50).
Clinical experience suggests that development of EPM is dependent on multiple
factors, and therefore it should not be surprising that neurologic signs are variable among
horses, or that signs might develop after variable periods or different ingested doses. The
predominance of pelvic limb disease in the challenged horses reported here reflects the
predominant presentation of naturally-occurring EPM cases. There is still no good
pathophysiological explanation for the development of devastating peracute disease in
animals known to be clinically normal only hours before (T. Cutler and R. MacKay,
unpublished observations). This must remain a concern for those investigating EPM
because it indicates either that the horse challenge protocol is incomplete (causing only
mild disease), or that the organism is adapting, changing the disease manifestation and that
peracute disease is becoming even rarer. Although adaptation of the parasite is unlikely to
occur over such a short period (at most 40 years), there is substantial evidence that most
horses are quite capable of withstanding S. neurona infection without developing EPM.
Necropsy findings are moderate or mild in all challenged horses. No protozoa were
identified in any sections examined. The meningeal hemorrhage in some horses may be
related to the collection of a maximal quantity of CSF immediately post-mortem and did
not occur in one horse where only 60 mL of CSF was recovered. Clinical impression was
that both horses #816 and #817 were improving prior to euthanasia, and this may explain
the end-stage neural disease that we identified (neuroaxonal degeneration and gliosis).
Even when sampling CSF twice monthly it appears that at most only a mild inflammatory
response occurs and does not induce a pleocytosis in CSF. In future work, it will be
important to sample frequently so as to identify more accurately the date of first
appearance of antibodies in the CSF and to increase the attempts to isolate parasite from
the CSF. When RBCs contaminated the CSF aspirate there was no evidence of residual
contamination on the next aspiration from that horse (14 days later).
In this study either 1 or 6 doses of 5 x 105 sporocysts were administered to each
horse. This dose was selected on the basis of extrapolations from our previous work and
a previous experiment where daily output of sporocysts was determined in opossums fed
S. falcatula-infected cowbirds (R. Porter & E. Greiner, personal communication). We
were also guided by the LD50 oral challenge doses reported for cattle (2 x 105 sporocysts
of S. cruzi, Fayer and Dubey, 1986) and sheep (105 sporocysts of S. tenella, Leek et al.,
1977) and seroconversion of pony foals in another study with a dose less than or equal to
1 x 106 sporocysts (Fenger et al., 1997a). No evidence exists to infer whether a single or
continuous exposure is more effective in inducing development of the natural disease
although the definitive host for many Sarcocystis sp. sheds sporocysts for a short period of
weeks to months rather than lifelong (Dubey et al., 1989). We suggest that disease
induction and seroconversion validate the dose selected and that the eventual model of
EPM will use a challenge similar to the multiple-dose protocol described herein. In large
animals acting as typical intermediate hosts an additional round of merogony probably
occurs before sarcocyst formation (Dubey et al., 1989). If a similar pathogenesis is
involved in causing EPM in horses, which are atypical intermediate hosts of S. neurona,
then ingested dose may only influence the rate of development and/or the severity of
disease. Although we have no data to predict whether merogony in the horse has a finite
or unlimited number of iterations, the latter appears more likely. The fact that some cases
of EPM have developed up to 2 years after export from EPM-endemic areas may lend
support to such an extended process (I. G. Mayhew, personal communication). After
challenge, horses were held in the isolation facility and received psyllium mucilloid orally
to prevent shedding of viable, unexcysted sporocysts in the horse pasture. No data are
available to validate either the duration or effect of the psyllium loading. However, lambs
orally-challenged with very high sporocyst doses (3.5 x 106) were still shedding viable
sporocysts in feces 7 days post-dosing (Munday, 1985). An even longer "washout" period
may therefore be indicated in horses, although the lambs were not fed psyllium to aid in
flushing unattached sporocysts out of the gut lumen.
There is considerable concern amongst equine clinicians about the interpretation of
western blot results of serum or CSF. The data presented in Tables 3-3 and 3-4 should
provide further assurance that general agreement within and between the laboratories
included in this study is excellent. When a horse is identified as seropositive the test result
is highly repeatable. The value of a negative western blot on serum is particularly
underscored. The data presented in Table 3-5 are difficult to interpret. The laboratory
results may simply be erroneous. However, when the results reported on the resubmitted
samples are considered, laboratory error seems an unsatisfactory explanation.
Alternatively, horse #817 may have responded to the natural point-challenge with a poor
humoral response which peaked and then dissipated quickly. When the horse was
experimentally challenged an anamnestic response ensued together with disease. The
repeated samples are compatible with this result. The 17 kD data from laboratory 2 also
corroborates this second possibility, and elevated from 5 at challenge to 75 on day 20 post
challenge. In contrast, the results for horse #816 were already elevated before challenge,
possibly reflecting disease due to natural exposure and not due to challenge. Considerably
more data will be necessary to determine the full value of the 17 kD protein densitometry
reading. Nonetheless, it remained at or below 5 in the environmental sentinel horses and,
therefore, in this limited instance was capable of differentiating the horses who were
naturally exposed and then experimentally challenged from the horses that were naturally
It has become popular to discount results of serum western blot tests simply
because seroprevalence is so high. The implication is that the test is providing unusable
results. An alternative, and perhaps more likely, explanation is that seroprevalence is high
but disease incidence is low because innate and humoral immunity usually prevent disease
(EPM) from occurring after infection with S. neurona.
The hay-challenged horse (#801) seroconverted and developed CSF antibodies at
30 and 60 days respectively. The longer time to first appearance of antibodies may be
related to a lower challenge dose ingested with the hay, or that the horse was adjusting
from a low level of nutrition and was partially immunocompromised as a result, thus
delaying the humoral response. In either case, although this is a single animal, the only
factor different between this individual and the others in the pasture was its diet. No other
horse was permitted access to the hay during the 4 weeks that it was fed to horse #801.
All horses had access to the holding pen area when hay was not being fed. The most likely
explanation is that exposure to S. neurona occurred consequent to ingesting contaminated
hay. An alternative explanation is that this horse alone had an additional point exposure to
contaminant S. neurona sporocysts. However, its poor nutritional state would suggest
that this horse was less likely, rather than more, to seroconvert after a group exposure.
The isolation pasture where horses were kept was successful in protecting the
environmental sentinels during the first experiment. All horses in the pasture experienced a
point exposure, presumably to S. neurona, and seroconverted after the conclusion of the
single-dose horse challenge and prior to the multiple-dose horse challenge. This
demonstrated that these horses were susceptible and were truly protected during the first
challenge. When horses were transported from the isolation paddock to the necropsy
facility, a truck and trailer were driven through the pasture and may have acted as a
mechanical vector. Extensive measures were taken to prevent wildlife access to the
isolation paddock because it was known that opossums were present around the facility.
No opossum was ever seen inside the enclosure, and so it is assumed that such was not the
source of contamination. Alternative sources of sporocyst contamination include the city
water supply and contaminated food given in error by technical personnel or visitors to the
compound. We have frequently observed turkey vultures (Cilia, Air auna) and other prey
birds preferentially scavenging opossum intestines during our roadkill collections.
Potentially, although perhaps less likely, prey birds may act as mechanical vectors as a
result. Because feed pellets are scattered on the paddock ground while the horses feed,
small birds are frequently present eating the lost food and are an additional potential
EPM has apparently increased in incidence recently (MacKay, 1997a), and
seroprevalence has been documented to be increasing (Granstrom, 1993; Saville et al.,
1997). It is important to distinguish seroconversion, which represents exposure to the
organism and probably infection, and the disease EPM which additionally involves
invasion of the central and/or peripheral nervous system by the parasite, asexual
replication at those sites, a resultant inflammatory response and the manifestation of
neurologic disease. An increased seroprevalence could be a result of increases in the
population or distribution of either the intermediate or definitive host, or both. Surveys
performed by the State of Kentucky recorded a decrease in the number of opossums killed
by trappers and furriers from 70,000 in 1982 to 2,000 in 1994 (Cramer, 1997). Removing
this pressure on the population could therefore permit an increase in feral opossum
numbers and result in further alterations in population dynamics. Roadkill surveys of
opossums are not available from that state, but the raccoon highway mortality index
increased by 386% between 1986 and 1991. It is possible that without trapper pressure
that opossum populations have dramatically risen (at the same time that seroprevalence
has), and that in turn has forced opossums to disperse (and change their distribution). The
intermediate host remains unknown and precludes speculation on its role in the emergence
It is clear from the combined results of this and previous studies (Fenger et al.,
1997a) that D. virginiana is the definitive host of both S. neurona and S. falcatula, and
probably of other Sarcocystis sp. (Tanhauser et al., 1999). The identification of opossum-
shed sporocysts categorized as non-neurona/non-falcatula indicates that a considerable
amount of work remains to be completed to understand the role of the opossum in
dispersing Sarcocystis sporocysts into the environment. Of ongoing concern is the failure
of any investigator to identify the true intermediate host(s) of S. neurona. This remains a
priority area if we are to be able to identify and endorse logical wildlife control and
preventative measures with confidence.
This is the first report of EPM induced in horses using characterized S. neurona
sporocysts. It seems unlikely that natural disease only occurs after ingestion of such
enormous numbers of sporocysts, and it is probable that additional uncontrolled factors
reduce the ID50 in horses that spontaneously develop EPM. The observed dose increment
necessary to cause disease in addition to antibody induction may reflect clinical experience
with prevalence of S. neurona antibodies compared to prevalence of EPM. The clinical
implications of each possibility (i.e. seroconversion alone or with disease) should be
considered when interpreting western blot results. Published serosurveys indicate that
approximately 50% of the horses in the USA have antibodies (Bentz et al., 1997; Blythe et
al., 1997; Saville et al., 1997), perhaps now an underestimate (T. Cutler and R. MacKay,
unpublished observations), whereas it has been suggested that only 1-2% of horses
actually have EPM (R. MacKay, personal communication). Thus the state of knowledge
is that opossums have an unusual position as host of both S. neurona and S. falcatula.
Secondly, the western blot test is capable of differentiating exposed and truly infected
horses from unexposed horses, and is highly repeatable. Sarcocystis neurona causes the
often devastating disease we know as EPM while S. falcatula causes disease in birds but
not horses. The horse model described herein remains to be validated and refined with a
larger group before it is used for more advanced work.
Table 3-1. Summary of single-dose horse challenge.
Horse Number of # 5 x 105 1st day 1st day 1st day Isolate
ID# days in Group sporocyst Serum CSF Neurol. ID#
quarantine doses WB+ve WB+ve signs
797 50 Challenge I 1 19 40 40 mix*
799 30 Challenge I 1 26 40 40 mix*
798 114 Sentinel I & II 0 WB-ve WB-ve none none
815 42 Sentinel I & II 0 WB-ve WB-ve none none
Challenge I = single-dose challenged horses; Sentinel I & II = environmental sentinel horses for
both studies; WB+ve = western blot results positive; WB-ve = western blot result negative
throughout *The inoculum mix for horses in challenge group I comprised 50% isolate #1112
and 16.67% each of isolates #1013, #1067 and #1071.
Table 3-2. Summary of multiple-dose horse challenge.
Horse Number of # 5 x 105 1st day Isolate
ID# days in Group sporocyst Neurol. ID#
__ quarantine doses signs
816 115 Challenge II 6 8 2009
817 21 Challenge II 6 13 2009
798 114 Sentinel I& II 0 none none
815 42 Sentinel I& II 0 none none
Challenge II = multiple-dose challenged horses; Sentinel I & II = environmental
sentinel horses for both studies; WB+ve = western blot results positive; WB-ve =
western blot results negative.
Table 3-3. Summary of western blot tests from laboratory 1 for horses in the single-dose
challenge, demonstrating consistency of test results.
Group Chall. Chall. Sentinel Sentinel Hay Quarantine
Horse ID# 797 799 798 815 801 816
Serum WB-ve consecutive tests 6 3 19 9 5 8
duration (d) 50 12 122 138 29 75
WB+ve consecutive tests 5 6 *1 n/a n/a 10 n/a
duration (d) 61 75 n/a n/a 94 n/a
CSF WB-ve consecutive tests 5 5 9 6 4 4
duration (d) 61 46 89 75 32 49
WB+ve consecutive tests 4 4 n/a n/a 5 n/a
duration (d) 49 49 n/a n/a 61 n/a
WB-ve= western blot result negative, WB+ve= western blot result positive, Chall.=
challenged horse; Hay= hay challenged horse; Quarantine= horse entering Sentinel
status; *1= 1 result discordant (negative).
Table 3-4. Summary of western blot tests from laboratory 1 and laboratory 2 for horses in
the multiple-dose challenge.
Horse and Sample # tests WB- Duration # tests Duration
Laboratory ve (days) WB+ve (days)
#816 Lab 1 Serum 8 75 5 70
CSF 4 49 5 56
#816 Lab 2 Serum n/a n/a 9 65
CSF n/a n/a 5 56
#817 Lab 1 Serum 3 10 *
CSF 0 0 *
#817 Lab 2 Serum n/a n/a 9 68
CSF n/a n/a 6 68
WB-ve= western blot result negative, WB+ve= western blot
result positive, *= results discordant, see Table 3-5.
Table 3-5. Results of western blot testing of horse #817 from laboratory 1.
Day # Sample Date Western Blot result Western Blot result
-39 Serum 18-May Negative
-31 Serum 26-May Negative <> Positive
-29 Serum 28-May Negative
-24 Serum 2-Jun Positive
-22 Serum 4-Jun Positive = Positive
-22 Serum 4-Jun Positive
-20 Serum 6-Jun Not tested Positive
-17 Serum 9-Jun Not tested Positive
-10 Serum 16-Jun Not tested Positive
-3 Serum 23-Jun Weak Positive = Weak Positive
0 Serum 26-Jun Not tested Weak Positive
4 Serum 30-Jun Not tested Weak Positive
6 Serum 2-Jul Not tested Weak Positive
13 Serum 9-Jul Negative
20 Serum 16-Jul Positive
46 Serum 11-Aug Positive
-22 CSF 4-Jun Positive = Positive
-22 CSF 4-Jun Positive
-22 CSF 4-Jun Weak Positive
-22 CSF 4-Jun Positive
-3 CSF 23-Jun Weak Positive* = Weak Positive
13 CSF 9-Jul Negative = Negative
20 CSF 16-Jul Positive*
46 CSF 11-Aug Weak Positive
Day # is from first dose; laboratory reports albumin concentration elevated in CSF
compared to expected range. Weak Positive is considered Negative for comparisons. ""
same result reported when tested in 1998 and 1999; "<>" results discordant
Figure 3-1. Densitometric comparison of 17 kD protein on western blot of CSF of
challenged horses compared to reference standard. Each horse was challenged on day 0.
-- Be Good #816
--- Pistol #817
0 20 40
CHALLENGE OF PUTATIVE INTERMEDIATE HOSTS OF S. NEURONA WITH
There is an urgent need to be able to generate S. neurona sporocysts under
controlled circumstances in the laboratory. Firstly, only by taking naive opossums and
feeding them characterized S. neurona sarcocysts would it be possible to assure that the
resulting sporocysts are a single population. Minor contamination with S. falcatula, for
instance, would complicate challenge of budgerigars because as few as 25 sporocysts /
gram bodyweight can cause death (Smith et al., 1989). Secondly, collecting wild
opossums and screening them for infection is costly, time consuming, unreliable, and an
unpredictable method of acquiring sufficient sporocysts for challenge. The intermediate
host of S. neurona needs to be identified before control of the sylvatic life cycle of the
parasite will be feasible. Therefore, the primary aim of this experiment was to find a
vertebrate host that would allow completion of the life cycle, and a secondary aim was to
find the true natural intermediate host or hosts for S. neurona.
Equine protozoal myeloencephalitis is most common in the eastern USA but
occurs across the country. It is likely that the intermediate host(s) have a similar
distribution. The North American opossum may be found across the continental USA and
therefore is unlikely to limit parasite spread (Gardner, 1982). Additionally, EPM occurs in
Central and South America (De Barros et al., 1986; Granstrom et al., 1992; MacKay et
al., 1992) and while Didelphis virginiana does not, the two other members of the genus
(D. marsupialis and D. albiventris) are found in those regions (Gardner, 1982).
Sarcocystisfalcatula can infect at least 3 orders of birds as intermediate host. It is quite
possible, therefore, that more than one intermediate host exists for S. neurona, because
the parasites are closely related genetically (Tanhauser et al., 1999). We considered as
intermediate host candidates birds which are common in the southeastern US and which,
by virtue of body size and population density, would likely be included in a free-ranging
Laboratory animals were selected in an attempt to fulfill the primary aim of
completing the life cycle. Psittacines are known to be very susceptible to infection by S.
falcatula (Smith et al., 1989). Even low dose infection can be sufficient to induce a fatal
pneumonitis. If the behavior of S. neurona in budgerigars (Melopsittacus undulatus) was
different (i.e. it did not cause fatal disease) it would provide a useful technique to
corroborate the results obtained from molecular tools and DNA sequencing. If S. neurona
could infect budgerigars and sarcocysts were produced then the primary aim would be
fulfilled. Immunologically intact NIH Swiss mice (Mus musculus) were also selected for
challenge as an alternative possibility to complete the S. neurona life cycle. Although S.
falcatula cannot infect mice (differentiating it from Toxoplasma gondii [Dubey, 1999]),
the ubiquitous use of the mouse in laboratory medicine made it a priority animal to screen
for S. neurona challenge.
Materials and Methods
Ten brown-headed cowbirds (Molothrus after 10 boat-tailed grackles (Cassidix
mexicanus), 10 European starlings (Sturnus vulgaris) and 10 redwing-blackbirds
(Agelaius phoeniceus) were obtained from the United States Department of Agriculture,
Gainesville, Florida (courtesy of Dr. Mike Avery). There were five males and five females
in each group. Ten female Bobwhite quail (Colinus virginianus) were donated by a local
breeder. Birds were randomly assigned to 5 groups of 2. Three S. neurona isolates and 2
type-2079 isolates were selected for administration to birds. Sporocyst isolates processed
as previously described (chapter three) were diluted in deionized water to a concentration
of 5000 sporocysts /mL. Each bird was administered 100 ptL containing 500 sporocysts
by dribbling the suspension in their beaks (Table 4-1). Any material lost by dripping was
noted. Quail were housed in a closed room at Animal Resources and were fed commercial
bird food. All other birds were maintained in covered-outdoor aviaries, separated by
species, and were fed crushed corn. All birds were euthanized by decapitation at 88 days
post-challenge except for the cowbirds, which were euthanized at 74 days post-challenge.
All muscle surfaces were examined for evidence of sarcocysts with and without a
dissecting microscope. Representative sections of muscles, brain, liver, lung, spleen, heart
myocardiumm) and kidney were taken and fixed in formalin for histopathological
Fifteen budgerigars were purchased from a local pet store. Fifteen sporocyst
isolates were diluted to a working concentration of 10,000 sporocysts /mL deionized
water (Table 4-2). Birds were housed individually in a single room. Cages were arranged
in 5 rows of 3 and had solid floors to minimize cross-contamination with feces. Birds
were challenged in the same room in which they were housed. Birds were first challenged
with 250 sporocysts each (for isolates, see Table 4-2) in 250 ptL deionized water by
dribbling the inoculum into the beak. Clinical signs were monitored for evidence of
respiratory distress, feather fluffing, tachypnea or grossly abnormal behavior. Birds were
re-challenged at day 27 because no evidence of disease had been noted. For the second
challenge, each bird was given 1000 sporocysts in 250 ptL deionized water by dribbling the
inoculum in to the beak. No loss of inoculum out of the beak was noted for any bird.
During the experiment, birds were re-housed twice because of circumstances beyond my
control. Birds that survived to day 90 were euthanized by decapitation. Necropsy was
performed as soon as possible after death in each case. All muscle surfaces were
examined for presence of sarcocysts with and without a dissecting microscope.
Representative sections of muscles, brain, liver, lung, spleen, heart myocardiumm) and
kidney were taken and fixed in formalin for histopathological examination.
Eighteen NIH Swiss mice (24 to 26 g) were purchased from a commercial
laboratory animal producer and assigned four per box in a single room. All mice were
female. Mice were identified by indelible markings on the tail, which were re-applied as
necessary to ensure readability. Mice were assigned randomly into 9 groups (Table 4-3).
All isolates were administered in 200 ptL deionized water. Five groups received 500
sporocysts of a single isolate per mouse, the 6h group received 100 sporocysts per mouse
of each of the same isolates (total 500 sporocysts per mouse), the 7h and 8th groups
received 500 sporocysts of a single isolate each and the 9h group received water only.
Mice were maintained according to Animal Resources protocol, fed a commercial
laboratory animal feed, and checked daily for evidence of overt clinical disease or behavior
change. At 88 days, mice were euthanized by exposure to CO2 for 15 minutes. Mice
were transferred to our laboratory, and were examined for presence of sarcocysts by gross
and dissecting microscope examination of all muscle surfaces and on cut section.
Representative sections were taken of appendicular and trunk muscles. Sections of
internal organs were also sampled for histopathological examination as follows: brain,
liver, lung, spleen, heart myocardiumm), kidney and tongue.
All birds belonging to families Icteridae, Sturnidae and Phasianinae survived the
experiment until euthanized at day 74 cowbirdss) or day 88 (all others). Some birds
belonging to Psittacidae (budgerigars) succumbed before conclusion of the experiment
(Table 4-2). All mice survived until they were euthanized at day 88.
Gross Necropsy and Histopathology: Non-Psittacidae
No bird had grossly visible sarcocysts on the surface or cut-section of any of the
muscles examined. No stages of Sarcocystis sp. or compatible lesions were identified in
any bird. Seven of 10 cowbirds had some degree of pulmonary mineralization with or
without granulomatous inflammation. Lesions were centered around major airways. Two
cowbirds had few, and one had many, intravascular microfilariae. One starling had an
extensive granulomatous hepatitis in multiple sections examined. Four other starlings had
small scattered inflammatory foci of macrophages and occasional heterophils in liver
sections. Individual hepatocellular necrosis was associated with some foci.
Two redwing blackbirds had pulmonary mineralization and granulomatous
inflammation around major airways. One bird (RWB#8) had microfilariae within the lung
Three birds had mild multifocal acute hepatic necrosis. There were no abnormal findings in
either the grackles or the quail.
Gross Necropsy and Histopathology: Psittacidae
In all cases cell necrosis was non-suppurative unless specifically indicated. Results
for all birds are summarized in Table 4-4.
Five budgerigars died spontaneously and one was euthanized because of tachypnea
and feather fluffing. Not all tissues were available for all birds. Five of the six birds were
examined by gross and histopathological methods. One bird (BUD-1, isolate #0000) was
too autolyzed to be included in the study. In a second bird (BUD-5, #1112) only the brain
was suitable for examination and it was normal. Three birds had been challenged with S.
falcatula (BUD-1, 3 & 13) and one each with S. neurona (BUD-10), type-1085 (BUD-5)
and an S. falcatula / type-1085 mixture (BUD-14). Three birds had merozoites present in
at least one organ (BUD-3, 10 & 13), and one other had what appeared to be merozoites
present (BUD-14). One of the S. falcatula-challenged birds had skeletal and cardiac
muscle sarcocysts (BUD-3), while two others (S. falcatula-challenged (BUD-13) and S.
neurona-challenged (BUD-10)) had skeletal muscle sarcocysts alone. Budgerigar 14
(#1092) had markedly distended liver sinusoids filled with infiltrates of degenerating
leukocytes. Moderate autolysis was present throughout which impaired definitive
identification of merozoites. Although the major blood vessels and airways of the lung
were infiltrated with inflammatory cells and associated necrotic cells, merozoites were not
seen. Budgerigar 3 (#1035) had moderate to severe sarcocystosis throughout the skeletal
muscle section. Mild inflammatory foci were associated with individual myofiber loss but
not with sarcocysts. Cardiac muscle contained large sarcocysts but also free merozoites
associated with infiltrated leukocytes. Hepatic sinusoids were distended with leukocytes
but merozoites were not definitively identified. The spleen was similarly affected.
The 9 remaining birds were euthanized on day 90. Two of these birds had
merozoites present in at least three organs (BUD-11 & 15). One bird had been challenged
with the original type-1085 isolate (#1085) while the other bird received an isolate of
unknown species (#1093). Budgerigar #15 (#1093) had a non-suppurative mild
encephalitis with merozoites present and with some necrotic cells. Merozoites were also
present intralesionally in sections of skeletal muscle and were associated with severe
diffuse myositis. The bird also had an acute pneumonitis with merozoites present
intralesionally, moderate diffuse necrotizing hepatitis and a subacute diffuse severe
splenitis with merozoites present.
When merozoites were present they were most common in liver, lung and spleen.
When sarcocysts were present they were not associated with inflammation. When
merozoites were present they were always associated with inflammatory cells,
predominantly lymphocytes and plasma cells and frequently with necrotic cells in the
vicinity. Other organs showed mild, unrelated changes that were not considered
Gross Necropsy and Histopathology: Mice
No grossly visible sarcocyst was found on the surface or cut-section of any of the
muscles examined. No stage of Sarcocystis sp. or compatible lesion was identified in any
mouse. Two mice had mild multifocal hepatitis without organisms present.
The data presented exclude brown-headed cowbirds, grackles, redwing blackbirds,
starlings, quail, and mice as natural or alternative intermediate hosts for S. neurona. In
addition, budgerigars appear unlikely to be able to complete the life cycle, although a
single bird administered an S. neurona isolate did die. None of the hosts challenged were
able to complete the life cycle of type-2079 isolates. All budgerigars challenged with S.
falcatula developed pneumonitis and died. Some isolates of type-1085 also caused
pneumonitis and death in budgerigars. The host infectivity spectrum of S. falcatula and
type-1085 therefore appear to overlap to include psittacines, but we can only exclude
some possible intermediate hosts of S. neurona and type-2079. Non-psittacines were not
challenged with type-1085 in this trial, and therefore the degree of overlap of host
infectivity between type-1085 and S. falcatula remains unknown.
All budgerigars that were challenged with S. falcatula sporocysts died or were
euthanized. One bird that was challenged with S. neurona (isolate #1071) also died: it
appears unlikely, although not impossible, that this bird developed pneumonitis due to S.
neurona infection. No other S. neurona-challenged bird became diseased or even had
subtle histopathological changes. It is possible that isolate #1071 included S. falcatula at a
concentration below the threshold detectable by our DNA-based classification tools.
Isolate #1071 has been retyped and again appeared to be a single population of S.
neurona. All birds were identified with leg bands and each bird received the same isolate
when re-challenged so inadvertent cross-contamination is considered very unlikely. It is
possible that the budgerigar accidentally ingested some S. falcatula, although we took
measures to prevent this. Clubb and Frenkel (1992) reported an acute fatal illness in Old
World psittacines held in an outdoor breeding collection in Florida which was reproduced
when cockroaches that had ingested opossum feces were fed to cockatoos and induced an
identical fatal illness. In those birds merozoites (presumably S. falcatula) were sufficiently
numerous to completely occlude pulmonary capillaries. Budgerigars were moved twice
during the experiment because of practicalities within Animal Resources, and this may be
an additional source of cross-contamination if birds and bird cages were not rematched
Sarcocystis type-1085 has not been reported previously, other than the defining
molecular work from my laboratory (Tanhauser et al., 1999). Therefore, we had no basis
on which to predict its behavior in birds. Biologically, it appears that type-1085 behaves
more like S. falcatula than like S. neurona in that it can infect, and kill, budgerigars.
Coincidentally, limited genomic-DNA sequence comparison shows that type-1085 is more
similar to S. falcatula than to S. neurona (Tanhauser et al., 1999). It has been recognized
for some time that S. falcatula is unusual as a Sarcocystis sp. in that it appears to be able
to use multiple different intermediate hosts (Tadros and Laarman, 1982). Box and
Duszynski (1978) have suggested that S. falcatula may be more than one species. When
originally described, Stiles (1893) reported S. falcatula and Balbianafalcatula as different
species parasitizing the cowbird, although they are now regarded as synonyms (Levine
1986). Box et al. (1984) have also demonstrated that S. falcatula, as described, may infect
passeriform, psittaciform and columbiform birds. In the latter experiment, infected
passeriforms were fed to opossums, the sporocysts generated were fed to columbiforms
through two cycles or to psittaciforms through one cycle causing infection in each. Thus,
even if Box's original suggestion is correct, S. falcatula as a single organism may infect
multiple species of bird. Our data are compatible with two similar sarcocystid parasites,
although they do not address whether type-1085 can infect brown-headed cowbirds or
grackles. In either case, further investigation of type-1085 will be required to characterize
its relationship to S. falcatula.
Our original hypothesis that budgerigars could act as a biological filter and
differentiate between S. falcatula and S. neurona is strongly supported. With the
exception of the budgerigar described above, all S. falcatula challenged budgerigars died
whereas no type-2079 or S. neurona-challenged birds died. Type-1085 appears to yield
intermediate results, and will require additional investigation. In addition, the exquisite
sensitivity of the budgerigar to S. falcatula-challenge may make it valuable in determining
whether any given feral opossum isolate has a small component of S. falcatula or possibly
type-1085 in it. With the administration of 2 x 104 or more sporocysts to a bird, a
threshold detection level of 5% or less may be attainable. The test would take 10-14 days
and in some instances may be almost as rapid as DNA-based techniques given the frequent
difficulty in excystation and also the common PCR inhibition thought to be due to other
components of opossum feces. The budgerigar-challenge system may even be more
sensitive than PCR for identifying small S. falcatula components in individual isolates.
Our results are in agreement with Marsh et al. (1997b) who reported fatal
pneumonitis in budgerigars after the administration of culture-derived S. falcatula
merozoites but no disease or pathologic lesion after the administration of S. neurona
merozoites. The data are complementary because Marsh et al. used merozoites for
challenge, whereas we had administered sporocysts. Bypassing the excystation stage did
not alter the ability of Sarcocystis sp. to cause disease in budgerigars.
The challenge protocol for the budgerigars was unusual in that there were two
time points when sporocysts were administered. The presence of sarcocysts in some of
the birds that died (range 20-29 days after second challenge, 47 56 days after first
challenge) is compatible with development from either the first or second challenge (Smith
et al., 1989). Sarcocysts may be present in cardiac muscle as early as day 7, but in skeletal
muscle rarely occur before day 28 (Smith et al., 1989). It is unclear whether the fatal
extent of the pneumonitis was due to the first or second administration because S.
falcatula-challenged birds succumbed soon after the second challenge (Smith et al.,
1987a). Smith et al. (1989) reported death after 39 days with a dose of 25 sporocysts per
gram bodyweight. At a dose of 60 sporocysts per gram, death occurred on approximately
day 28. The initial dose that we administered was between 10 15 sporocysts per gram
while the second dose was 40 60 sporocysts per gram. The survival of all birds to 27
days after low dose challenge agrees with Smith et al. (1989). It is quite possible that
some S. falcatula-challenged birds may have developed disease and died if we had waited
longer before re-challenging them. However, our primary interest was in S. neurona and
therefore we chose to re-challenge the birds with a half-order of magnitude increase in
dose. The fact that some birds died 10 13 days after the higher dose challenge appeared
to be temporally related to that second challenge, but is compatible with either.
It is worthy of note that some budgerigars had moderate to severe infiltrates in
association with intralesional merozoites on some sections whereas other sections of the
same organ appeared normal or had non-specific cellular infiltrates. In light of this finding,
it is possible that we did not identify all birds or organs that were infected. Because of the
small size of the birds, autolysis had a rapid onset and this was thought to partially obscure
definitive identification of merozoites on some sections. In Table 4-4, where
appropriately-sized and stained 1 2 [tm-wide bodies were seen, but could not be
definitively identified, the symbol "?" designates the possible presence of merozoites.
The budgerigar (BUD-15) that was challenged with an unknown Sarcocystis
species is of particular interest because myositis was severe with many intralesional
merozoites present. No sarcocysts were present, however. It is possible that sarcocysts
were rare and were present on adjacent sections, but it appears more likely that in this
instance the parasite was not able to reach sarcocyst formation. The severe inflammation
is not compatible with sarcocyst formation but rather with merogony. The observed
pathology may represent infection of an atypical host for this Sarcocystis species. No
other bird had intralesional merozoites in skeletal muscle. Budgerigar #3 had some free
merozoites but sarcocysts were numerous. Further classification of isolate #1093 and
challenge of additional birds may provide evidence to explain this unusual finding.
Clinical signs were useful in identifying budgerigars that were developing
pneumonitis. Most birds developed feather fluffing and some degree oftachypnea. It
would have been useful to have sequentially measured body weight but this is difficult to
perform accurately in live budgerigars. A number of budgerigars exhibited significant loss
of pectoral muscle at necropsy, although this was obscured by feathering during life. It is
likely that development of this group of clinical signs is more useful as an end-point than
death because it would permit intervention and euthanasia and also better quality tissue
The development of sarcocysts in budgerigars is not novel, but is apparently
uncommon and was first reported in 1982 (Box and Smith). Presumably, most birds do
not survive long enough to develop skeletal muscle sarcocysts (present in an immature
form as early as 28 days post infection) and either cardiac muscle was not examined or
sarcocysts were not present there. Smith et al. (1989) have stated that cardiac muscle
sarcocysts rarely mature, and are absent in budgerigars killed many months after challenge.
Regression of myocardial sarcocysts was not observed in our trial because no S. falcatula-
challenged birds were alive after 40 days.
The presence of merozoites in the liver, lung and spleen of most S. falcatula-
challenged budgerigars is in agreement with previously published data (Smith et al.,
1987a, Smith et al., 1987b). Merozoites were most abundant in pulmonary vasculature,
but also were numerous in the liver sinusoid. This is also in agreement with previous
The lack of sarcocyst development in the birds challenged with other Sarcocystis
sp. was disappointing but not surprising. Sarcocystis neurona may be a more typical
Sarcocystis sp. and have a narrow intermediate host range. The range of aberrant hosts
that have developed disease after infection with S. neurona or S. neurona-like organisms
stands in contrast to this, however (Dubey et al., 1987; Dubey et al., 1990; Dubey and
Hedstrom, 1993; Dubey et al., 1994; Klumpp et al., 1994; Dubey et al., 1996; Lapointe et
al., 1998;). If S. neurona were capable of infecting brown-headed cowbirds and grackles,
we might have expected to identify a mixed infection in laboratory opossums fed naturally-
infected cowbirds in previous experiments (Dame et al., 1995). The presence of only
incidental findings in all birds examined by histopathology provides strong evidence that
future investigations into possible hosts should be directed towards other bird species,
other mammals, or even amphibians.
We did not attempt to identify sarcocysts in any species of bird by digesting
muscle, although a pepsin and trypsin digestion method was described by Box and Smith
(1982). It may be more sensitive than gross examination alone but that method failed to
identify the rare sarcocysts in pigeons (Columba livia) which subsequently infected naive
opossums. The early conclusion that pigeons did not complete the life cycle of S. falcatula
was revised when it was demonstrated that pigeons do in fact form limited numbers of
sarcocysts when challenged (Box et al., 1984).
The location of the pulmonary lesions identified in the brown-headed cowbirds and
redwing blackbirds (centered on major airways) suggests that birds aspirated a portion of
the inoculum. The inoculum was administered by dripping into the buccal cavity to avoid
the possible trauma of passing an oro-esophogeal tube in wild birds. Birds were held
firmly during the challenge, but were permitted limited control of their heads. No
evidence of respiratory or other distress was noted during challenge or within the 12 hours
before returning the birds to the aviaries.
The microfilariae seen in redwing blackbird #8 possibly could have been
merozoites rather than fragments of microfilariae; however, this is considered unlikely
because so few were present, neither its group mate RWB#7 nor other challenged birds
had merozoites identified, and the birds were wild-caught adults with natural parasite
burdens. Many of the cowbirds also had intravascular microfilariae definitively identified
and had been housed in adjacent aviaries.
The hepatitis described in 3 blackbirds was mild and was considered unlikely to be
clinically significant. Additionally, no organisms were identified in any of the foci.
However, we did not attempt to recover viable organisms by tissue culture of
homogenized organs. This could have been a useful corroboration of the histopathology
results, or would have been a more sensitive test if organisms were isolated.
In retrospect, and in light of the results of the other experiments, it is not
surprising that no mice challenged here developed sarcocysts. Marsh et al. (1997a) had
demonstrated that S. neurona administration to nude mice resulted in development of
encephalitis. The suggestion that the immune status was a critical factor in determining the
susceptibility of those mice to S. neurona is supported by our results. In another
experiment (Dubey and Lindsay, 1998), nude and gamma interferon-knockout mice were
fed S. neurona sporocysts and became lethargic and developed encephalitis. Sarcocystis
neurona was grown in tissue culture from a liver / spleen / brain homogenate taken from a
nude mouse on day 11. Protozoa were identified in neural tissue as early as day 21. In
contrast, S. falcatula was harmless to both species of mouse when given as either
merozoite or sporocyst. Finally, Dubey fed sporocysts from two wild-caught (Dubey et
al., 1998) opossums to budgerigars, nude mice, and gamma-interferon knockout mice. No
mouse fed sporocysts of S. falcatula developed infection. Mice challenged with S.
neurona developed encephalitis, and merozoites were confirmed to be S. neurona by
immunohistochemistry. A third Sarcocystis sp. (possibly the species we designate type-
2079, [S. Tanhauser, personal communication]) described in that paper was also
administered to mice. Disease was similar to that induced by S. neurona but schizonts and
merozoites, which were predominantly found in the liver, did not react with S. neurona
antibodies. Additionally, sarcocysts were found in leg muscles of 2 mice killed on days 50
The presence of hepatitis in two mice in our study may be coincidence or may
represent subtle disease due to Sarcocystis sp. infection in partially-resistant mice. One of
two mice challenged with a type-2079 isolate (#1058) had a mild necrotizing, focal,
suppurative hepatitis infiltrated with degenerate neutrophils. No organisms were identified
with Giemsa stain. Its group mate was normal. The second mouse was 1 of 2 challenged
with S. neurona isolate #1112 and developed necrosuppurative hepatitis with multiple
small foci of acute hepatocellular necrosis. Giemsa staining of slides did not reveal
organisms. The second group mate was normal. Although disease cannot be completely
excluded, the NIH Swiss mice challenged here offer less potential for understanding the
life cycle of opossum-shed Sarcocystis sp. than do the nude or gamma-interferon
knockout mice used by others.
Box and Smith noted in 1982 that pigeons, which are Columbiforms, were
intermediate in their susceptibility to S. falcatula infection. Merogony was present but the
authors felt pigeons were resistant to muscle meronts. Although in that instance
sarcocysts were later identified (Box et al., 1984), the salutary lesson is that infection in
partially-susceptible prey (or intermediate hosts) is likely to be subtle and all means to
detect it should be employed. We are confident that we have not erroneously excluded
the true intermediate host of S. neurona, but it is possible that merogony did occur at low
levels in some of these birds and was not identified.
Future challenge of mice should be attempted in animals that have been selectively
bred to be specifically immunodeficient. The identification of some birds and mice with
mild nonspecific hepatitis is of concern. This may represent unrelated disease, but it is
also possible that we missed evidence of merozoites. Future efforts in these areas would
be well improved with the addition of a tissue culture protocol to attempt recovery of any
viable organisms present. Dubey has demonstrated the potential value of this
corroborative evidence (Dubey et al., 1998).
Members of the genus Sarcocystis (Lankester, 1882) undergo asexual proliferation
by processes of merogony and external budding and replication by endodyogeny.
Sarcocysts are typically present in striated muscle and merozoites are elongate. The
sarcocyst form of S. neurona has yet to be described, although it clearly must exist. The
only known host is the horse which is considered aberrant because sarcocysts of S.
neurona have not been identified in horses and because the organism has a neurotropism
in horses. Affected horses, therefore, present no risk to group mates, nevertheless,
outbreaks of disease have been reported (Fenger et al., 1997b). In contrast with other
Sarcocystis sp. the tendency for S. neurona (or S. neurona-like organisms) to infect
atypical hosts and cause neurologic disease has been reported on numerous occasions.
Encephalitis due to S. neurona (or S. neurona-like organisms) has been reported in a
colony of mink (Dubey and Hedstrom, 1993), a striped skunk (Dubey et al., 1996), a
rhesus monkey (Klumpp et al., 1994), a cat (Dubey et al., 1994), a steer (Dubey et al.,
1987), harbor seals (Lapointe et al., 1998) and caused fatal disease in a raccoon (Dubey et
al., 1990). This is a very broad range of hosts by any comparison. The fact that aquatic
mammals can be affected is interesting. Hepatitis was recently reported in polar bears
caused by a Sarcocystis sp., but the merozoites did not stain with S. neurona antibodies
(Garner et al., 1997). It is quite possible that some of the S. neurona-like organisms are
not S. neurona at all, but that epitopes have been conserved among some Sarcocystis sp.
The strong neurotropism of S. neurona and the host spectrum are very uncharacteristic of
Sarcocystis sp. in general, and mean some question remains about the final classification of
S. neurona. The encephalitis experimentally induced in laboratory mice by two different
Sarcocystis sp. (Dubey et al., 1998) is the logical extension of this literature and
capitalizes on the commonality of immunosuppression in most of the aberrant hosts
affected. The mouse model opens a new line of investigation which should prove very
In summary, the results presented suggest that unlike S. falcatula, neither S.
neurona nor type-2079 can use the brown-headed cowbird or grackles as intermediate
hosts. In addition no quail, starling or redwing blackbird developed convincing evidence
of disease after challenge with either S. neurona or type-2079. Type-1085 isolates can
behave similarly to S. falcatula and induce fatal pneumonitis in budgerigars. Merozoites
can be found in multiple organs, and the pathology is similar to that seen with S. falcatula
infection. However, not all type-1085 organisms caused disease, and no bird developed
sarcocysts. The experimental information provided is incomplete and further evaluation is
certainly warranted. Immunocompetent mice are not a suitable intermediate host
substitute for S. neurona, S. falcatula, type-1085 or type-2079 on the basis of the limited
replicate pairs described here. The information presented is in agreement with previously
published molecular evidence that the opossum can be infected with at least 4 different
types of Sarcocystis sp. Furthermore, these data suggest that the 4 different classifications
are different species on the basis of their individual ability to infect the different hosts
challenged and reported here. Immunodeficient laboratory animals may provide the next
major advance in understanding the life cycle of these opossum-borne Sarcocystis sp., and
show considerable promise as an alternative model to study S. neurona.
Table 4-1. Administration of sporocysts to brown-headed cowbirds, boat-tailed grackles,
Bobwhite quail, European starlings and redwing blackbirds.
Bird Bird Bird Bird Bird Number of Isolate Species Test
ID# ID# ID# ID# ID# sporocysts ID# Identity results
BHCB-1 BTG-1 BWQ-7 ES-1 RWB-1 500 1013 neurona 10 n
BHCB-2 BTG-2 BWQ-8 ES-2 RWB-2 500 1013
BHCB-3 BTG-3 BWQ-5 ES-3 RWB-3 500 1062 type-2079 sequenced
BHCB-4 BTG-4 BWQ-6 ES-4 RWB-4 500 1062
BHCB-5 BTG-5 BWQ-1 ES-5 RWB-5 500 1112 neurona 10 n
BHCB-6 BTG-6 BWQ-2 ES-6 RWB-6 500 1112
BHCB-7 BTG-7 BWQ-9 ES-9* RWB-7 500 1071 neurona 10 n
BHCB-8 BTG-8 n/a ES-10 RWB-8 500 1071
BHCB-9 BTG-9 BWQ-3 ES-7 RWB-9 500 1058 type-2079 sequenced
BHCB-10 BTG-10 BWQ-4 ES-8 RWB-10 500 1058
BHCB= brown-headed cowbird, BTG= boat-tailed grackle, BWQ= Bob White quail, ES=
European starling, RWB= redwing blackbird. Bird ES-9* died shortly after challenge due to a
severe mite infestation. Bird BWQ-10 died prior to challenge and is not shown in the table.
Brown-headed cowbirds were euthanized on day 74 while all other birds were euthanized on day
88. Test results are number of marker positions indicating the Sarcocystis sp. in the isolate is
falcatula-like (f) or neurona-like (n).
Table 4-2. Administration of sporocysts to budgerigars and summary of results.
Bird Number of ID# of Species Test Day # Manner of
ID# Group sporocysts isolate identity results died death
BUD-1 1 1000 0000 falcatula 10 f 13 died
BUD-2 2 1000 1051 neurona 10 n 90 experiment ended
BUD-3 3 1000 1035 falcatula 10 f 28 euthanized early
BUD-4 4 1000 1129 unknown -- 90 experiment ended
BUD-5 5 1000 1114 type-1085 3 n, 1 f 20 died
BUD-6 6 1000 1013 neurona 10 n 90 experiment ended
BUD-7 7 1000 1112 neurona 10 n 90 experiment ended
BUD-8 8 1000 1058 type-2079 sequenced 90 experiment ended
BUD-9 9 1000 1062 type-2079 sequenced 90 experiment ended
BUD-10 10 1000 1071 neurona 10 n 29 died
BUD-11 11 1000 1085 type-1085 8 n, 2 f 90 experiment ended
BUD-12 12 1000 1086 type-1085 8 n, 2 f 90 experiment ended
BUD-13 13 1000 1089 falcatula 10 f 20 died
BUD-14 14 1000 1092 falcatula & mixture 6 died
BUD-15 15 1000 1093 unknown --- 90 experiment ended
Day # killed is from date of second challenge (with 1000 sporocysts), which was 27 days after
initial challenge with 250 sporocysts per bird. Test results are numbers of marker positions
indicating the Sarcocystis sp. in the isolate is falcatula-like (f) or neurona-like (n).
Table 4-3. Administration of sporocysts to NIH Swiss mice.
Mouse Group Number of Isolate Species Test Notes on Day #
ID# sporocysts ID# identity results administration killed
Ibar 1 500 1051 neurona 10 n OK 88
dot 1 bar 1 400-500 1051 1 drop out 88
2 bar 2 500 1058 type-2079 sequenced OK 88
dot 2 bar 2 500 1058 __OK 88
3 bar 3 400-500 1013 neurona 10 n 175 uL (not 200 uL) 88
dot 3 bar 3 400-500 1013 175 uL (not 200 uL) 88
4 bar 4 500 1085 type-1085 8 n, 2 f OK 88
dot 4 bar 4 500 1085 __OK 88
5 bar 5 500 1071 neurona 10 n OK 88
dot 5 bar 5 500 1071 __OK 88
6 bar 6 500 Mix* [mixture] n/a OK 88
dot 6 bar 6 500 Mix* __OK 88
7 bar 7 500 1112 neurona 10 n OK 88
dot 7 bar 7 500 1112 __ OK_ 88
8 bar 8 500 0000 falcatula 10 f OK 88
dot 8 bar 8 500 0000 __OK 88
black tail Control -- water [water] n/a OK 88
black tail Control -- water __ __OK 88
mix* is a composite of the following isolates #1051, #1058, #1013, #1085 and #1071. n/a= not
applicable. Test results are number of marker positions indicating the Sarcocystis sp. in the
isolate is falcatula-like (f) or neurona-like (n).
8n rn 0A
LI K | I I I l l I
LI 1 z
as CA aCAlA C C
0 0 00 0)
0Ci i i Z
l as a 1
-L mI s i4 m s
-'' z ^3
0c cs '
1F IKH _^. ____ L---- ^^ ----- ^-
Figure 4-2. Photograph of a boat-tailed grackle (Cassidix mexicanus).
Figure 4-3. Photograph of a European starling (Sturnus vulgaris).
Figure 4-4. Photograph of a redwing blackbird (Agelaiusphoeniceus).
Figure 4-5. Photograph of Bobwhite quail (Colinus virginianus).
Figure 4-6. Photograph of budgerigars (Melopsittacus undulatus)
DIDELPHIS VIRGINIANA: DEFINITIVE HOST OF MULTIPLE SARCOCYSTIS SP.
The North American opossum (Didelphis virginiana) is the only identified
definitive host ofS. neurona (Fenger et al., 1995). The intermediate host(s) in this
heteroxenous life cycle is (are) as yet unidentified. Levine (Levine and Tadros, 1980;
Levine, 1986) reported 4 Sarcocystis sp. ofDidelphis sp. opossums: 2 use D. virginiana
as a definitive host (S. falcatula Stiles, 1893; Box et al., 1984) and S. rileyi (Stiles, 1893)
and 2 use D. marsupialis as an intermediate host (S. garnhami and S. didelphidis).
Sarcocystisfalcatula had been extensively studied and re-described by Box and colleagues
in the 1980s (Box and Duszynski, 1980; Box and Smith, 1982; Box et al., 1984).
Sarcocystis neurona was named when first recovered from diseased equine spinal cord
and grown in cell culture (bovine monocyte M617) by Dubey et al. (1991). All attempts to
recreate the life cycle of S. neurona must incorporate collection of infected opossums as a
source of sporocysts. Ultimately, the identification of an intermediate host would allow
generation of sporocysts in the laboratory by experimental infection of definitive and
intermediate hosts. Large scale collection of opossums would provide not only a source of
sporocysts, but also would provide descriptive data of the natural population that might
lead to a better understanding of the natural spread of the organism and perhaps identify
risk factors for the development of EPM in horses.
It was postulated that S. neurona was synonymous with S. falcatula (Dame et al.,
1995) and, in fact, the DNA sequence of the 18s rRNA gene of S. falcatula was identical
to that published for S. neurona (Fenger et al., 1994). The inability to recreate EPM in
horses with S. falcatula sporocysts from laboratory opossums fed infected brown-headed
cowbirds (Molothrus ater) forced reconsideration of that hypothesis (Cutler et al., 1999).
Molecular tools have now been reported that demonstrate S. neurona and S. falcatula are
different, and that additional Sarcocystis sp. are shed by the opossum (Tanhauser et al.,
1999). Equine protozoal myeloencephalitis has been reproduced both using an
uncharacterized sporocyst homogenate (Fenger et al., 1997a) and using characterized S.
neurona sporocysts (chapter three).
We undertook a large scale collection of road killed and trapped opossums as a
source of sporocysts for our host challenge experiments. It was suspected that the
descriptive data collected during this survey might be useful in determining risk factors for
infection of individual opossums. Although a number of state agencies collect data on
numbers of nuisance animals killed on the highways, we are not aware of any data
describing opossum-borne Sarcocystis sp. The results reported here describe the
molecular classifications of 87 sporocyst isolates from 419 opossums collected over a 2-
year period. Factors which might affect shedding pattern and intensity of infection, such as
age, gender, weight, time of year and location are of specific interest and are also
Materials and Methods
Opossums were collected either as roadkill or were live-trapped in the state of
Florida between November 1996 and December 1998. Posters and radio advertising was
used to maximize reporting of roadkilled opossums (Figure 5-1). Collections were made in
a discontinuous fashion because of the seasonality of opossum movements (and hence
numbers killed on roads) and also to permit examination of collected material. Overall
numbers collected were in part driven by the need to accumulate S. neurona sporocysts
for horse challenge experiments.
Gender and weight were recorded and a unique identification number was
assigned. Some samples were provided to my laboratory as intestinal tracts only, (without
other information). Necropsies were performed as soon as possible after collection. The
gastrointestinal tract was isolated and removed intact through an incision in the midline of
the ventral abdominal wall. The mesentery was stripped off and the intestinal contents
milked into a clean, labeled container. A section of distal ileum was incised, flattened and
the mucosal surface exposed. Excess ingesta were washed off the mucosa with deionized
water and then mucosa was collected by scraping with a microscope slide held at a 45-
degree angle. A 2-g sample of feces and a 1-mL sample of intestinal-mucosal scrapings
were taken from the original samples and each was homogenized in 10 mL deionized
water. Mucosal scrapings were passed through cheesecloth into labeled 15 mL tubes and
were centrifuged for 10 minutes at 800 g. Pellets were re-suspended in saturated
Sheather's sugar solution and transferred onto coverslips by centrifugation for 10 minutes
at 800 g. Coverslips were scanned at 100 x magnification to determine if sporocysts were
present. Sporocysts were frequently noted to have a pink hue if examined within 12 hours
of removing the intestinal tract. When sporocysts were identified, the remaining
gastrointestinal tract mucosa from that animal was recovered. Infected live opossums were
killed (Beuthanasia-D Special; Schering Plough Animal Health, Kenilworth, New Jersey)
and their gastrointestinal tracts recovered. Mucosal scrapings were mixed in to an equal
volume of 5.25% NaOCl (commercial bleach) on ice for 30 minutes and stirred every 10
minutes. Tissue aggregates were removed by pouring the sporocyst suspension through
gauze mesh. Bleach was removed by 2-3 washes in deionized water. The sporocysts were
pelleted at 800 g for 10 minutes between washes and finally stored at 4 C in phosphate-
buffered saline (PBS) with 100 U/mL penicillin, 100 [tg/mL streptomycin and 50 [tg/mL
gentamicin to prevent bacterial overgrowth. The final sporocyst count of each isolate was
established by counting at 400 x magnification using a hemocytometer. Periodically,
samples were recounted to determine if spontaneous excystation was occurring in the
Density Purification of Sporocysts
A density gradient was prepared by underlayering a 2 mL aliquot of isolate
(containing 10,000 sporocysts), with 3 mL each of 60%, 30% and finally 20% Percoll
gradient. Sample tubes were centrifuged in a swinging bucket rotor SW41 (Beckman
L870 centrifuge; Beckman Instruments, California) at 22,500 g for 30 minutes at 25 C.
Most sporocysts were present at the interface between 30% and 60% percent Percoll. This
material was collected and washed twice with Hanks' buffered saline solution (HBSS),
and finally stored in HBSS with 100 U/mL penicillin, 100 tlg/mL streptomycin and 50
tlg/mL gentamicin. Sporocysts so purified were used for DNA extraction when DNA
extracts from unpurified samples failed to amplify (due to PCR inhibition).
Excystation of Sporocysts
Aliquots of 10,000 sporocysts each were pelleted by centrifugation and
resuspended in 200 ptL undiluted equine bile (pH 7.0) containing 2% w/v purified trypsin
(Sigma T-8642; Sigma Chemical Company, St. Louis, Missouri). Sporocysts were then
incubated at 37 C for up to 12 hours to allow maximal excystation. Excystation was
confirmed by examination at 100 x magnification.
Polymerase Chain Reaction
Excysted sporozoites were pelleted by centrifugation for 10 minutes at 1,600 g in a
microcentrifuge. DNA was extracted and then amplified using up to 6 separate pairs of
RAPD primers (Tanhauser et al., 1999) as previously described. A negative control was
included which contained all components of the PCR mixture except template. The
possibility of contamination by extraneous amplicons was minimized by performing
reaction setup, sample extraction, and amplification / analysis in three separate
laboratories. Reaction tubes were closed in the room in which the reaction mixtures were
prepared and only opened after amplification. PCR products were analyzed by agarose gel
electrophoresis using a 1% agarose gel (NuSieve 3:1; FMC, Rockport, Maine).
Up to six molecular tools were used to characterize most isolates. Briefly, a tool
consists of two specific primers used to amplify a DNA sequence and a corresponding
restriction endonuclease digestion enzyme pair capable of positively identifying a neurona-
like or falcatula-like sequence. Identifications were by positive criteria, i.e. neurona-like
isolates were cut like S. neurona only, and falcatula-like isolates were cut like S. falcatula
only. One tool amplifies a different length product from neurona-like isolates than from
falcatula-like isolates. Final categorization depends on the collective results from all tools:
certain combinations of results characterize the non-falcatula/non-neurona isolates termed
type-1085 (named after the initial isolate numbers showing those characteristics). Isolates
designated type-2079 were classified after sequencing of portions of the ITS sequence
because those isolates do not amplify with most of the molecular tools utilized. The DNA
sequences are compared with the sequences of both S. neurona and S. falcatula to ensure
that they are different.
Restriction Endonuclease Digestion
Amplified DNA from isolates was first run on an 1.5% agarose gel to demonstrate
that DNA was present in sufficient quantity to be clearly visible after digestion. Two
aliquots of 10 ptL of PCR-amplified product were incubated with 5 uL of one or other
restriction endonucleases in a waterbath at 37 C for 1-2 hours according to manufacturer's
instructions (Pharmacia, Kalamazoo, Michigan). At the conclusion of the digest, another
agarose gel is prepared with the following lanes: ladder, undigested product, product
digested with first enzyme, product digested with second enzyme. In each case, the digest
enzyme pairs are selected to positively identify a neurona-like or falcatula-like sequence.
Four hundred nineteen opossums were collected between November 1996 and
November 1998. Eighty seven were infected (21% infection prevalence). Descriptive data
are available for 81 of those isolates, 78 of which were classified as single isolates.
Summary data are shown by month (Table 5-1) and by county (Table 5-2). Two skunks
(M. mephitis) were also collected and were negative.
Table 5-3 shows the average, median, minimum and maximum yield for each
species or type and for the total collection. Table 5-4 shows the shedding of sporocysts
by weight and in Table 5-5 the data are further classified by gender. Table 5-6 shows the
breakdown of shedding in opossums by gender.
Categorization of Isolates
All positive and negative controls appeared as expected, confirming that no
accidental contamination of samples occurred during preparation or handling, and that
reaction conditions were suitable for amplification of DNA. The breakdown of the 81
isolates by species and type examined is given in Table 5-3. Excystation of sporocysts of
most isolates was poor (average 15%; range 10-40%).
Equine protozoal myeloencephalitis is a common and often devastating disease in
horses in the United States. The apparent increase in disease incidence and confusion
about the life cycle of the causative organism required further investigation of the
definitive host and dissemination of infective sporocysts. In a review of all Sarcocystis sp.
known at the time, Levine (1986) listed just one (S. falcatula) that utilized the opossum as
a natural definitive host. In addition, experimental evidence was available that the
opossum could become infected with, and shed, S. rileyi (Levine, 1986). It had been
speculated that S. falcatula might not be a single species, but rather be several very similar
parasites that infected different birds (Box and Duszynski, 1978). Early horse-challenge
experiments suggested that sporocysts from more than one Sarcocystis species was shed
by opossums (Fenger et al., 1997a; Cutler et al., 1999). Further investigation using
molecular techniques confirmed this. Consequently, our laboratory has recently reported a
series of molecular tools which differentiate among S. neurona, S. falcatula and other
Sarcocystis sp. parasites of the opossum (Tanhauser et al., 1999). Those tools were used
to characterize sporocysts from naturally infected opossums and permit the description of
sporocysts in this paper. These data again call in to question whether S. falcatula, as
defined, is a single organism or not. The isolate that we have designated type-1085 can at
least infect and kill budgerigars (chapter four), and may well behave similarly to S.
It is now clear that opossums can shed sporocysts of at least 4 distinguishable
DNA types. These include at least 2 species (S. neurona and S. falcatula). It remains to
be demonstrated that the other classifications we have used actually represent independent
species. It is possible that 1 or more of them do not even belong to the genus Sarcocystis.
The identification of possible additional Sarcocystis sp. is not surprising on the basis of the
ability of opossum-derived sporocysts to infect different hosts as described elsewhere in
this thesis. Indeed, some of these species may already be known in their intermediate
hosts. A number of traditional parasitology questions involving life cycle are therefore
Some of the isolates collected have been administered to horses: S. neurona
sporocysts induced the disease EPM (chapter three) whereas S. falcatula did not (Cutler
et al., 1999).
The most recent data available in the United States indicate that approximately half
of US horses have antibodies against S. neurona (Bentz et al., 1997; Blythe et al., 1997;
Saville et al., 1997). Our experiences attempting to identify a large group of seronegative
horses is that seroprevalence may already be higher than that (T. Cutler and R. MacKay,
unpublished observations). On the basis of these data, it is somewhat surprising that the
prevalence of infection among opossums reported in this paper is so low (21%).
Furthermore, to date, only 3.8% (16) of the total (419) have been identified as shedding S.
neurona. Because of difficulty extracting DNA from low concentration isolates, and
possibly because of inhibitors contained in the samples, the slight majority of isolates (40)
have yet to be definitively typed. Therefore, the proportion of S. neurona isolates may
currently be underestimated. Of the other isolates classified, 13 were S. falcatula, 6 were
type-1085 and 3 were type-2079. Six isolates were classified as mixed infections of 2
species. Three of these isolates had only a small percentage of the second species present
and for data purposes in this paper are considered single isolates of the predominant
Prevalence of infection was highest in opossums recovered in the 2nd quarter and
was similar in both years (mean, 35%, Table 5-1). This quarter also included the highest
collection rates. It is possible that an interaction exists between diet and opossum
movement, or consequences of searching for a new territory, such as stress. Opossums
recovered in the 1st quarter were almost as likely to be infected (mean, 21%) as in the 2nd
quarter, while in the 3rd and 4th quarters far fewer opossums were infected (10% and 12%
respectively). There was a strong statistical difference between prevalence in the 2nd
quarter and other quarters in 1997 (p=l x 10-6), between 1st and 2nd quarters compared to
3rd and 4th quarters in both years (p=3 x 10-6, p=5 x 10-11), and a strong tendency between
2nd quarter of 1998 and other quarters combined (p=0.012). The increased prevalence in
opossums in the 2nd quarter may indicate a seasonal change in either the numbers of
intermediate hosts or the proportion of intermediate hosts that are infected. The data may
support a seasonal host which winters in Florida. If shedding of sporocysts lasts 4 -6
months after infection, few opossums are likely to be still shedding in the last quarters of
the year. In the 3rd and 4th quarters of 1998, we additionally began collecting opossums
from Alachua County Animal Control (ACAC) and it is possible that this changed the
prevalence in our collection. However, the prevalence in the 3rd quarter of 1997 was
similar to 1998 (13% vs. 9%). The prevalence of Sarcocystis in opossums from ACAC for
3rd and 4th quarters is low (17 of 155, 11%) but not statistically different from 3rd quarter
prevalence in 1997 (p=0.77). Nonetheless, a number of factors might affect prevalence in
the ACAC opossums. All animals collected by ACAC are live-trapped and may be
healthier than, or just different from, the roadkill population. Secondly, ACAC collects
opossums from across Alachua County whereas we noted a higher prevalence of infection
in opossums collected in cities. Thirdly, our roadkill population may be biased by the
motivation of people notifying us of roadkill or live-trapped opossums. This latter reason
appears unlikely because many calls came from radio-listeners in addition to farm
managers and horse owners.
Two changes in collection protocol occurred during this study and are relevant to
our report. Firstly in June 1997 we significantly increased advertising and more
vigorously pursued collecting roadkill opossums. Secondly, in July 1998 we began
collecting opossums from Alachua County Animal Control as described above. It is not
possible to determine the significance of the effects of these alterations, but we feel that
the data from June 1997 through July 1998 are likely most representative because we
collected a large percentage of all opossums road killed in Alachua County during that
time. The fact that the other seasonal data are comparable suggests that it is appropriate
to make some early conclusions.
Most opossums were collected from Alachua and Marion counties, because they
were closest to our College. Data by county are presented in Table 5-2, although because
of the disproportionate numbers in each group, it is hard to make many inferences. We
did note that many of the infected opossums, including most of the S. neurona isolates,
were collected in Alachua County and specifically from the city of Gainesville (population
approximately 100,000) rather than from the rural areas surrounding it. Additionally,
prevalence of infection in opossums collected in Gainesville was higher than other areas.
Opossums have adapted well to being city dwellers (Gardner, 1982), and it may be that
intermediate hosts are more commonly found, or more commonly infected, in cities.
There were some associations between the type of Sarcocystis sp. recovered from infected
opossums and area of the state the opossum was collected. Five of 6 isolates identified as
type-1085 were from South Florida (Dade and Broward counties) and the 6h was from
Marion County. Four of those isolates had greater than 2 x 106 sporocysts recovered
compared with only 22 of the remaining 75 isolates of all other species. It is noted that
the collections systems were different for these counties, and therefore these data are
likely to be biased. In 1998, the prevalence of infection per county and total dropped
because of the greater numbers of opossums collected in the 4h quarter, and the low
prevalence of infection during that period (Table 5-1). The overall prevalence drop from
1997 to 1998 was not statistically significant (p=0.11), but had a tendency to significance
when Broward County data was excluded (p=0.027). As expected there is no statistical
difference in prevalence of Broward County opossums between the years (p=0.11).
From the isolates already classified there is a trend between type and intensity of
shedding. Type-1085 (mean 1.2 x 107) and type-2079 (mean 2.4 x 107) were shed in the
highest numbers (Table 5-3). The lowest recovery was 3.4 x 105 for a type-1085 isolate
and 5.8 x 105 for a type-2079 isolate. Sarcocystisfalcatula was recovered in the lowest
numbers (mean, 7 x 105), while S. neurona was intermediate (mean, 2.4 x 106). Some
isolates of each type were recovered in very low numbers. Two possibilities may explain
the data. Either the intensity of shedding was always low in that individual opossum, or it
was collected near the beginning or end of infection and, therefore, of shedding. In
support of the second possibility, many Sarcocystis sp. are known to infect definitive hosts
for only a short period of time (Dubey et al., 1989). For some isolates, an additional cause
was the severely decomposed state of the gastrointestinal tract, preventing proper
recovery technique. Therefore, the differences in mean sporocyst recovery between
Sarcocystis types is likely underestimated.
The lowest infection intensities of all the species tend to overlap, but this is
probably an artificial confounding due to either poor recovery technique or a very early or
very late infection.
In a previous experiment conducted by investigators in my group, naturally-
infected brown-headed cowbirds were fed to opossums and sporocyst shedding was
monitored by examining fecal samples (R. Porter and E. Greiner, personal
communication). The numbers of S. falcatula sporocysts shed by those laboratory
opossums appears to be somewhat different than the numbers shed by the opossums
naturally infected with S. falcatula. The peak shedding in those laboratory opossums was
2 x 105 sporocysts per gram (R. Porter and E. Greiner, personal communication). With a
'standardized' fecal pellet weighing 30 g, up to 6 x 106 sporocysts could be deposited at
once. This appears to be considerably higher than the average numbers recovered in the
opossums reported here. Even though the comparison is not direct, the implication of such
intense fecal shedding is that large numbers of sporocysts were present in the gut mucosa.
In multiple experimentally-induced S. falcatula infections in opossums, Box and Smith
(1982) reported recoveries of between 1 and 6.3 x 108 sporocysts using their gut-digestion
method. Data from these experimental S. falcatula infections appear to correlate well. Of
note, in the latter study more sporocysts were recovered when the opossums were
euthanized soon after infection. Even if Box's digestion method increases yield 10-fold,
the discrepancy in sporocyst recovery between naturally-infected and experimentally-
infected opossums is very large. Brown-headed cowbirds experimentally challenged with
5,000 S. falcatula sporocysts had a very severe sarcocyst burden (Dame et al., 1995; E.
Greiner, personal communication). If natural S. falcatula, and possibly S. neurona,