Michigan State
University

This is to certify that the
dissertation entitled

Molecular Epidemiology and Evolution of
Sarcocystis neurona, The Etiologic Agent of Equine
Protozoal Myeloencephalitis (EPM)

presented by

Hany M. Elsheikha

has been accepted towards fulfillment
of the requirements for the

PhD. degree in Large Animal Clinical Sciences

 

 

 

MSU is an Affirmative Action/Equal Opportunity Institution

 

PLACE IN RETURN Box to remove this checkout from your record.
TO AVOID FINES return on or before date due.
MAY BE RECALLED with earlier due date if requested.

 

DATE DUE

DATE DUE

DATE DUE

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

2/05 c:/ClRC/DateDue.indd-p.15

 

 

—_—

MOLECULAR EPIDEMIOLOGY AND EVOLUTION OF
SARCOCYSTIS NEURONA, THE ETIOLOGIC AGENT OF EQUINE PROTOZOAL
MYELOENCEPHALITIS (EPM)

By

Hany M. Elsheikha

A DISSERTATION

Submitted to
Michigan State University
in partial fulfillment of the requirements
for the degree of

DOCTOR OF PHILOSOPHY
Department of Large Animal Clinical Sciences

2004

ABSTRACT
MOLECULAR EPIDEMIOLOGY AND EVOLUTION OF SARCOCYSTIS NEURONA,
THE ETIOLOGIC AGENT OF EQUINE PROTOZOAL MYELOENCEPHALITIS
(EPM)
By

Hany M. Elsheikha

Equine Protozoal Myeloencephalitis (EPM) is a common, serious neurological
disease of horses that can lead to disability and death. EPM is caused by the ubiquitous
protozoan parasite Sarcocystis neurona, Horses acquire this parasite from the feces of
opossums of the genus Didelphis when they ingest the sporocyst stage in contaminated
food. Inadequate understanding of the parasite’s species identity, genetic makeup, host
range, and the mechanisms it uses to cause disease in horses has hampered effective
diagnosis and management of the disease. The most basic information needed is how to
determine which Sarcocystis strains are S. neurona and which are not and how genetic
traits are transferred within the population. The main hypothesis investigated was that S.
neurona undergoes genetic recombination during the sexual phase of parasite
development leading to strains with variable genetic traits that may include increased
virulence. The studies described here included the assembly of a large collection of S.
neurona and other related cyst-forming coccidia, development of methods for
purification of the sporocyst stage from intestinal cells of opossums and the merozoite
stage from tissue culture and the use of DNA sequencing of the small subunit ribosomal
RNA (SSUrRNA) gene-and major surface antigen gene 1 (SAGI) as well as sequence of
the 25/396 diagnostic marker. Amplified Fragment Length Polymorphism (AF LP)

methodology was also used as a whole-genome fingerprinting method to develop a

“species identity” or “genetic fingerprint” for S. neurona strains. This method was
applied to characterize the available Sarcocystis strains that have been assessed by
preliminary phenotypic and molecular marker methods such as PCR-RFLP as “neurona-
like”.

The contributions of the present work are: (1) evaluation of reproducible and
sensitive methods for classification of strains of S. neurona from horses and opossums to
the species and subspecies level based on SAGl gene sequence data and AF LP markers;
(2) elucidation of the nature and extent of genetic variation and phenotypic heterogeneity
of Sarcocystis spp. isolated from horses, opossums, and other vertebrates; and (3) use of
phylogenetic information to assess the distribution of S. neurona in equids and wild hosts
and to predict the risk of introduction into horses. Methods developed in this work can
be used to identify to the species and subspecies level any S. neurona isolate, to track
sources of S. neurona for horses, and to test whether S. neurona undergoes genetic

recombination to produce new strains with pathogenic potential.

DEDICATION

With love, I dedicate my dissertation to
My beloved wife, Mom, and the spirit of my Dad

iv

ACKNOWLEDGEMENTS

Many people have made the completion of this work possible. First of all,

I would like to thank my major advisor Dr. Linda S. Mansfield for the very productive
ideas we shared. Dr. Mansfield offered me every possible way of support to accomplish
this work. Dr. Mansfield was always a great guidance and constructive challenge for me.
I also want to thank my co-major advisor Dr. Mahdi A. Saeed for always being an
inspiring figure in my career and for his help to come to Michigan State University.

I would also like to thank my committee members for their continuous and
generous support and cooperation throughout this work. Thank you, Dr. Scott D.
Fitzgerald. I thank you not only because you allowed me to conduct part of my
dissertation research at the Animal Health Diagnostic Laboratory, but especially for the
times in which you reached out to help as an advisor and as a friend. I also would like to
thank Dr. Thomas S. Whittam for providing me with the privilege of carrying out part of
my dissertation research in the Microbial Evolution Laboratory. His love of science and
his dedication to his work inspire me. My sincere thanks to Dr. Harold C. Schott and Dr.
Jeffrey P. Massey for their encouragement and enthusiasm and for providing me with
guidance in my research and in my career.

Many thanks to the past and present staff at the Parasitology Section, Diagnostic
Center for Population and Animal Health; the Whittam and the Mansfield laboratory
members; and most of all to Alice Murphy, David Lacher, Seth Walk, and Lindsey
Ouellette, Adam Nelson, Geetha Parthasarathy, David Wilson, Dr. Julia Bell, Kathryn
Jones, Lakeisha Cummingham, and Dr. Mary Rossano. I had great moments with all of

them and very friendly relationships.

I am also indebted to the Department of Large Animal Clinical Sciences staff for
helping me in various ways during my stay in the department and for providing me with
all the necessary advice and support to accomplish my academic goals.

Financial support was offered by the Egyptian government; Department of Large
Animal Clinical Sciences, College of Veterinary Medicine, Michigan State University;
and the Grayson Jockey Club Research Foundation.

I want to thank my wife Dr. Nashwa for her love and support in every possible
way she could and for being next to me during this exciting time of my life. And of
course my beloved mother, although a few thousand miles away, she has fully supported
me and encouraged me in every step of my life. Finally, I wish to honor the memory of

my father who I missed since my childhood but his spirit has always been next to me.

vi

TABLE OF CONTENTS

LIST OF TABLES ........................................................................................................... ix
LIST OF FIGURES ......................................................................................................... xi
KEY TO ABBREVIATIONS ....................................................................................... xvi
THE RESEARCH PROBLEM (SCOPE AND SIGNIFICANCE) ............................... 1
References ................................................................................................................ 4
HYPOTHESE STESTED .................................................................................................. 5
CHAPTER 1: LITERATURE REVIEW ........................................................................ 8
Life cycle of Sarcocystis neurona ............................................................................ 8
Sarcocystis neurona and Equine Protozoal Myeloencephalitis ............................... 9
Taxonomic status of S. neurona ............................................................................. 11
Previous genetic and phylogenetic studies of Sarcocystis neurona ....................... 12
Genetic variation among Sarcocystis neurona isolates ......................................... 16
References .............................................................................................................. 19
CHAPTER 2: PURIFICATION OF SARCOCYSTIS NE URONA MATERIALS ..... 24
[1] Purification of S. neurona sporocysts from opossum (Didelphis virginiana) using

potassium bromide discontinuous density gradient centrifugation ............................... 25
Abstract .................................................................................................................. 26
Introduction ............................................................................................................ 27
Materials and Methods ........................................................................................... 29
Results .................................................................................................................... 36
Discussion .............................................................................................................. 39
References .............................................................................................................. 42

[2] Generally applicable methods to purify intracellular coccidia from cell cultures and

to quantify purification efficacy using quantitative PCR .............................................. 44
Abstract .................................................................................................................. 45
Introduction ............................................................................................................ 46
Materials and Methods ........................................................................................... 48
Results .................................................................................................................... 57
Discussion .............................................................................................................. 71
References .............................................................................................................. 73

CHAPTER 3: PHYLOGENETIC RELATIONSHIPS OF SARCOCYSTIS
NE URONA ISOLATES OF HORSE AND OPOSSUM TO OTHER CYST-

FORMING COCCIDIA DEDUCED FROM SSUrRNA GENE SEQUENCES ........ 74
Abstract .................................................................................................................. 75
Introduction ............................................................................................................ 76
Materials and Methods ........................................................................................... 79

vii

Results .................................................................................................................... 89
Discussion ............................................................................................................ 101
References ............................................................................................................ l 08

CHAPTER 4: SARCOCYSTIS NE URONA MAJOR SURFACE ANTIGEN GENE 1
(SA GI) SHOWS EVIDENCE OF HAVING EVOLVED UNDER POSITIVE

SELECTION PRESSURE ............................................................................. - ............... 113
Abstract ................................................................................................................ 114
Introduction .......................................................................................................... l 15
Materials and Methods ......................................................................................... 117
Results .................................................................................................................. 121
Discussion ............................................................................................................ 130
References ............................................................................................................ 1 3 6

CHAPTER 5: PHYLOGENETIC CONGRUENCE OF SARCOCYSTIS NE URONA
ISOLATES IN THE UNITED STATES BASED ON SEQUENCE ANALYSIS AND

RESTRICTION FRAGMENT LENGTH POLYMORPHISM ................................ 138
Abstract ................................................................................................................ 139
Introduction .......................................................................................................... 140
Materials and Methods ......................................................................................... 142
Results .................................................................................................................. 149
Discussion ............................................................................................................ 163
References ............................................................................................................ 167

CHAPTER 6: GENETIC VARIATIONS AMONG SARCOCYSTIS NEURONA AS

REVEALED BY AFLP MARKERS ............................................................................ 169
Abstract ................................................................................................................ 169
Introduction .......................................................................................................... 1 70
Materials and Methods ......................................................................................... 172
Results .................................................................................................................. l 78
Discussion ............................................................................................................ 183
References ............................................................................................................ 1 87

CHAPTER 7: DISCUSSION AND CONCLUDING REMARKS ............................ 189
DISCUSSION ...................................................................................................... 189
CONCLUSIONS .................................................................................................. 202
FUTURE DIRECTIONS ..................................................................................... 206
REFERENCES .................................................................................................... 212

viii

LIST OF TABLES

CHAPTER 2

[1] Purification of Sarcocystis neurona sporocysts from opossum (Didelphis
virginiana) using potassium bromide discontinuous density gradient centrifugation

Table 1. Comparison between mucosa] scraping (current method) and potassium
bromide discontinuous gradient centrifugation method for the isolation of
Sarcocystis spp. sporocysts from opossum small intestine ........................ 37

[2] Generally applicable methods to purify intracellular coccidia from cell cultures
and to quantify purification efficacy using quantitative PCR

Table 1. qPCR primers ............................................................................................. 55
Table 2. Total DNA concentration (ng/ul) in purified and unpurified preparations
of DNA estimated to each contains 1000 merozoites/ZOOul estimated by
UV absorbance spectroscopy ..................................................................... 62
Table 3. Total and parasite DNA content of defined admixtures of crude and
column purified preparations of B. darlingi estimated by UV absorbance
and qPCR targetting parasite rDNA .......................................................... 62
Table 4. Effect of column purification on the detection of host rDNA amplification
in preparations estimated to contain 1000 merozoites/ZOOul .................... 70
Table 5. Effect of column purification on the detection of parasite rDNA
amplification in preparations contain 1000 merozoites/ZOOul .................. 70
Table 6. Effect of column purification on the detection of parasite rDNA in lng/ul
total DNA ................................................................................................... 70
CHAPTER 3
Table 1. Details of Apicomplexan taxa included in the study ................................. 80
Table 2. Names, locations, and DNA sequence of primers used for amplification
and sequencing of portion of SSUrRNA gene from Sarcocystis neurona
isolates from horses and opossums. ........................................................... 82
Table 3. The log likelihood scores (- 1n likelihood) of the topologies of ML, NJ,

and MP trees of the SSUrRNA under best-fit models of nucleotide
evolution selected by LRT and AIC testing using program modeltest. TrN
= (Tamura and Nei 1993), GTR = (Yang, 1994). ...................................... 97

ix

Table 4. Proportion of pairwise nucleotide differences between SSUrRNA
sequences of Sarcocystis spp. and other organisms determined by

Uncorrected ("p") distance analysis of alignment ...................................... 98
CHAPTER 4
Table 1. DNA sequence differences of SA G1 gene of S. neurona isolates based on
pairwise comparisons using uncorrected ("p") distance matrix ............... 122
CHAPTER 5
Table 1. Details of the Sarcocystis species analyzed in the present study. ............ 143
CHAPTER 6
Table 1. Origin, source, year of isolation of taxa used in this study. ..................... 173
Table 2. Selected primer combinations and polymorphism rates for AF LP analysis
of 15 Sarcocystis neurona strains. ........................................................... 179

CHAPTER 2

LIST OF FIGURES

[1] Purification of Sarcocystis neurona sporocysts from opossum (Didelphis
virginiana) using potassium bromide discontinuous density gradient centrifugation

Figure 1.

Figure 2.

Figure 3.

Centrifuge tubes showing a white band (arrows) of Sarcocystis neurona
sporocysts afier potassium bromide discontinuous gradient centrifugation
(A) at 3,000 g, 30 min, and 4°C. (B) At 16,000 g, 1 h, and 4°C ............... 32

Photomicrographs of cleaned and crude S. neurona sporocysts. Crude
sporocysts were collected and cleaned from intestinal mucosa of opossum.
(A, B) Phase contrast microscopy of crude and cleaned sporocysts. Scale
bar = 25 ul (C) Oblique illumination of purified sporocysts. Note the
residual body (white arrow), sporozoite (black arrow), oocyst wall (black
arrowhead), and sporocyst wall (white arrowhead). Scale bar = 100 pl ...36

Results of random amplified polymorphic patterns using the primers
JNBZS/JD396 and restriction enzyme digestion with Hinfl and HindIII of
sporocysts DNA preparations from Sarcocystis neurona isolates from
Michigan opossums (MIOP), S. neurona (MIHl) isolate from horse, and
S. falcatula (Cornell) isolate from opossum. M IOO-bp ladder marker; P
PCR product; Hf Hinfl digest; Hd HindIII digest. Note, after HindIII
digestion of PCR 334 bp amplicons, S. neurona amplicon is unaffected
while S. falcatula amplicon yields bands at 164, 108, and 62 bp. After
Hinfl digestion, S. neurona amplicon yields bands at 180 and 154 bp while
S. falcatula amplicon yields bands at 170 and 164 bp. .............................. 38

[2] Generally applicable methods to purify intracellular coccidia from cell cultures
and to quantify purification efficacy using quantitative PCR

Figure 1.

Figure 2.

Figure 3.

Illustration of the filter assembly. It consisted of a removable 250-ml
borosilicate glass funnel (l) and support base (2). An anodized aluminum
spring clamp (3) sandwiched a cellulose filter pad or filter paper (4)
between funnel and support base. Connection to SOO-ml Erlenmeyer filter
flask (5) is made with a silicone stopper (6). Parasite suspension with
debris (7) and relatively pure parasite suspension (8) ............................... 50

Photomicrographs of S. neurona merozoites. (A) Crude merozoite
preparations before purification. Scale bar = 50 ul. (B) Merozoites (m)
completely free of host cell (hc) contaminants and extraneous debris after
purification using PD-lO column. Scale bar = 25 ul ................................. 57

Purification of Sarcocystis neurona merozoites (A) and Besnoitia darlingi
tachyzoites (B) cultured in Bovine turbinate cells using PD-10 columns,
filter pads, and filter papers compared to unpure (crude) merozoites.

xi

Number of parasites is expressed as x 106. All experiments were
conducted in triplicate ................................................................................ 59

Figure 4. Silver-stained SDS-PAGE gel of whole-cell protein preparations of
Sarcocystis neurona merozoites showing differences in the protein profiles
of crude S. neurona merozoites cultured in BT cells and PD-10 column-
purified merozoites. M, molecular weight markers (Broad Range
Standards; Bio-Rad); Lanes 1 to 3 contain the following amounts of S.
neurona protein extract of column- purified merozoites: 1, 2p]; 2, Sul; 3,
10ul. Lane 4 contains 15p] of S. neurona protein extract of un-pure
merozoites. The numbers on the left indicate the molecular masses of
proteins in kilodaltons (KDa) used as standard markers ........................... 60

Figure 5. Analytical sensitivity of PCR analyses using the diagnostic primers
JNB25/J D396 on DNA Obtained from crude Sarcocystis neurona
merozoites cultured in BT cells and PD-lO column-purified S. neurona
merozoites. Lanes: M, molecular size marker of a 100bp ladder; 1, 2, 3
unpure merozoites (1000, 100, 10, respectively); 4, 5, 6 column- purified
merozoites (1000, 100, 10, respectively); 7, a positive control; 8 a negative
control. Numbers on the right and on the left of the gel are DNA fragment
sizes (in base pairs). ................................................................................... 63

Figure 6. Amplification plot (fluorescence vs. cycle number) in quantitative PCR
assays. (A) Proportional delay in the amplification of parasite rDNA as
crude B. darlingi are diluted, in serial 5-fold steps, over 4 orders of
magnitude. (B) After normalizing crude (red) and column-purified (blue)
extracts to a total DNA concentration of lng/ul, amplification of host 18S
rDNA is delayed. (C) Note amplification of B. darlingi l8S rDNA is
accelerated. (D) Note amplification of B. darlingi hsp70 is accelerated ...66

CHAPTER 3

Figure 1. TEM of S. neurona merozoites cultured in equine dermal (ED) cells and
budding out of the residual body (RB). They had conoids (CO),
micronemes (MN), and dense granules (DG). Note the presence of
dividing nucleus (N) and nucleolous (NO) and the absence of rhoptries,
which exist in S. falcatula. Scale bar = 1pm ............................................. 89

Figure 2. Strict consensus of 117 most parsimonious trees inferred from analysis of
347 phylogenetically informative characters of the nuclear SSUrRNA
gene of 17 Sarcocystis spp. and 11 related taxa and rooted with three
Outgroup taxa. Neighbor-joining (NJ) tree topology based on Tajami Nei
model was almost consistent with the MP trees. Values above and below
the nodes represent bootstrap resampling results (% of 100 replications for
MP analysis) and (% of 1000 replications for NJ algorithm), respectively.
Numbers at the nodes ((1) indicate decay indices. Bootstrap values are
reported only for clades present in >50% of replicates. GenBank

xii

Figure 3.

Figure 4.

CHAPTER 4

Figure 1.

Figure 2.

Figure 3.

accession numbers of the SSUrRNA sequences of the organisms are given
in parentheses. To the right taxonomic affiliation is presented. Branch
lengths are proportional to distance. 50% majority role tree has identical
topology ..................................................................................................... 94

Maximum likelihood phylogeny of SSUrRN A among the Sarcocystidae,
rooted with three outgroup taxa constructed with a GTR + I + G model of
substitution, (-1nL: 7885) heuristic search, random stepwise addition,
TBR, equal character weight, with MULPARS in effect. The log
likelihood score of the phylogeny is —7884.844. Numbers above nodes
represent percent bootstrap support of 100 non-parametric replications.
Branch lengths are proportional to the hypothesized amount of inferred
evolutionary change (ML distances), as shown by the scale bar. Sequences
of the isolates in bold are obtained in this study. Sequences of the isolates
in bold are obtained in this study ............................................................... 96

Phylogram based on SSUrRNA sequence alignment of 31 Sarcocystidae
taxa. Tree was constructed using the Bayesian inference and rooted on E.
tenella, C. cayetanensis, and C. bigenetica. Posterior probabilities are
displayed at each node. Sequences of the isolates in bold are obtained in
this study .................................................................................................. 100

Maximum parsimonious phylogenetic tree from analysis of 208
phylogenetically informative characters of SA GI showing relationships
among the Sarcocystis neurona isolates (rooted to T oxoplasma gondii).
The first of the two numbers at the nodes represent bootstrap resampling
results based on maximum parsimony analysis (MP, % of 1,000
replicates). The second number represents the bootstrap support using
neighbor-joining algorithm (NJ, % of 1,000 replicates). Numbers in
parentheses after taxon names refer to GenBank accession nos., followed
by the host and origin of isolate. Tree statistics are length (L)=290,
consistency index (CI)=O.997, retention index (RI)=0.998 excluding
uninforrnative characters .......................................................................... 123

Diversity in SA GI gene of Sarcocystis neurona. Plot of the variable
nucleotide sites (A) and amino acid sites (B) in pairwise comparisons of
the two alleles of SA GI gene with percentage of the nucleotide
polymorphism shown above the domains. General three-domain structure
model of Sarcocystis neurona SA GI gene is drawn above with the intron
region shaded in grey. The arrow marks the signal peptide sequence
region. Each breakpoint in the domains denotes the location of an
alignment gap. Each vertical line marks the location of a nucleotide or
amino acid difference between the two alleles sequence ......................... 126

Nucleotide alignment of 2 SA GI sequences representing the 2 alleles of S.
neurona SA G] that shows predicted cleavage site for a signal peptide

xiii

(underlined) as identified by Hyun et al., (2003). Dots indicate sequence
identity and dashes indicate alignment gaps ............................................ 127

Figure 4. Plot of the number of substitutions per 100 sites for synonymous (pg) and
nonsynonymous (pN) sites between H1 and H2 SA G1 alleles in a 30-codon
sliding window. The difference (pN_ps) is a measure of the level of

selective constraint on various parts of the molecule .............................. 128
Figure 5. Selective pressures acting upon single codon sites of the SA GI gene using
Approximate Likelihood Ratio (ARS) analysis. ...................................... 129
CHAPTER 5
Figure 1. Map of the United States illustrating the number and distribution of S.
neurona isolates analyzed from different animal hosts in different states in
the present study. ..................................................................................... 150
Figure 2. Nucleotide sequence alignment of the 25/396 marker sequences of

Sarcocystis neurona isolates compared with other Sarcocystis spp. and

- coccidian taxa. Sequences were truncated to the length of published
sequences obtained from opossums from Florida and Mississippi. Bases
that are identical to those of the prototype sequence are indicated by
periods, missing bases are indicated by hyphen, and bases that are
different from those of the prototype sequence are shown. Note the
presence of 2 inserted bases (TA at positions 270 and 271) in Neospora
caninum and in 10 out of 11 of Michigan S. neurona isolates ................ 153

Figure 3. Neighbor-joining (NJ) phylogenetic tree of Sarcocystis neurona isolates
from various hosts and different localities in the US inferred from
sequences of the 25/239 marker using General Time Reversible (GTR)
model. One Sarcocystis sp. isolate (GenBank Accession no. AF 323943)
was used as outgroup. Bootstrap confidence values were based on 1,000
replicates; only those values that exceed 60% are shown above the node.
Branches without bootstrap values represent polytomies that could not be
resolved because the sequences were almost identical. The scale bar
represents the calculated distance value. Branch lengths are drawn
proportional to the amount of genetic change. The isolate designation and
locality follows taxon. For details of isolate designation please refer to

Table 1 ..................................................................................................... 157
Figure 4. Maximum parsimony (MP) phylogenetic tree of Sarcocystis neurona

isolates from various hosts and different localities in the US inferred from

sequences of the 25/239 marker ............................................................... 158
Figure 5. Neighbor-joining (NJ) phylogenetic tree of Sarcocystis neurona isolates

from various hosts and different localities in the US inferred from

xiv

Figure 6.

Figure 7.

CHAPTER 6

Figure 1.

sequences of the 25/239 marker using Kimura two-parameter distance
model ........................................................................................................ 159

Phylogenetic tree of Sarcocystis neurona isolates obtained by minimum
evolution (ME) method ............................................................................ 160

Diversity revealed by RF LP analyses of PCR products of representative
isolates of S. neurona from Michigan. (A) RF LP pattern obtained by
restriction digestion of the 25/396 marker (334 bp amplicons). Lanes 1 to
8, 334 bp undigested amplicons; lanes 9 to 16, samples treated with Hinfl;
lanes 17 to 24, samples treated with HindIII. (B) RF LP pattern obtained by
restriction digestion of the 33/54 marker (1100 bp amplicons). Lanes 1 to
8, 1100 bp undigested amplicons; lanes 9 to 16, Hian restriction profile;
lanes 17 to 24, DraI restriction profile. Lanes M, molecular size standards
are shown to the leftmost and rightmost of the figure in 100 bp-sized
bands; (V) denote empty lanes. Sources of isolates of the amplicons and
restriction analyses: MIH7 (lanes 1, 9, and 17), MIHlO (lanes 2, 10, and
18), MIOPl (lanes 3, 11, and 19), MIOP7 (lanes 4, 12, and 20),
MSUOP/CBS (lanes 5, 13, and 21), MSUOP/CB6 (lanes 6, 14, and 22),
MSUOP/cat2 (lanes 7, 15, and 23), MSUOP/Raccoonl (lanes 8, l6, and
24) ............................................................................................................ 162

An UPGMA dendrogram based on AF LP markers obtained with 9 primer
pairs. Numbers shown at different nodes represent percentage confidence
limits obtained in the bootstrap analysis. Nodes without numbers had
bootstrap values of less than 50. The scale shown above is the measure of
genetic similarity calculated according to J accard similarity coefficient
prepared from AF LP banding patterns of 15 S. neurona strains. Taxon
names and designations are indicated on the right side of the panel. ...... 181

XV

AFLP

ANOVA

ATCC

BT

CSF

Ct

CV-l

DMEM

ED

EMEM

EPM

HBSS

Kbr

ML

MP

NJ

PBS

PCR

RAPD

SAGl

SSUrRNA

UPGMA

KEY TO ABBREVIATIONS
Amplified fragment length polymorphism
Analysis of variance
American Type Culture Collection
Bovine turbinate cell line
Cerebrospinal fluid
Cycle at which fluorescence significantly exceeds baseline levels
African green monkey (Cercopithecus aethiops) kidney cells
Dulbecco’s minimal essential medium
Equine dermal cell line
Eagle’s minimal essential medium
Equine protozoal myeloencephalitis
Hanks’ balanced salt solution
Potassium bromide
Maximum Likelihood
Maximum Parsimony
Neighbor Joining
Phosphate buffered saline
Polymerase chain reaction
Random amplified polymorphic DNA
Maj or surface antigen genel
Small subunit ribosomal RNA gene

Unweighted Pair Group Method with Arithmatic Mean

xvi

THE RESEARCH PROBLEM (SCOPE AND SIGNIFICANCE)-

Equine protozoal myeloencephalitis (EPM) is a serious neurological disease of
horses in the Americas caused by the protozoal parasite, Sarcocystis neurona (Dubey et
al., 2001b). Over 50% of the horses in the US have been infected with S. neurona based
on finding antibodies against the parasite in their blood (Bentz et al., 1997; Blythe et al.,
1997; MacKay et al., 2000; Dubey et al., 2001b; Bentz et al., 2003; Rossano et al., 2003).
The potential risk factors and characteristics associated with transmission of this
pathogen are not fully understood. Several very closely related Sarcocystis organisms
have been identified as S. neurona based on molecular procedures. However, these
organisms have been shown to have different biological characteristics. Even though
they all infect the opossum as the definitive host (with sexual stages), they do not all
infect the same intermediate hosts (with the asexual stages). Also, they do not all
produce disease in the gamma interferon gene knockout mice (y-IFN-KO) that have been
used to mimic infection in the horse (Cheadle et al., 2001; Dubey etal., 2001a, 2001b).
These findings suggest that they vary in important ways that are likely to affect the type
of disease produced or even whether they are able to produce disease in horses.
Veterinarians have observed that not all horses exposed to the parasite develop the
disease, and when disease exists, clinical signs vary from mild ataxia to paralysis,
recumbency, and death. Also, different horses have different responses to treatment,
suggesting that strain variations in the parasite may be an important factor in the
treatment of EPM. Strain variation may contribute to problems in diagnosis of EPM and

will certainly affect any vaccine that is made to protect horses.

With the increasing awareness for the potential of S. neurona as an important
cause of economic losses in the horse industry, knowledge of the mechanisms by which it
causes disease are of great importance. In this regard, understanding the extent and nature
of genetic variation in S. neurona and closely related species is an essential prerequisite
to determining the epidemiology and transmission of this pathogen. It is especially
important to know whether genotypes of S. neurona are stable (indicative of clonal
propagation) or unstable (due to frequent genetic recombination). For example, in a
related protozoan T oxoplasma gondii, population studies demonstrated that this organism
is genetically stable, and the small amount of variation defined three clonal lineages
where phenotypes such as virulence are associated with a single lineage (Howe and
Sibley, 1995). This situation may be due to the fact that only one isolate of T oxoplasma is
usually found in a cat at a given time; therefore, the potential for sexual recombination
(which can only occur in the host in which sexual reproduction takes place) is low. Other
protozoa have high potential for sexual recombination, which generates diversity and
enhances virulence attributes (Tibayrenc and Ayala, 2002; Mallon et al., 2003). The
population structure of S. neurona has not been determined yet. However, the molecular
tools and knowledge supporting the possibility of sexual recombination in this parasite
are now available. It is well known that sexual reproduction of S. neurona takes place in
the gut of opossums. Also, the recovery of many Sarcocystis strains from the gut of a
single opossum shows that they co-exist and could undergo recombination (Elsheikha et
al., 2004). Therefore, to test adequately the “recombination model” or “non-clonal

model” theory for S. neurona and to determine population structure, a population genetic

study was conducted in which data were analyzed in terms of allelic and genotypic
frequencies, rather than simply assessing genetic distances as was done in the past.

The objectives of present research were (1) to determine the population structure,
phylogenetic relationships, and extent of genetic diversity among Sarcocystis spp. '
infecting horses and opossums in relation to closely related species using sequencing of
multiple genetic markers and the Amplified Fragment Length Polymorphism (AF LP)
methods that provide a genetic fingerprint and (2) to use this knowledge for thorough
understanding of the molecular epidemiology and evolution of S. neurona at the
molecular level. The specific aims of this work were: (1) to develop an accurate,
reproducible, reliable and sensitive method for genotyping of clinical strains of S.
neurona from horses and environmental strains from opossums to subspecies level based
on molecular markers; (2) to elucidate the nature and extent of genetic variation as well
as the heterogeneity of Sarcocystis species isolated from horses and opossums; (3) to
improve the systematic classification of S. neurona strains according to their evolutionary
relationships; (4) to use phylogenetic information to predict the pattern of distribution of
S. neurona among equines and wild host species; and (5) to refine the definition of S.
neurona species, strains, clones, and populations using molecular evolutionary analysis.
This work was focused on S. neurona, but the genetic fingerprints of S. neurona isolates
were compared to diverse but related organisms such as S. falcatula, B. darlingz’, T.
gondii, and N. caninum. The work appears to be the first wide—ranging comparative study

of S. neurona and related species from domesticated and wild animal hosts.

REFERENCES

Bentz BG, Granstrom DE, Stamper S (1997) Seroprevalence of antibodies to Sarcocystis
neurona in horses residing in a county of southeastern Pennsylvania. J Am Vet Med
Assoc 210:517—518

Bentz BG, Ealey KA, Morrow J, Claypool PL, Saliki JT (2003) Seroprevalence of
antibodies to Sarcocystis neurona in equids residing in Oklahoma. J Vet Diagn Invest
62597—600

Blythe LL, Granstrom DE, Hansen DE, Walker LL, Bartlett J, Stamper S (1997)
Seroprevalence of antibodies to Sarcocystis neurona in horses residing in Oregon. I Am
Vet Med Assoc 210:525-527

Cheadle MA, Tanhauser SM, Scase TJ, Dame JB, Mackay RJ, Ginn PE, Greiner EC
(2001) Viability of Sarcocystis neurona sporocysts and dose titration in gamma-
interferon knockout mice. Vet Parasitol 95:223—231

Dubey JP, Lindsay DS, Kwok OC, Shen SK (2001a) The gamma interferon knockout
mouse model for Sarcocystis neurona: comparison of infectivity of sporocysts and
merozoites and routes of inoculation. J Parasitol 872117—1173

Dubey JP, Lindsay DS, Saville WJA, Reed SM, Granstrom DE, Speer CA (2001b) A
review of Sarcocystis neurona and equine protozoal myeloencephalitis (EPM). Vet
Parasitol 95289—1 31

Elsheikha HM, Murphy AJ, Mansfield LS (2004) Prevalence of Sarcocystis species
sporocysts in Northern opossums (Didelphis virginiana). Parasitol Res 93:427—431

Howe DK, Sibley LB (1995) T oxoplasma gondii comprises three clonal lineages:
correlation of parasite genotype with human disease. J Infect Dis 172:561—1566

MacKay RJ, Granstrom DE, Saville WJ, Reed SM (2000) Equine protozoal
myeloencephalitis. Vet Clin North Am Equine Pract 16:405—425

Mallon M, MacLeod A, Wastling J, Smith H, Reilly B, Tait A (2003) Population
structures and the role of genetic exchange in the zoonotic pathogen Cryptosporidium
parvum. J Mol Evol 561407—417

Rossano MG, Kaneene JB, Marteniuk JV, Banks BD, Schott HC 2nd, Mansfield LS
(2003) A herd-level analysis of risk factors for antibodies to Sarcocystis neurona in
Michigan equids. Prev Vet Med 57:7—13

Tibayrenc M, Ayala FJ (2002) The clonal theory of parasitic protozoa: 12 years on.
Trends Parasitol 18:405—41 0

HYPOTHESES TESTED
Whether in nature the parasitic protozoan Sarcocystis neurona reproduces
sexually or clonally remains largely unsettled. This ignorance may seem surprising given
that the matter is of great medical consequence: the strategies for developing vaccine and
curative drugs as well as for diagnosis and treatment are different for an organism that
propagates clonal, than for an organism that propagates with frequent sexual

recombination.

Overall hypothesis
The overall hypothesis tested in this work was that: (H0) S. neurona has
panamictic genetic structure due to frequent genetic recombination during the sexual
phase of parasite development in the intestinal cells of the definitive host leading to
strains with variable genetic traits including increased virulence. An alternate hypothesis
(H1) was that S. neurona has a clonal population structure and virulence is associated
with a single lineage. Testing of this hypothesis can provide a clearer understanding of

the population structure and transmission characteristics of this ubiquitous parasite.

Specific hypotheses
CHAPTER 3
Hypothesis 1: S. neurona strains from opossums and horses form a monophyletic group.
Ho: S. neurona strains form a single monophyletic group.

H]: S. neurona strains are not monophyletic.

Hypothesis 2: Phylogenies of S. neurona and its hosts are consistent with a common
history.

Ho: Associated hosts and parasites have identical phylogenies.

H1: There are different phylogenies for hosts and parasites.

Rationale: These two hypotheses were investigated by determining the phylogenetic
relationships of S. neurona strains obtained from horses and opossums in relation to other
cyst-forming coccidia based on sequences of the small subunit ribosomal RNA

(SSUrRNA) gene.

CHAPTER 4

Hypothesis 3: SAGl gene of S. neurona evolved under positive selection pressure.
Ho: The SAGl gene of S. neurona has evolved under diversifying selection pressure.
H1: The SAGl gene has not evolved under positive selection pressure.

Rationale: Sequences of the SAGl gene of multiple S. neurona strains were used to

investigate this hypothesis.

CHAPTER 5

Hypothesis 4: S. neurona strains have a clonal population structure.

Ho: S. neurona strains exhibit phylogenetic congruence and clonality.

H1: S. neurona strains include diverse, distinct genotypes.

Rationale: Sequences of the diagnostic 25/396 marker were obtained from multiple S.
neurona strains from different hosts in the United States and were used to explore this

hypothesis.

CHAPTER 6

Hypothesis 5: A genome-wide fingerprinting analysis using AF LP resolves the intra-
species relationships among S. neurona.

Ho: AF LP resolves the intraspecific relationships among S. neurona.

H1: AF LP does not resolve the intraspecific relationships among S. neurona.
Rationale: To investigate this hypothesis, the phylogenetic relationships between S.
neurona strains were constructed based on fingerprinting profiles of S. neurona strains

obtained by the AF LP technique.

CHAPTER 1
LITERATURE REVIEW

Life cycle of Sarcocystis neurona. Sarcocystis species have an obligatory, two-
host life cycle (F ayer et al., 1982; Dubey et al., 1989a). Sarcocystis species replicate
asexually in the muscles or other tissues of a prey intermediate host, where they form
tissue cysts that are infective to an appropriate carnivorous definitive host. The parasite
reproduces sexually in the intestine of the definitive host, ultimately producing sporocysts
that are passed in feces. Ingestion of these sporocysts by an intermediate host species
continues the life cycle. Opossums have been identified as the natural definitive host of
Sarcocystis neurona (Fenger et al., 1995, 1997). A recent study revealed that domestic
cats can serve as intermediate hosts for S. neurona when they are experimentally infected
with S. neurona sporocysts obtained fi'om naturally infected opossums (Dubey et al.,
2000). Tissue cysts produced in these cats were infective to opossums and were
subsequently confirmed to be S. neurona. Also, numerous S. neurona strains have been
isolated from various geographic regions and biological hosts in the US, including nine-
banded armadillos, skunks, raccoons, sea otters, harbor seals, and cats (Dubey et al.,
2000; Dubey et al., 2001b; Cheadle et al., 2001b, 2001c; Tanhauser et al., 1999). Horses
are infected with the parasite when they ingest sporocysts in food contaminated with
opossum feces (Fenger et al., 1995). The sequence of events that occurs when the
parasite enters the horse is not known. It has been speculated that in the horse the
parasite enters the blood vessels from the gut and then invades the central nervous system
(rather than muscle), destroying cells as it multiplies. The only stage of the parasite found

to date in horses is the asexual merozoite stage in neural cells. S. neurona sarcocysts are

absent fi'om other horse tissues suggesting that horses are an accidental or “dead-end”
host for the parasite. As such, an infected animal cannot transmit the parasite to another
horse or back to the opossum. The absence of S. neurona sarcocysts in horses raised
some concerns regarding the authenticity of horses as a true intermediate host (I.H).
Sarcocystis neurona and equine protozoal myeloencephalitis (EPM). S.
neurona is the most commonly diagnosed neurological disease EPM in the horse
population in North America (Dubey et al., 2001 a). Although other coccidia have been
also associated with the distinct clinical manifestations of EPM, only S. neurona is well
known to be responsible for causing EPM disease. Approximately 50% (20 to 90%) of
horses in the US have been infected by this organism as evidenced by antibodies against
the parasite in their serum (Bentz et al., 1997; Blythe et al., 1997; MacKay et al., 1997,
2000; Bentz et al., 2003; Rossano et al., 2003). Market research conducted with equine
veterinarians revealed that EPM was diagnosed in 220,000 horses in 1998 with a
conservative estimate of 10,000 to 160,000 new cases every year. Approximately 70% of
diseased horses improved with therapy, but less than one half returned to normal
function. EPM is a serious, growing problem for the horse industry in the US. Animals
affected by EPM can demonstrate a variety of clinical abnormalities, and signs can vary
in severity from mild lameness to death. EPM-affected horses are unable to attain their
ultimate performance. Despite treatment, horses with EPM can be affected so severely
that they must be euthanatized, or they can develop permanent disabilities, which can
affect their athletic activities. Treatment of EPM is expensive, and both mildly and
severely affected horses often require treatment for extended periods without guarantee

of success (MacKay et al., 2000; Dubey et al., 2001a).

Economists from the USDA estimated that the average cost to the equine industry
in the US due to EPM, is $27 million annually (Dubey et al., 2001a). This figure includes
estimates for costs related to diagnosis, treatment, and maintenance of horses while they
are unable to be used for their intended purpose, loss of stakes payments, and the value
and transport costs of horses that die or are euthanatized. This estimate does not account
for opportunity losses because affected horses are unable to attain their ultimate
performance potential. The National Animal Health Monitoring System’s Equine ‘98
study rated “Gathering information about this disease (EPM)” as one of the highest
priorities for the equine industry (NAHMS, 1998). Veterinarians throughout the US have
diagnosed this disease with increasing frequency since commercial tests to detect S.
neurona infection have become available. However, conventional diagnostic methods for
EPM have suffered from problems with false positive tests. That is, normal non-
neurological horses often have S. neurona antibodies in serum and cerebrospinal fluid.

Sarcocystis neurona was first isolated and characterized in 1991 by Dubey and
colleagues. Even today there are few definitive morphological characteristics with which
to identify Sarcocystis taxa that infect horses. Also, there have been difficulties in
characterizing S. neurona strains cultured in the laboratory in order to clearly identify
species. At least four species of Sarcocystis have been described in the opossum
including S. neurona (Dubey et al., 1991), S. falcatula (Box and Duszynski, 1980), S.
speeri (Dubey and Lindsay, 1999), S. lindsayi (Dubey et al., 2001c), and each species has
been considered in turn as a possible causative agent of EPM in horses. Recently,
another Sarcocystis sp. that cycles between opossum and the brown-headed cowbird

(Molothrus ater) was identified and found to causes neurological disease in the mouse

10

model of EPM (Mansfield et al., unpublished data). All of these species have sexual
stages (sporocysts) in the opossum gut that cannot be easily distinguished
morphologically in fecal specimens or at necropsy, and this further complicates the
determination of parasite diversity.

Variation between isolates has been observed for some biological and molecular
characters. Not all horses exposed to the parasite develop the disease; but when disease
exists, clinical signs can vary from mild ataxia to paralysis, recumbency, and death
(MacKay et al., 2000). Parasite diversity may also be reflected in variable responses to
treatment in horses infected with S. neurona suggesting that strain-to-strain variation may
be a factor in the clinical management of EPM. The fact that horses are exposed to
multiple Sarcocystis spp. in the environment may complicate the understanding of the life
cycle of S. neurona and of the pathogenesis and diagnosis of EPM. Clearly, additional
features must be found to aid in the identification and characterization of Sarcocystis spp.
The application of sensitive molecular approaches has obviated the need for expensive,
cumbersome, and time-consuming biological infection studies as the sole means of
distinguishing species.

Taxonomic status of Sarcocystis neurona. The authors of early studies named
and described different Sarcocystis spp. on the basis of host occurrence and the
appearance of the muscle sarcocysts (Wenvon, 1926; Boch et al., 1979). Following the
discovery in 1972 of the obligate heteroxenous (requiring two hosts) life cycle for the
parasite, species were described on the basis of both intermediate and definitive host
specificity (Heydron and Matuschka, 1981). Then, reports started to identify and to

classify S. neurona mainly on phenotypic characteristics such as ultrastructural analysis,

11

host specificity (the ability to replicate/cause disease in certain hosts), details of the life
cycle, or location of the developmental stages. Accurate identification of a Sarcocystis
sp. using solely morphological criteria may not be adequate, yet distinguishing the true
host for a given Sarcocystis sp. from related Sarcocystis is critical to understanding and
controlling of infection. Indeed, a new system of nomenclature was proposed for those
species in which both host species were known (Heydron et al., 1975). However, cross-
transmission studies have indicated that the host specificity of certain species is not as
strict as previously thought. Furthermore, although certain morphological and
ultrastructural characteristics of mature muscle cysts have been used to identify species
(Mehlhom, 1976), some studies have revealed slight morphological heterogeneity
occurring among cysts presumed to have belonged to single species.

The identification of S. neurona has relied on biological infection patterns in a
variety of animals. Biotyping based on susceptibility of interferon gamma knockout mice
to S. neurona infection and the pattern of grth in tissue culture were the first methods
used to differentiate between S. neurona, S. falcatula, and S. speeri. Whereas S. neurona
and S. speeri sporocysts infect immunodeficient mice but not budgigars, the reverse is
true for parasites designated S. falcatula (Marsh et al., 1997a; 1997b; Dubey and Lindsay,
1998). However, bioassay and development patterns in tissue culture systems are time-
consuming and thus not-very suitable tools for systematic and epidemiological studies.

Previous genetic and phylogenetic studies of S. neurona. Phylogenetic analysis
can provide a quantitative estimate of genetic relatedness among S. neurona strains and
patterns of divergence between members of this species and other Sarcocystis species. In

the past decade, various methods have been developed for the identification and typing of

12

S. neurona strains. The first genetic analysis of S. neurona was done using a random
amplified polymorphic DNA assay to compare S. neurona to Sarcocystis spp.,

T oxoplasma gondii, and several Eimeria spp. (Granstrom et al., 1994). Subsequently, the
nuclear small subunit-ribosomal SSUrRNA gene was amplified from cultured S. neurona
organisms, cloned, and sequenced (Fenger et al., 1994). Sequence information was used
to design a PCR assay to screen DNA extracts from sporocysts shed by various candidate
definitive hosts. The SSUrRNA gene from sporocyst DNA was cloned and sequenced
and found to be identical to the SSUrRNA gene sequence of culture-derived S. neurona.
Following the identification of the opossum as the definitive host (F enger et al., 1995),
the SSUrRNA gene sequence of S. falcatula, which cycles between opossums and
various birds, was determined (Fenger et al., 1994). While the sequences were identical,
there is increasing evidence that suggests that Sarcocystis isolates from opossums were
not a single uniform species, and subsequent biologic, morphologic, and molecular
evidence showed that S. neurona and S. falcatula were distinct species (Dame et al.,
1995; Marsh et al., 1999; Cheadle et al., 2001a).

Phylogenetic analyses of the SSUrRNA gene from members of the family
Sarcocystidae have been used to study relatedness and to help complete parasite life
cycles. Initially, it was believed that protozoa co-evolved with the definitive host (Barta,
1989). The SSUrRNA gene sequence of S. neurona and results from experimental
infection studies placed it in the felid clade of Sarcocystis (F enger et al., 1994). A recent
analysis using SSUrRNA gene sequences from many more Sarcocystis spp. suggested
intermediate host (5) were involved. S. neurona was placed in the non-ruminant clade

and was found to be most closely related to S. mucosa, a parasite believed to use the

13

Bennetts wallaby (Dayrus sp.) as a definitive host (Jakes, 1998; Jenkins et al., 1999).
Recently, molecular data in the form of DNA sequence of the small subunit ribosomal
RNA genes (SSUrRNA) and the D2 domain of the large subunit ribosomal RNA genes
have been used to reconstruct phylogenies and to examine the relationships of S. neurona
and other Sarcocystis spp. (Slapeta et al., 2003).

In spite of a wide array of analytical methods being available, the characterization
of S. neurona strains below the species level remains a demanding task. Generally the
lack of discriminatory power of serological methods greatly limits their applicability to
the typing of S. neurona. Protein profiling methods such as gel electrophoresis and
immunoblotting (Granstrom et al., 1993) also possess little discriminatory power and are
time-consuming, require large samples for analysis, and may be difficult to interpret.
Various DNA-based methods have been used for interspecies differentiation of
Sarcocystis strains from horse and opossum. They include PCR-mediated restriction
fragment length polymorphism and arbitrarily primed PCR (Tanhauser et al., 1999).
However, these methods differ in their taxonomic range, discriminatory power,
reproducibility, and ease of interpretation. In addition, these genetic typing assays also
have drawbacks in that they require a relatively large amount of high-quality DNA and
are difficult to standardize between laboratories. Thus, the need for high-resolution
analytical tools, which will facilitate the typing of S. neurona on a routine basis, still
exists. The ideal genotyping method produces results that are invariable from laboratory
to laboratory and allows unambiguous comparative analyses and the establishment of
reliable databases. One of the newest and most promising methods is amplified fragment

length polymorphism (AF LP) analysis (Blears et al., 1999; V05 et al., 1995), developed

14

by Keygene BV, Wageningen, The Netherlands. This method combines universal
applicability with high powers of discrimination and reproducibility. An increasing
number of reports describe the use of AF LP analysis for animal genetic mapping, medical
diagnostics, phylogenetic studies, and microbial typing (Blears et al., 1999). For
example, AF LP analysis revealed a remarkably high degree of genetic diversity within
Cryptosporidium parvum strains (Blears et al., 2000).

AF LP is a whole-genome fingerprinting method based on selective amplification
of restriction fragments (Vos et al., 1995; Blears et al., 1999). The AF LP reaction is a
multistep procedure, which combines the power of PCR with the inforrnativeness of
restriction enzyme analysis. The procedure includes the preparation of an AF LP template:
genomic DNA digested with two restriction endonucleases that produce cohesive
fragment ends and that cut DNA with different frequencies (rare cutter [RC] and frequent
cutter [FC]). Following digestion, genomic restriction fragments are modified by ligation
of synthetic adapters (RC adapter and F C adapter) with ends complementary to those of
the restriction fragments. Thus, after the ligation step, genomic restriction fragments
have termini with known sequences. Such an AF LP template is submitted to a highly
stringent PCR amplification with primers fully complementary to their targets (adaptor
and restriction site sequence), with additional nucleotides at the 3’ end, are used as
selective oligomers to amplify a subset of ligated fragments. Only those fragments with
complementary nucleotides extending beyond the restriction site are amplified by the
selective primer under stringent annealing conditions. Fluorescent dye-labeling one of the
AF LP primers (usually the RC primer) allows the detection of only a subset of the

hundreds of amplified restriction fragments. Polymorphisms are identified by the

15

presence or absence of DNA fragments after separation on polyacrylarnide gels or a DNA
sequencer. The technique has several advantages over other DNA fingerprinting systems.
AF LP usually yields more complex banding patterns than most of the available DNA
fingerprinting methods, which may increase discrimination between strains under
analysis. The strategy of using two restriction enzymes and selective amplification
provides a great flexibility in designing the typing protocols optimal for the given species.
The most important of these are the capacity to inspect an entire genome for
polymorphism without prior knowledge of the target DNA sequence and the
reproducibility of the method. These features, combined with the automation and a high-
throughput analysis, make AF LP superior to the currently used molecular techniques. In
the present study, an AF LP-based procedure suitable for the inter- and intraspecies
differentiation of S. neurona species of clinical and environmental origin was evaluated.

Genetic variation among Sarcocystis neurona isolates. There is little
information regarding genetic variation of S. neurona isolated from horses with EPM and
other Sarcocystis spp. isolated fi‘om opossums. Recent studies of S. neurona support the
evidence of isolate diversity obtained by phenotypic and western blot analysis of
merozoite proteins (Mansfield et al., 2001; Marsh et al., 2001). S. neurona is
differentially distributed in multiple mammalian hosts and may be maintained through
different transmission cycles in nature. Meningoencephalitis due to a S. neurona-like
protozoan was reported in marine mammals: Pacific harbor seals (Phoca vitulina
richardsi) (Lapointe et al., 1999) and Alaskan sea otters (Enhydra lutris) (Rosonke et al.,
1999). The parasitic zoites in infected tissues reacted with S. neurona antibodies

although the habitat and behavior of the harbor seals would most likely exclude contact

16

with opossum feces, and it would appear highly unlikely that these marine mammals
became infected from the same sporocyst source as the horse. Thus, it is possible that the
protozoa described in numerous species, including raccoons (Dubey et al., 1990),
northern gannets (Spalding et al., 2002), sheep (Dubey et al., 1989b), skunks (Dubey et
al., 1996), and mink (Dubey and Hedstrom, 1993) as “S. neurona-like”, are actually
different species with varying life cycles. To date, the population structure and the
relationships among S. neurona strains threatening equine health remain ill defined.
Further, the question should be raised about how much variation exists between different
S. neurona strains that infect horses.

These previous studies have found considerable evidence of genetic heterogeneity
among Sarcocystis strains from different species of vertebrates, and there is now
mounting evidence suggesting that a series of host-adapted genotypes/strains/species of
the parasites exists. Unfortunately, recognition of S. neurona as the causative agent of
EPM in horses and other mammals is probably an oversimplification. First, there might
be some strains containing a mixture of genotypic alleles. Second, studies to date have
failed to identify recombinant genotypes, suggesting that most genotypically defined
strains are reproductively isolated populations. This contrasts with the present taxonomy
of the species, which includes only one recognized genotype. These studies clearly
indicate that the current classification of Sarcocystis spp. affected horses and the species
level taxonomy of S. neurona does not appear to reflect molecular phylogenetic analyses
or biological reality, and this situation is likely to be a limitation in epidemiological
investigations. Indeed, current molecular data provide support for some of the earlier

taxonomic separations on the basis of host occurrence (Wenvon, 1926).

17

Although the previous studies have confirmed genetic diversity among Sarcocystis
strains from horses and opossums, most have been small studies with few cases.
Therefore, there is a need for population-level information on the population structure
and genetic diversity of S. neurona and non-S. neurona isolates. A larger study
encompassing diverse populations is necessary before we can have a more complete
understanding of the population structure and molecular epidemiology of S. neurona and
other Sarcocystis spp. Further sampling and molecular description will almost certainly
show that S. neurona actually comprises several cryptic species. Determining the
population structure and understanding the nature and extent of genetic heterogeneity
among S. neurona strains is a task of great importance from an economic and equine
health standpoint in the US. This information will be valuable for understanding
virulence traits and for unraveling the broad evolutionary history of this species.
Improved fingerprinting methods are required if fundamental questions about

epidemiology and risk factors are to be addressed.

18

 

REFERENCES

Barta JR (1989) Phylogenetic analysis of the class Sporozoea (phylum Apicomplexa
Levine, 1970): evidence for the independent evolution of heteroxenous life cycles. J
Parasitol 75: 195—206

Bentz BG, Granstrom DE, Stamper S (1997) Seroprevalence of antibodies to Sarcocystis
neurona in horses residing in a county of southeastern Pennsylvania. J Am Vet Med
Assoc 210:517—518

Bentz BG, Ealey KA, Morrow J, Claypool PL, Saliki JT (2003) Seroprevalence of
antibodies to Sarcocystis neurona in equids residing in Oklahoma. J Vet Diagn Invest
6:597—600

Blears MJ, De Grandis SA, Lee H, Trevors J T (1999) Amplified fragment length
polymorphism (AF LP): review of the procedure and its applications. J Ind Microbiol
Biotechnol 21 :99—1 14

Blears MJ, Pokomy NJ, Carreno RA, Chen S, De Grandis SA, Lee H, Trevors J T (2000)
DNA Fingerprinting of Cryptosporidium parvum Strains Using Amplified Fragment
Length Polymorphism (AF LP). J Parasitol 86:838—841

Blythe LL, Granstrom DE, Hansen DE, Walker LL, Bartlett J, Stamper S (1997)
Seroprevalence of antibodies to Sarcocystis neurona in horses residing in Oregon. J Am
Vet Med Assoc 210:525—527

Boch J, Bierschenck A, Erber M, Weiland G (1979) Sarcocystis und T oxoplasma-
Infektionen bei Schlachtschafen in Bayem. Berl Munch Tierarztl Wochenschr 92: 137—
141

Box ED, Duszynski DW (1980) Sarcocystis of passerine birds: sexual stages in the
opossum (Didelphis virginiana). J Wildl Dis 18:209—215

Cheadle MA, Dame JB, Greiner EC (2001a) Sporocyst size of strains of Sarcocystis shed
by the Virginia opossum (Didelphz's virginiana). Vet Parasitol 95:305—311

Cheadle MA, Tanhauser SM, Dame JB, Sellon DC, Hines M, Ginn PE, MacKay RJ,
Greiner EC (2001b) The nine-banded armadillo (Dasypus novemcinctus) is an
intermediate host for Sarcocystis neurona. Int J Parasitol 312330—335

Cheadle MA, Yowell CA, Sellon DC, Hines M, Ginn PE, Marsh AE, Dame JB, Greiner

EC (2001c) The striped skunk (Mephitis mephitis) is an intermediate host for Sarcocystis
neurona. Int J Parasitol 31 :843—849

19

Dame JB, MacKay RJ, Yowell CA, Cutler TJ, Marsh A, Greiner EC (1995) Sarcocystis
falcatula from passerine and psittacine birds: synonymywith Sarcocystis neurona, agent
of equine protozoal myeloencephalitis. J Parasitol 81 :93 0—935

Dubey JP, Davis SW, Speer CA, Bowman DD, de Lahunta A, Granstrom DE, Topper
MJ, Hamir AN, Cummings JF, Suter MM (1991) Sarcocystis neurona n. sp. (Protozoa:
Apicomplexa), the etiologic agent of equine protozoal myeloencephalitis. J Parasitol
77 :2 1 2—2 1 8

Dubey JP, Hamir AN, Hanlon CA, Topper MJ, Rupprecht CE (1990) Fatal necrotizing
encephalitis in a raccoon associated with a Sarcocystis-like protozoan. J Vet Diagn Invest
2:345—347

Dubey JP, Hamir AN, Niezgoda M, Rupprecht CE (1996) A Sarcocystis neurona-like
organism associated with encephalitis in a striped skunk (Mephitis mephitis). J Parasitol
82: 1 72—1 74

Dubey JP, Hedstrom OR (1993) Meningoencephalitis in mink associated with a
Sarcocystis neurona-like organism. J Vet Diagn Invest 5: 467—471

Dubey JP, Lindsay DS (1998) Isolation in immunodeficient mice of Sarcocystis neurona
from opossum (Didelphis virginiana) faeces, and its differentiation fiom Sarcocystis
falcatula. Int J Parasitol 28:1823—1828

Dubey JP, Lindsay DS (1999) Sarcocystis speeri n. sp. (Protozoa: Sarcocystidae) from
the opossum (Didelphis virginiana). J Parasitol 85:903—909

Dubey JP, Lindsay DS, Saville WJA, Reed SM, Granstrom DE, Speer CA (2001a) A
review of Sarcocystis neurona and equine protozoal myeloencephalitis (EPM). Vet
Parasitol 95:89—131

Dubey JP, Speer CA, Fayer R (1989a) Sarcocystosis of Animals and Man. CRC Press,
Inc, Boca Raton

Dubey JP, Speer CA, Munday BL, Lipscomb TP (1989b) Ovine sporozoan
encephalomyelitis linked to Sarcocystis infection. Vet Parasitol 34:159—163

Dubey JP, Saville WJ, Stanek JF, Lindsay DS, Rosenthal BM, Oglesbee MJ, Rosypal
AC, Njoku CJ, Stich RW, Kwok OC, Shen SK, Hamir AN, Reed SM (2001b) Sarcocystis
neurona infections in raccoons (Procyon lotor): evidence for natural infection with
sarcocysts, transmission of infection to opossums (Didelphis virginiana), and
experimental induction of neurologic disease in raccoons. Vet Parasitol 100:117—129

Dubey JP, Rosenthal BM, Speer CA (2001c) Sarcocystis lindsayi n.sp. (Protozoa:

Sarcocystidae) from the South American opossum, Didelphis albiventris from Brazil. I
Eukar Microbiol 48:595-603

20

Dubey JP, Saville WJ, Lindsay DS, Stich RW, Stanek JF, Speer CA, Rosenthal BM,
Njoku CJ, Kwok OC, Shen SK, Reed SM (2000) Completion of the life cycle of
Sarcocystis neurona. J Parasitol 86: 1276—1280

Fayer R, Dubey JP, Leek RG (1982) Infectivity of Sarcocystis spp. from bison, elk,
moose and cattle for cattle vis sporocysts from coyotes. J Parasitol 681681—685

Fenger CK, Granstrom De, Langemeier J L, Gajadhar A, Cothran G, Tramontin RR,
Stamper S, Dubey JP (1994) Phylogenetic relationship of Sarcocystis neurona to other
members of the family Sarcocystidae based on the sequence of the small ribosomal
subunit gene. J Parasitol 79:966—975

Fenger C K, Granstrom DE, Langemeier J L, Stamper S, Donahue J M, Patterson J S,
Gajadhar AA, Marteniuk JV, Xiaomin Z, Dubey JP (1995) Identification of opossums
(Didelphis virginiana) as the putative definitive host of Sarcocystis neurona. J Parasitol
81 :91 6—9 1 9

Fenger CK, Granstrom DE, Gajadhar AA, Williams NM, McCrillis SA, Stamper S,
Langemeier J L, Dubey JP (1997) Experimental induction of equine protozoal
myeloencephalitis in horses using Sarcocystis sp. sporocysts from the opossum
(Didelphis virginiana). Vet Parasitol 68: 199—2 1 3

Granstrom DE, Dubey JP, Davis SW, Payer R, Fox JC, Poonacha KB, Giles RC, Comer
PF (1993) Equine protozoal myeloencephalitis: antigen analysis of cultured Sarcocystis
neurona merozoites. J Vet Diagn Invest 5:88—90

Granstrom DE, McPherson J M, Gaj adhar AA, Dubey JP, Tramontin R, Stamper S
(1994) Differentiation of Sarcocystis neurona from eight related coccidia by random
amplified polymorphic DNA assay. Mol Cell Probes 82353—356

Heydron AO, Gestrich R, Mehlhom H, Rommel M (1975) Proposal for a new
nomenaclature of the Sarcosporidia. Z Parasienkd 48:73—82

Heydron AO, Matuschka FR (1981) Zur Endwirtspezifitat der vom Hund ubertragenen
sarkosporidienarten. Z Parasitenkd 66:231—234

J akes KA (1998) Sarcocystis mucosa in Bennetts wallabies and pademelons from
Tasmania. J Wildl Dis 34:594—599

Jenkins MC, Ellis JT, Liddell S, Ryce C, Munday BL, Morrison DA, Dubey JP (1999)
The relationship of Hammondia hammondi and Sarcocystis mucosa to other heteroxenous
cyst-forming coccidia as inferred by phylogenetic analysis of the 18S SSU ribosomal
DNA sequence. Parasitology 119235—142

Lapointe JM, Duignan PJ, Marsh AE, Gulland FM, Barr BC, Naydan DK, King DP,
Farman C, Burek KA, Lowenstine J L (1998) Meningoencephalitis due to a Sarcocystis

21

neurona-like protozoan in Pacific Harbor seals (Phoca vitulina richardsi). J Parasitol
84:1184—1189

MacKay RJ (1997) Equine protozoal myeloencephalitis. Vet Clin North Am Equine Pract
1 3 :79—96

MacKay RJ, Granstrom DE, Saville WJ, Reed SM (2000) Equine protozoal
myeloencephalitis. Vet Clin North Am Equine Pract 16:405—425

Mansfield LS, SchottII HC, Murphy AJ, Rossano MG, Tanhauser SM, Patterson J S,
Nelson K, Ewart SL, Marteniuk JV, Bowman DD, Kaneene J B (2001) Comparison of
Sarcocystis neurona strains derived from horse neural tissue. Vet Parasitol 95267—178

Marsh AE, Barr BC, Tell L, Bowman DD, Conrad PA, Ketcherside CC, Green T (1999)
Comparison of the internal transcribed spacer, ITS-1, from Sarcocystisfalcatula strains
and Sarcocystis neurona. J Parasitol 85:750—757

Marsh AE, Barr BC, Tell L, Koski M, Greiner E, Dame J, Conrad PA (1997a) In vitro
cultivation and experimental inoculation of Sarcocystisfalcatula and Sarcocystis neurona
merozoites into budgerigars (Melopsittacus undulatus). J Parasitol 85:1189—1192

Marsh AE, Barr BC, Lakritz J, Nordhausen R, Madigan J E, Conrad PA (1997b)
Experimental infection of nude mice as a model for Sarcocystis neurona-associated
encephalitis. Parasitol Res 83 :706—711

Marsh AE, Johnson PJ, Ramos-Vara J, Johnson GC (2001) Characterization of a
Sarcocystis neurona isolate from a Missouri horse with equine protozoal
myeloencephalitis. Vet Parasitol 95: 143—154

Mehlhom HW, Hartley WJ, Heydron A0 (1976) A comparative ultrastructural study of
the cyst wall of 13 Sarcocystis species. Parasitologica 12:451—467

National Animal Health Monitoring System (NAHMS Equine) (1998) Needs assessment
survey results. United States Department of Agriculture: Animal Plant Health Inspection
Service: Veterinary Services, Fort Collins, CO, #N207.597

Rosonke BJ, Brown SR, Tomquist SJ, Synder SP, Garner MM, Blythe LL (1999)
Encephalomyelitis associated with a Sarcocystis neurona-like organism in a sea otter. J
Am Vet Med Assoc 215:1839—1842

Rossano MG, Kaneene JB, Marteniuk JV, Banks BD, Schott HC 2nd, Mansfield LS
(2003) A herd-level analysis of risk factors for antibodies to Sarcocystis neurona in
Michigan equids. Prev Vet Med 5727—13

Slapeta JR, Modry D, Votypka J, J irku M, Lukes J, Koudela B (2003) Evolutionary

relationships among cyst-forming coccidia Sarcocystis spp. (Alveolata: Apicomplexa:

22

Coccidea) in endemic African tree vipers and perspective for evolution of heteroxenous
life cycle. Mol Phylogenet Evol 27:464—475

Spalding MG, Yowell CA, Lindsay DS, Greiner EC, Dame JB (2002) Sarcocystis
meningoencephalitis in a northern gannet (Moms bassanus). J Wildl Dis 38:432—437

Tanhauser SM, Yowell CA, Cutler TJ, Greiner EC, MacKay RJ, Dame JB (1999)
Multiple DNA markers differentiate Sarcocystis neurona and Sarcocystisfalcatula. J
Parasitol 85:221—228

Vos P, Hogers R, Bleeker M, Reijans M, van de Lee T, Homes M, Frijters A, Pot J,
Peleman J, Kupier M, Zabeau M (1995) AF LP: a new technique for DNA fingerprinting.
Nucleic Acids Res 23:4407—4414

Wenvon CM (1926) Protozoology. Bailliere, Tindall and Cox, London

23

CHAPTER 2

PURIFICATION OF SARC 0C YS T IS NE URONA MATERIALS

This chapter is based on the following two studies:

1. Purification of Sarcocystis neurona sporocysts from opossum (Didelphis
virginiana) using potassium bromide discontinuous density gradient

centrifugation.

2. Generally applicable methods to purify intracellular coccidia from cell cultures

and to quantify purification efficacy using quantitative PCR.

24

 

1. Purification of Sarcocystis neurona sporocysts from opossum (Didelphis
virginiana) using potassium bromide discontinuous density gradient

centrifugation

Published as original paper:

Hany M. Elsheikha 1, Alice J. Murphy 2, Scott D. Fitzgerald 2, Linda S. Mansfield 2’ 3,
Jeffrey P. Massey 4, and Mahdi A. Saeed 1 (2003) Purification of Sarcocystis neurona
sporocysts from opossum (Didelphis virginiana) using potassium bromide discontinuous

density gradient centrifirgation. Parasitology Research 90: 104—109.

1 Molecular Epidemiology Laboratory, Food Safety and Toxicology Center, Michigan
State University, MI 48824, East Lansing, USA.

2 Animal Health Diagnostic Laboratory, Michigan State University, MI 48824, East
Lansing, USA.

3 Department of Microbiology and Molecular Genetics, Michigan State University, MI
48824, East Lansing, USA.

4 Molecular Biology Section, Michigan Department of Community Health, MI 48909,

Lansing, USA.

25

ABSTRACT

This report describes a new, inexpensive procedure for the rapid and efficient
purification of Sarcocystis neurona sporocysts from Opossum small intestine. S. neurona
sporocysts were purified using a discontinuous potassium bromide density gradient. The
procedure provides a source of sporocyst wall and sporozoites required for reliable
biochemical characterization and for immunological studies directed at characterizing
antigens responsible for immunological responses by the host. The examined isolates
were identified as S. neurona using random amplified polymorphic DNA primers and
restriction endonuclease digestion assays. This method allows the collection of large
numbers of highly purified S. neurona sporocysts without loss of sporocyst viability as
indicated by propidium iodide permeability and cell culture infectivity assays. In

addition, this technique might also be used for sporocyst purification of other Sarcocystis

spp.

26

 

INTRODUCTION

Since the early 19905, the protozoan parasite Sarcocystis neurona has been
increasingly recognized as an important pathogen infecting the equine central nervous
system, causing equine protozoal myeloencephalitis (EPM) (Fenger et al., 1997; Dubey et
al., 2001; Mansfield et al., 2001; Marsh et al., 2001). Sarcocystis spp. are ubiquitous
parasites and have an obligatory two-host life cycle, with sexual reproduction and
sporocyst formation occurring in the definitive host and schizogony and sarcocyst
formation occurring in the intermediate host (Gardiner et al., 1988). Although, the life
cycle is not completely understood, natural and experimental transmission studies with S.
neurona indicate that the organism lacks host specificity and can be transmitted via the
fecal-oral route between opossums, the definitive host (F enger et al., 1995, 1997; Dubey
and Lindsay, 1998) and other mammalian intermediate hosts of several different species
such as cats, armadillos, skunks, and raccoons (Dubey et al., 2000; Cheadle et al., 2001c,
2001d; Hamir and Dubey, 2001). Horses are exposed to multiple Sarcocystis spp. in the
environment that may complicate the understanding of the life cycle of S. neurona, and
the pathogenesis and diagnosis of EPM. S. dasypi and S. diminuta sarcocysts have been
detected from the tongues of arrnadillos in North America (Lindsay et al., 1996). In
addition, at least three species of Sarcocystis (S. neurona, S. falcatula, S. speeri) have
been shown to use the North American opossum (Didelphis virginiana) as their definitive
host (Fenger et al., 1995; Tanhauser et al., 1999; Cheadle et al., 2001a). These species
have the sporocyst stage in the opossum small intestine and are difficult to differentiate

based on morphological characteristics alone.

27

 

Some reports describe sporocyst isolation from opossum small intestine, but none
provides pure and sufficient parasite material to carry out genetic, biochemical, and
immunological investigations (Murphy and Mansfield, 1999; Cheadle et al., 2001a;
Dubey et al., 2001; Ellison et al., 2001). Although sporocysts represent a valuable source
of stage specific parasite material, there are limitations concerning their use, including:
(1) purification is time-consuming and provides low yield, and (2) there are some
Sarcocystis species that produce low numbers of sporocysts and/or shed sporocysts
intermittently in the feces during infection. Also, infections with multiple Sarcocystis
species may occur. The infected small intestine contains several developmental stages of
the parasite. The ability to retrieve intracellular Sarcocystis oocysts/sporocysts stages has
been hampered by the lack of a method for separating the parasitized cells (situated
exclusively on the brush border layer of the mucosa) from the majority of non-parasitized
cells of the intestinal wall. Therefore, there is a need to develop alternatives to current
methods to allow not only research into the biology of the Sarcocystis life cycle, but also
the characterization of the antigens and the molecular constituents of the sporocyst stage
of the parasite.

The objectives of the present study were (1) to maximize the sporocyst yield to
assure the isolation of a quantitatively representative pure sporocyst population for
subsequent studies, (2) to genotype the investigated Sarcocystis spp. sporocysts, (3) to
use a dye permeability assay using propidium iodide (P1) to determine whether the
permeability of the sporocyst wall to P1 would be changed by exposure to a potassium
bromide (Kbr) gradient, and (4) to perform a cell culture infectivity assay to examine the

effects of Kbr purification on the sporozoites' viability.

28

 

MATERIALS AND METHODS

Parasite isolates. Two Sarcocystis species isolates used during this study were
obtained from naturally infected opossums (D. virginiana) that had been killed by
automobiles on the roadways in Michigan and that had been submitted for necropsy to
the Wildlife Services Agency of the Michigan Department of Natural Resources, and the
Animal Health Diagnostic Laboratory, Michigan State University.

Sample collection and sporocyst isolation. Fecal specimens were examined for
the presence of Sarcocystis sporocysts using saturated NaCl (360 g 1'], sp. gr. 1.21) fecal
floatation. It is helpful to start with fecal specimens that have large numbers of
sporocysts, however, some animals may harbor the parasite and not shed sporocysts.
Some opossums that have been tested negative for Sarcocystis spp. infection using the
fecal floatation test were found to be heavily infected with the parasite in the small
intestine by examining smears from mucosal scrapings (unpublished data). Therefore, the
opossums' intestines were initially screened for Sarcocystis spp. sporocysts by mucosal
scrapings, and samples found to be positive for the presence of sporocysts were processed
according to the method described by (Murphy and Mansfield, 1999) with some
modifications. A few segments of the small intestine from each animal were removed and
washed with 0.01 M phosphate-buffered saline, pH 7.4 (PBS). A scraping of mucosa was
observed at 200x magnification (Reichert microscope) to confirm the presence of oocysts
and/or sporocysts. Then, the remaining small intestine was flushed with sterile PBS to
remove the contents before being slit lengthwise. The intestinal content was collected
separately. The mucosa was scraped off using the edge of a glass slide to collect the

mucosa with the oocysts and/or sporocysts contained within the brush border. The

29

 

mucosal scrapings were washed three times with PBS by centrifugation at 500 g for

10 min. The pellet was resuspended in 3 vol. of pepsin-NaCl-HCI (0.65% pepsin w/v,
0.86% NaCl w/v, 1% conc. HCl v/v) and incubated at 37°C for 1.5 h with frequent
mixing. The slurry was washed 3 times with PBS as mentioned previously and the pellet
stored at 4°C in 3 vol. of Hank's balanced salt solution (HBSS) plus 100 U ml'1 of
penicillin, 100 ug ml'lof streptomycin, and 1.25 pg ml‘1 of fungizone until further use. A
1to 3 ml aliquot of the semi-digested mucosa was concentrated by centrifugation for

10 min at 500 g. The pellet was decontaminated by suspending in 10 ml of a 10% (v/v)

commercial bleach solution consisting of 1 m1 household bleach (5.25% sodium

 

 

hypochlorite) and 9 ml sterile deionized distilled water, and incubating on ice for 5 min
with frequent agitation. The bleach solution was removed by centrifugation (8 min at
3,000 g) at 4°C and the pellet was washed twice with sterile PBS (pH 7.2). The
supernatant was removed and the pellet resuspended in 3 ml of T ris-EDTA buffer (Tris
base 50 mM, EDTA 10 mM).

Fat aggregates, bacteria, and debris contaminate the sporocysts when gut slurry
with a high fat content is used. Since the yield and purity of the recovered sporocysts is
highly influenced by fat content, two washing procedures were used to decrease the fat
content and to improve the sporocyst yield and purity. In the case of specimens that
contain excess fat, the sporocyst suspension was vigorously mixed with an equal volume
of ethyl ether. Two layers formed; the upper (ethyl ether) layer contained the lipid
material, while the sporocysts remained in the aqueous phase. The upper layer was
removed and the sporocysts washed twice by centrifugation at 3,000 g for 10 min, 4°C in

Tris-EDTA buffer to remove residual ether. The final samples with low fat content were

30

repeatedly washed by centrifugation at the same conditions in Tris-EDTA buffer until a
clear supernatant was obtained. The final pellet obtained in both cases was resuspended
in a minimal volume of Tris-EDTA buffer before being applied to the gradient.
Discontinuous Kbr density gradient. A discontinuous gradient of Kbr (P9881,
Sigma, St. Louis, Mo., USA) was prepared. The gradient consisted of three solutions of
5, 15, and 25% (w/v) Kbr in Tris-EDTA buffer. From 7 to 8 ml each of 25, 15, and 5%
ice-cold Kbr solutions was carefully layered from bottom to top into 30 ml Nalgene
Centrifuge, Oak Ridge Style 3118 transparent polycarbonate tubes (Sigma). From 3 to
4 ml of each concentrated mucosal homogenate was carefully layered on the top of each
gradient. The gradients were centrifuged in a CR/CT4.12 centrifuge equipped with a M4
swinging bucket rotor at 3,000 g for 30 min at 4°C. Afier centrifugation, 3 clear white,
distinct band was observed at the interface between the 5% and 15% Kbr solutions
(Figure 1). This white band contained S. neurona sporocysts free of debris and bacteria.
The cleaned sporocysts (Figure 2B) were almost entirely free of debris (Figure 2A). The
band was gently aspirated at the interface using a sterile Pasteur pipette, and the
recovered sporocysts were diluted threefold with sterile 1>< PBS (pH 7.2) and centrifuged
at 3,000 g for 10 min at 4°C. The pellet was resuspended and washed twice using the
same buffer and centrifirgation conditions. The final recovered pellet was resuspended in
sterile l>< PBS. The yield of sporocysts was determined microscopically by counting the

number of sporocysts using a hemocytometer.

31

 

Figure 1. Centrifuge tubes showing a white band (arrows) of Sarcocystis neurona
sporocysts afier potassium bromide discontinuous gradient centrifugation. (A) At

3,000 g, 30 min, and 4°C. (B) At 16,000 g, 1 h, and 4°C. Images in this dissertation are
presented in color.

32

Finally, purified sporocysts were resuspended in a minimal volume of storage
media 6. g. HBSS containing an antibiotic-antimycotic mixture, as mentioned previously,
at 4°C. Each lot of purified sporocysts was tested by the dye permeability assay after
purification as described below. In some cases, higher centrifugal forces were needed to
enhance the purity of the final suspension. Here, the Kbr gradient was centrifuged under
conditions described for cesium chloride gradient purification of Cryptosporidium
parvum oocysts (Taghi-Kilani and Sekla, 1987), 16,000 g for 1 h at 4°C using a Sorvall
RC-SC superspeed, refrigerated centrifuge equipped with a Sorvall SM-24 rotor.

Morphometric characterization. Sporocyst type was elucidated using
morphological analysis (Cheadle et al., 2001a). An aliquot of 20 ul of purified sporocysts
from each isolate was quickly mixed with two warm drops of melted agar (2 mg
BACTO-agar powder (Difco)/ 100 ul 0.2 M Tris-HCl, pH 8.0). The sporocyst agar
suspension was placed on the center of a clean glass slide and covered with a clean glass
coverslip. The slide was then observed using light microscopy (Reichert) for
morphological characteristics. Then, immersion oil with 1000X magnification was
employed for the measurement of the sporocysts. A laser scanning microscope (LSM
Zeiss 210) with 488 nm laser line was used to take images (Figure 2). At least 25
sporocysts per isolate were measured using an ocular micrometer calibrated against its
stage micrometer slide after the agar had hardened. Minimum and maximum values are
given, followed in parentheses by the arithmetic mean and standard deviation. All
measurements were in micrometers.

Genotypic identification. The genotype of Sarcocystis species sporocysts was

identified according to the method described by Tanhauser et al., (1999). Genotyping of

33

 

the isolate was carried out using DNA extracted from purified sporocysts from opossums
using the Dneasy Tissue kit (Qiagen). Pure sporocysts were subjected to freezing in
liquid nitrogen followed by thawing for six cycles in order to break the oocyst and
sporocyst wall. Then, the sporozoite DNA was extracted according to the manufacturer's
instructions. Genotypic analysis of the DNA from Sarcocystis species isolates was
performed using PCR analysis with primers JNBZS/JD396 (Tanhauser et al., 1999). The
primers were used to amplify a 334 bp amplicon from the sporocyst DNA of the isolates.
The amplified 334 bp PCR products were subsequently digested with HindIII or Hinfl

restriction enzymes to differentiate bands characteristic of S. neurona and S. falcatula. S.

 

speeri does not amplify with these primers (S. Tanhauser, personal communication). The
digested DNA products were subjected to electrophoretic separation on a 2% agarose gel
with appropriate size markers, stained with ethidium bromide, visualized under UV light
and compared with an S. neurona (MIHl) isolate cultivated from horse neural tissue and
an S. falcatula (Cornell) isolate obtained from sporocysts shed from an opossum fed a
wild-caught grackle (Quiscalus sp.).

Dye permeability assay. The PI exclusion assay was used to indicate sporocyst
viability and potential infectivity. The dye permeability assay was used as described
previously (Campbell et al., 1992). This assay was initially used to assess the viability of
C. parvum oocysts. We have adapted this assay to evaluate the viability of Sarcocystis
spp. sporocysts. A stock solution of PI (Sigma, P4170) 1 mng'1 in 0.1 M PBS, pH 7.2
was added to aliquots of the purified sporocysts suspended in 1.5 ml microcentrifuge
tubes at a ratio of 1:10 (v/v). The contents of the tubes were mixed gently and the tubes

were incubated in the dark at 37°C for 2 to 3 h. Each sample was examined by both

34

epifluorescence and differential interference contrast (DIC) microscopy. At least 50
sporocysts in each sample were examined and scored for viability.

Cell culture infectivity assay. Maintenance, passaging, and growth of equine
dermal cells (ATCC, CCL-57) (American Type Culture Collection, Manassas, Va.) were
performed as described by Mansfield et al., (2001). For infectivity assays, cells were
grown on sterile 12-mm2 coverslips positioned at the bottom of 24-well tissue culture
plate in a 5% C02 atmosphere at 37°C to approximately 70-90% confluence (24-48 h).
Cultures were infected with sporozoites obtained from sporocysts before and after

purification as described by Murphy and Mansfield, (1999). Cultures were placed in a 5%

 

C02 incubator at 37°C. During in vitro cultivation, the medium was changed every i
3 days to maintain the pH within the range 7.2-7 .6. Both light inverted and

epifluorescence microscopy were used to assess plaques in the infected cell monolayers.

After incubation for 4 weeks, the medium was aspirated and the infected monolayers

were washed once with 1X PBS and coverslips were fixed in 10% phosphate-buffered

formalin for 30 min, placed in 100% methanol (room temperature) for 10 min, and

stained with Giemsa. Coverslips were attached to glass microscope slides with Perrnount

(Fisher Scientific, Fair Lawn, NJ, USA) and examined by light microscopy to observe

the sporozoite and schizont stages of the life cycle.

35

RESULTS

We developed a protocol for collecting the S. neurona parasites from the brush
border of the intestinal mucosa, thereby increasing the ratio of parasites to intestinal cells
and reducing other debris and contaminants. By centrifugation on a discontinuous Kbr
gradient, it was possible to recover and purify Sarcocystis sporocysts from the intestinal
mucosa. The use of commercial bleach for 5 min prior to gradient centrifugation greatly
facilitated the separation of oocysts and sporocysts from intestinal cells and fecal debris.
We routinely isolated 10-20><106 sporocysts in 3-4 h without the use of specialized or
expensive equipment. Oocysts were ovoid to subspherical, smooth walled, colorless, and
lacked oocyst residuum, polar granule or micropyle. The sporocysts were ovoid to round,
9.2-11.3X6.0—8.1 (10.3:t0.55><7.16:t0.56) with length to width ratios of 1.3-1.6. The

sporocyst wall was colorless, thin and smooth. All sporocysts examined possessed four

sporozoites and a distinct residuum (Figure 2C).

 

Figure 2. Photomicrographs of cleaned and crude Sarcocystis neurona sporocysts. Crude
sporocysts were collected and cleaned from intestinal mucosa of opossum. (A, B) Phase
contrast microscopy of crude and cleaned sporocysts. Scale bar = 25 pl. (C) Oblique
illumination of purified sporocysts. Note the residual body (white arrow), sporozoite
(black arrow), oocyst wall (black arrowhead), and sporocyst wall (white arrowhead).
Scale bar = 100 pl.

36

The total mean recovery of sporocysts after applying them to the gradient was
86.89ziz4.1% and the recovered sporocysts were greater than 98% pure (Table 1). We
compared the current method (mucosal scraping) for isolation of Sarcocystis spp.
sporocysts from opossum small intestine to the Kbr-gradient centrifugation method and
found that the purity and viability of the isolated sporocysts as well as the freedom of

bacterial contamination were improved (Table 1).

Table 1. Comparison between mucosal scraping (current method) and potassium
bromide discontinuous gradient centrifugation method for the isolation of Sarcocystis
spp. sporocysts from opossum small intestine.

 

 

 

 

 

 

 

. . 'Current method (mucosal {Potassium bromidediscontrnuousgradient
Criteria ; . i .
, scraping) gmethod
{Purity (%) F40 ,, -_ l>98 H -
Viability

- i - . 2t .
(%) 192 3 0 6 g98 68 107
{Bacteria lTNTC immutable

o The relative purity was assessed microscopically by counting the number of
sporocysts and all other particles or debris of any size/microscopic field (averaged
over ten randomly selected fields). This procedure was performed on each specimen
before and after purification. In both cases, the ratio of sporocysts versus debris was
compared to quantify purity percentage before and after Kbr-gradient purification.

o The viability of sporocysts was determined using the propidium iodide exclusion
assay.

0 TNTC = too numerous to count.

37

1
L.....___. 4M.L~-V_WL.-.;L_-L- _ .

RFLP analysis showed banding patterns characteristics of only S. neurona for
these two opossum isolates. They both had bands at 180, and 154 bp with Hinfl digestion,

and a single band at 334 bp with HindIII digestion (Figure 3).

MIOP MIHl Cornell
M PHer PHfHd PHfHd

 

Figure 3. Results of random amplified polymorphic patterns using the primers
JNB25/JD396 and restriction enzyme digestion with Hinfl and HindIII of sporocysts
DNA preparations from Sarcocystis neurona isolates from Michigan opossums (MIOP),
S. neurona (MIHl) isolate from horse, and S. falcatula (Cornell) isolate from opossrun.
M 100-bp ladder marker; P PCR product; Hf Hinfl digest; Hd Hindlll digest. Note, afier
HindIII digestion of PCR 334 bp amplicons, S. neurona amplicon is unaffected while S.
falcatula amplicon yields bands at 164, 108, and 62 bp. After Hinfl digestion, S. neurona
amplicon yields bands at 180 and 154 bp while S. falcatula amplicon yields bands at 170
and 164 bp.

The viability of the purified sporocysts was estimated using propidium iodide dye
exclusion assay and was found to be 98.68i1.07%. Successful cultivation of the two S.
neurona isolates was observed from sporozoites obtained from the crude and the clean
sporocysts, before and after gradient purification, respectively. The in vitro infectivity
assays showed that S. neurona sporozoites obtained from Kbr-purified sporocysts were
viable and maintained their infectivity afier purification. Generally, cultures infected with

sporozoites obtained from purified sporocysts appeared morphologically healthier than

those infected with the sporozoites obtained from crude sporocysts.

38

DISCUSSION

The goal of these experiments was to develop a reliable method for the
purification of large quantities of viable S. neurona or any other Sarcocystis spp.
sporocysts. The infected opossum gut contains populations of oocyst and sporocyst stages
of the parasite, but its use for obtaining purified material of these intracellular Sarcocystis
stages has been hampered by the lack of a method for separating the parasitized cells
(situated exclusively on the brush border layer of the mucosa) from the majority of the
non-parasitized cells of the intestinal wall. In the present study, the potassium bromide

discontinuous gradient was used to isolate pure S. neurona oocysts/sporocysts from the

 

intestinal mucosa of opossums (D. virginiana). Potassium bromide discontinuous i
gradient has been effectively used for the purification of C. parvum oocysts (Entrala et
aL,2000)

Under natural conditions, the oocyst wall is so thin that it often ruptures
mechanically. Free sporocysts, released into the intestinal lumen, are passed in feces
(Dubey et al., 1989). The recovery of many intact oocysts post-purification that still have
wall and enclose two sporocysts (Figure 2C) may result in increased viability of the
purified sporocysts due to the protection afford by the cell wall.

Based on the measured dimensions of the sporocysts, the Sarcocystis spp. isolates
under investigation appeared most closely related to the S. neurona strain (Cheadle et al.,
2001a). However, there were no unique internal morphological characteristics of the
isolates. The use of agar overlay helped to hold the sporocysts in position, which in turn

prevented deformity of the sporocysts and preserved their dimensional structure.

39

For in vivo and in vitro studies of S. neurona, it is important to use well-defined
isolates of S. neurona because at least five strains with structurally similar sporocysts
have been described from Opossum isolates. Molecular characterization of Sarcocystis
sporocysts collected from opossum has been shown to be practical compared to the time-
consuming and expensive use of mice to type Sarcocystis isolates (Cheadle et al., 2001b).
In the present study, PCR-RF LP analysis identified the two opossum isolates as S.
neurona based on characteristics of their banding patterns. However, some opossums
have more than one Sarcocystis spp. This purification method should provide the basis
for future experiments to develop clonal isolates from individual oocysts or sporocysts.

This method improved the viability of the purified samples. PI is a fluorogenic
vital dye that indicates the presence of live organism. PI exclusion from the nucleus is
based on membrane integrity, and organisms are considered dead (non-viable) when PI is
included in the nucleus (Griffiths et al., 1994). Kbr is a hyperosmotic compound, and the
recovery of sporocysts is affected by sporocyst viability, since only water impermeable
sporocysts can be recovered. The viability of the Kbr-purified S. neurona sporocysts was
remarkably high. This enrichment concentration of viable sporocysts has been previously
reported in C. parvum oocysts, when hyperosmotic sucrose density or zinc sulfate
flotation techniques were used for oocyst purification (Bukhari and Smith, 1995).

Infectivity is an important issue that is currently addressed mainly by using
animal models. Cell culture seems to be a more cost-effective, reproducible and practical
approach. S. neurona can be cultivated in equine dermal cells for 6 months with no
apparent loss of viability. S. falcatula grow briefly and die in 4-6 weeks (L.S. Mansfield

and A]. Murphy, unpublished data). The sporozoites obtained from purified sporocysts

40

of the two parasite isolates continued to grow in equine dermal cells for at least 2 months,
which supports the notion that the genotype of the parasite isolates is S. neurona.

This technique is easy to perform, convenient, and allows the rapid purification of
S. neurona sporocysts which will in turn enhance research into the biology of host cell-
parasite interactions, and enable amplification of parasite materials for further
immunological, biochemical, and molecular studies. Axenic sporocysts, suitable for in
vitro culture, can be obtained using this method. Therefore, it could also be used for the
assessment of the viability of S. neurona sporocysts isolated fi'om opossums. This
technique is a general method that can be applied for the purification of sporocysts of any

other Sarcocystis species from opossum small intestine.

41

REFERENCES

Bukhari Z, Smith HV (1995) Effects of three concentration techniques on viability of
Cryptosporidium parvum oocysts recovered from bovine feces. J Clin Microbiol
33:2592—2595

Campbell AT, Robertson LJ, Smith HV (1992) Viability of Cryptosporidium parvum
oocysts: correlation of in vitro excystation with inclusion or exclusion of fluorogenic vital
dyes. Appl Environ Microbiol 58:3488—3493

Cheadle MA, Dame JB, Greiner EC (2001a) Sporocyst size of isolates of Sarcocystis
shed by the Virginia opossum (Didelphis virginiana). Vet Parasitol 95:305—311

Cheadle MA, Tanhauser SM, Scase TJ, Dame JB, MacKay RJ, Ginn PE, Greiner EC
(2001b) Viability of Sarcocystis neurona sporocysts and dose titration in gamma-
interferon knockout mice. Vet Parasitol 95:223—231

Cheadle MA, Tanhauser SM, Dame J B, Sellon DC, Hines M, Ginn PE, MacKay RJ,
Greiner EC (2001c) The nine-banded armadillo (Dasypus novemcinctus) is an
intermediate host for Sarcocystis neurona. Int J Parasitol 31 :330—335

Cheadle MA, Yowell CA, Sellon DC, Hines M, Ginn PE, Marsh AE, Dame JB, Greiner
EC (2001d) The striped skunk (Mephitis mephitis) is an intermediate host for Sarcocystis
neurona. Int J Parasitol 31 :843—849

Dubey JP, Lindsay DS (1998) Isolation in immunodefecient mice of Sarcocystis neurona
fiom opossum (Didelphis virginana) faeces, and its differentiation from Sarcocystis
falcatula. Int J Parasitol 28:1823—1828

Dubey JP, Speer CA, F ayer R (1989) General biology: In: Sarcocystosis of animal and
man. CRC Press, Boca Raton, pp 2—13

Dubey JP, Saville WJ A, Lindsay DS, Stich RW, Stanek J F, Speer CA, Rosenthal BM,
Njoku CJ, Kwok OCH, Shen SK, Reed SM (2000) Completion of the life cycle of
Sarcocystis neurona. J Parasitol 86:1276—1280

Dubey JP, Mattson DE, Speer CA, Hamir AN, Lindsay DS, Rosenthal BM, Kwok OCH,
Baker RJ, Mulrooney DM, Tomquist SJ, Gerros TC (2001) Characteristics of a recent
isolate of Sarcocystis neurona (SN7) from a horse and loss of pathogenicity of isolates
SN6 and SN7 by passages in cell culture. Vet Parasitol 95:155-166

Ellison SP, Greiner E, Dame JB (2001) In vitro culture and synchronous release of
Sarcocystis neurona merozoites from host cells. Vet Parasitol 95:251—261

Entrala E, Molina-Molina J, Rosales-Lombardo M, Sanchez-Moreno M, Mascaro-
Lazcano C (2000) Cryptosporidium parvum: oocysts purification using potassium

42

bromide discontinuous gradient. Vet Parasitol 92:223—226

F enger CK, Granstrom DE, Langemeier J L, Stamper S, Donahue J M, Patterson J S,
Gajadher AA, Marteniuk JV, Xiaomin Z, Dubey JP (1995) Identification of opossums
(Didelphis virginiana) as the putative definitive host of Sarcocystis neurona. J Parasitol
81 :916—919

Fenger CK, Granstrom DE, Gajadher AA, Williams NM, McCrillis SA, Stamper S,
Langemeier J L, Dubey JP (1997) Experimental induction of equine protozoal
myeloencephalitis in horses using Sarcocystis sp. sporocysts from the opossum
(Didelphis virginiana). Vet Parasitol 68:199—213

Gardiner CH, Fayer R, Dubey JP (1988) An atlas of protozoan parasites in animal tissues.
Agriculture Handbook, vol 651. USDA, Agricultural Research Service, pp 40—45

Griffiths JK, Moore R, Dooley S, Keusch GT, Tzipori S (1994) Cryptosporidium parvum
infection of Caco-2 cell monolayers induces an apical monolayer defect, selectively

increases transmonolayer permeability, and causes epithelial cell death. Infect Immun
62:45 06—45 14

Hamir AN, Dubey JP (2001) Myocarditis and encephalitis associated with Sarcocystis
neurona infection in raccoons (Procyon lotor). Vet Parasitol 95:335—340

Lindsay DS, McKown R, Upton SJ, McAllister CT, Toivio-Kinnucan MA, Veatch J K,
Blagbum BL (1996) Prevalence and identity of Sarcocystis infections in armadillos
(Dasypus novemcinctus). J Parasitol 82:518—520

Mansfield LS, Schott HC, Murphy AJ, Rossano MG, Tanhauser SM, Patterson J S,
Nelson K, Ewart SL, Marteniuk JV, Bowman D, Kaneene JB (2001) Comparison of
Sarcocystis neurona isolates derived from horse neural tissue. Vet Parasitol 952167—178

Marsh AE, Johnson PJ, Ramos-Vara J, Johnson GC (2001) Characterization of a
Sarcocystis neurona isolate from a Missouri horse with equine protozoal
myeloencephalitis. Vet Parasitol 952143—154

Murphy AJ, Mansfield LS (1999) Simplified technique for isolation, excystation, and
culture of Sarcocystis species from opossums. J Parasitol 85:979—981

Taghi-Kilani R, Sekla L (1987) Purification of Cryptosporidium oocysts and sporozoites
by cesium chloride and Percoll gradients. Am J Trop Med Hyg 36:505—508

Tanhauser SM, Yowell CA, Cutler TJ, Greiner EC, MacKay RJ, Dame JB (1999)

Multiple DNA markers differentiate Sarcocystis neurona and Sarcocystisfalcatula. J
Parasitol 85:221—228

43

2. Generally applicable methods to purify intracellular coccidia from cell cultures

and to quantify purification efficacy using quantitative PCR

Submitted for publication as original paper:

HM. Elsheikha], EM. Rosentha12,A.J. Murphyj, D.B. Dunamsz, D. A. Neelis’, and LS.
Mansfield“4 (2004) Generally applicable methods to purify intracellular coccidia from
cell cultures and to quantify purification efficacy using quantitative PCR. Journal of

Microbiological Methods.

1 Department of Large Animal Clinical Sciences, College of Veterinary Medicine,
Michigan State University, East Lansing, MI 48824, USA.

2 Animal Parasite Diseases Laboratory, Agricultural Research Service, United States
Department of Agriculture, Beltsville Agricultural Research Center, Animal and Natural
Resources Institute, Building 1080, BARC-East, Beltsville, MD 20705, USA.

3 Diagnostic Center for Population and Animal Health, College of Veterinary Medicine,
Michigan State University, East Lansing, MI 48824, USA.

4 Department of Microbiology and Molecular Genetics, College of Veterinary Medicine,

Michigan State University, East Lansing, MI 48824, USA.

44

ABSTRACT

The objective of this study was to evaluate the utility of a simple, efficient, and
rapid method for the isolation of Sarcocystis neurona merozoites and Besnoitia darlingi
tachyzoites from cultured cells. The efficacy of this purification method was assessed by
microscopy, SDS-PAGE, Western Blotting, immune-fluorescence, and three novel
quantitative PCR assays. Culture medium containing host cell debris and parasites was
eluted through PD-lO desalting columns. This purification method was compared to
alternatives employing filtration through a cellulose filter pad or filter paper. The
estimated recovery of S. neurona merozoites purified by the coltunn method was 82% (i:
3.7) of the original merozoites with 97.5% purity. In contrast, estimated recovery of S.
neurona merozoites purified by filter pad and filter paper was 40% and 30% with 76%
and 83% purity, respectively. The same procedures were applied to purify B. darlingi
tachyzoites from cultured cells. Of the original cultured B. darlingi tachyzoites, 94% (i
2.5) were recovered from the PD-lO column with 96.5%, purity whereas percentage
recovery of B. darlingi tachyzoites purified by filter pad and filter paper were 51% and
35% with 84% and 88% purity, respectively. All described methods maintained sterility
so that purified parasites could be subsequently cultured in vitro. However, purification
using a PD-lO column minimized parasite loss and the loss of viability as determined by
the trypan blue dye exclusion assay, the rate of parasite production, and plaque forming
efficiency in tissue culture. Moreover, column-purified parasites improved the sensitivity
of an immune-fluorescent (IFA) analysis and real-time quantitative PCR assays targeted
to parasite 18S ribosomal DNA and hsp70 genes. This technique appears generally

applicable for purifying coccidia grown in tissue cultures.

45

INTRODUCTION

Sarcocystis neurona is the primary causative agent of equine protozoal
myeloencephalitis (EPM) in the Americas (Dubey et al., 2001b). S. neurona was first
described in 1991 after its isolation from a horse (Dubey et al., 1991). It belongs to the
phylum Apicomplexa, a large group of mainly intracellular parasites. Besnoitia darlingi
is an Apicomplexan parasite employing opossum (Didelphis virginiana) as an
intermediate host and cat as a definitive host (Dubey et al., 2002). Equine dermal cells
support the development of S. neurona in cell culture (Mansfield et al., 2001). In
addition, bovine monocytes, equine kidney cells, cardiopulmonary endothelial cells, deer
testes, vero cells, Afiican green monkey (Cercopithecus aethiops) kidney cells (CV-1),
rat myoblasts, and bovine turbinate (BT) cells have been employed successfully to
support S. neurona replication (Dubey et al., 2001b). Bovine monocytes, CV-l cells, and
BT cells also have been used to cultivate and maintain B. darlingi tachyzoites (Dubey et
al., 2002; Elsheikha et al., 2004).

One of the major obstacles to research with S. neurona and other coccidia is the
difficulty of obtaining abundant parasites free of host cell contaminants. Such pure
parasite suspensions would facilitate studies of host-parasite interaction and are needed in
order to extract protein and genetic materials for biochemical, immunological, and
molecular analyses. Difficulty in obtaining sufficient numbers of S. neurona merozoites
free of contaminating host cells or cell debris is generally recognized. Approaches have
been reported for separating the parasite from host cells. Dubey et al., (2001a) reported
the isolation of merozoites from cell culture by filtration through a syringe filter.

However, the syringe filters tend to clog easily. Lindsay and Dubey, (2001) passed the

46

merozoite suspension through a 27-gauge needle attached to a 10 ml syringe in order to
rupture host cells, and then passed the suspension through a sterile 3pm filter to remove
cellular debris. However, the parasite yield is low using this method. In another study, a
calcium ionophore (A23187) was used for synchronous release of parasites from host
cells (Dame et al., 2001). This method was slow, complicated for large-scale preparation
and resulted in preparations that contained appreciable host cell contamination. Although
useful, none of these approaches are well suited for producing an abundance of pure
merozoites desirable for immunological, biochemical, and genetic analyses.

Quantitative real-time PCR provides a sensitive, accurate and fast method for
estimating the abundance of DNA templates by monitoring the accumulation of
fluorescent PCR products. In assays that approach 100% efficiency, in which a doubling
of product occurs with each successive amplification cycle during the exponential phase,
the number of cycles required to significantly exceed background fluorescence (Ct) is
inversely proportional to the initial template concentration (Klein et al., 2002; Ginzinger,
2002)

Herein we describe the adaptation of a previously described approach using a PD-
10 desalting column (Hemphill et al., 1996) to isolate viable over 80% of the S. neurona
merozoites and over 90% of the B. darlingi tachyzoites cultivated in bovine turbinate
cells at a purity of over 95%, and describe new real-time quantitative PCR assays to
estimate the relative and absolute concentrations of parasite and host DNA in such
samples. The described method compares favorably to known alternatives in parasite
recovery, purity, and viability. Additionally, the procedure improves the sensitivity of

immune—fluorescent staining and of diagnostic PCR amplifications.

47

MATERIALS AND METHODS

Parasite strains and tissue cultures. Sarcocystis neurona strains MIHl and SN3
were maintained in bovine turbinate (BT) cell monolayers. BT cells were originally
purchased from American Type Culture Collection (ATCC, CRL-1390, Manassas, VA,
USA). BT cells were used between passages 15-17 and maintained in 8-10 m1 of
complete culture medium using 25-cm2 plastic tissue culture flasks. Cells were fed
Eagle's minimum essential medium (EMEM) with Earle's salt supplemented with L-
glutamine, 10,000 U ml‘l penicillin G sodium, 10,000 pg ml'1 dihydrostreptomycin, 250
pg ml'1 of amphotericin B, nonessential amino acids, 100 mM sodium pyruvate, and 5 to
10% heat inactivated fetal bovine serum (F BS) (Gibco-Invitrogen, Grand Island, NY,
USA). Cells were incubated at 37 °C in 5 % COz/95% air. The B. darlingi MIBDl strain
was initially obtained from cysts isolated from a naturally infected opossum (D.
virginiana) from Michigan in 2002. Methods used for isolation and maintenance of this
strain were previously described (Elsheikha et al., 2004). Stock cell cultures were also
maintained by lifting the monolayers with trypsin-EDTA (0.25% trypsin; 1 mM
EDTA.4Na, Gibco) once a week and transfening the cells to new culture flasks.
Approximately 5X 1040f S. neurona_merozoites and 3 x 104 B. darlingi tachyzoites were
used to inoculate new flasks of BT cells. The development of mature schizonts occurred
in 3 to 5 days in both parasites after inoculation of the flasks.

Purification of S. neurona merozoites. Sarcocystis neurona merozoites were
harvested from their feeder cell cultures when about 60—80% of the BT host cells were
lysed. Free merozoites were removed from the tissue culture flasks by collecting the

medium supernatant. A large portion of the remaining parasites were released from their

48

host cells by gently shaking the side of the flasks 2 to 3 times. A few milliliters of
EMEM was added and the merozoites were collected. The preparation containing
merozoites and host cell debris was mixed thoroughly, divided into three equal aliquots,
and submitted to purification procedures by PD-lO column, cellulose filter pad, or filter
paper. All the purification procedures were performed inside a biological safety cabinet
under sterile conditions.

Purification using a PD-10 column. The merozoite suspension was washed
twice in a buffer solution by centrifugation at 1,500g for 5 min at 4°C. Buffer solutions
used were either sterile 1x PBS or 1x Hanks’ balanced salt solution (HBSS). The final
pellet was resuspended in 2.5 ml of buffer solution and applied to a PD—10 column filled
with sephadex G-25M (Amersham Biosciences, Piscataway, NJ, USA), previously
equilibrated with approximately 25 ml of the same buffer. The parasites were eluted with
3.5 ml of the same buffer used in equilibration of the column. The eluted, purified
merozoites were used for determination of recovery, purity, and viability as well as for
evaluating the utility of parasite materials purified by this method for molecular analyses.

Purification using cellulose filter pad and filter paper. The cellulose filter pad
method for purifying the merozoite suspension required a glass filter assembly
GVIillipore, Bedford, Massachusetts, USA) (Figure l). A 47-min glass holder apparatus
was designed to handle large volumes of liquids. An autoclavable cellulose filter pad
(cat. no. AP1004700, Millipore) was initially rinsed with sterile 1x PBS, pH 7.2 or
EMEM immediately prior to use. Then, the merozoite suspension was passed through
the filter pad and the filtrate was collected. The filtrate was centrifuged at 1,500g for 5

min at 4°C. The pellet was resuspended in 3.5 ml of 1x PBS. The filter paper method

49

was performed identically using the same conditions, but by using Whatrnan filter paper
(cat. no. 1001125, VWR Intemational,West Chester, PA, USA) instead of a filter pad.
The purified merozoites in both cases were used for determination of recovery, purity,

and viability.

 

Figure 1. Illustration of the filter assembly. It consisted of a removable 250-ml
borosilicate glass funnel (l) and support base (2). An anodized aluminum spring clamp
(3) sandwiched a cellulose filter pad or filter paper (4) between funnel and support base.
Connection to 500-ml Erlenmeyer filter flask (5) is made with a silicone stopper (6).
Parasite suspension with debris (7) and relatively pure parasite suspension (8).

50

Determination of merozoites’ yield and purity. In all experiments, the yield of
merozoites was determined microscopically by counting individual intact merozoites
before and after purification using the four outside squares and one central square of a
standard hemocytometer chamber (Sigma, Saint Louis, MO, USA) at 200X magnification
on a light microscope (Reichert-Jung, Pegasus Scientific, Frederick, MD, USA). The total
number of merozoites/m1 was estimated by averaging the counted squares and corrected
for dilution and volume using the formula (Total merozoites counted in 5 squares/5) x 104
= merozoites/ml. Tubes containing the purified merozoites in 3.5-ml aliquots were
vortexed for 10 seconds before counting. The purity of the final merozoite preparation
was evaluated microscopically by differential counting of intact motile merozoites versus
any other recognizable particles or cellular debris. Percentage purity was quantified by
comparing the number of merozoites to the number of debris particles before and after
purification.

Assessment of merozoite viability. The viability of the purified merozoites was
assessed by two means (1) trypan blue dye exclusion assay and (2) by evaluating the rate
of S. neurona merozoite production and plaque forming efficiency in BT cell culture.
Equal numbers (~ 12 x 105) of crude merozoites and merozoites purified by PD-lO
column, filter pad, or filter paper were inoculated onto T-25 flasks containing a
monolayer of BT cells in complete EMEM. The cultures were incubated at 37°C in 5%
C02. T25 tissue culture flasks were mock-inoculated with complete EMEM and served
as negative controls. Media was changed 36 h post-inoculation to remove extracellular
merozoites and then at 3 to 4 day intervals thereafter. The inoculated cultures were

monitored daily using a light inverted microscope (Carl Zeiss, Opton, Columbia, MD,

51

USA) to follow the development of plaques and production of merozoites. The numbers
of S. neurona plaques (infected foci) were estimated 10 days post infection (DPI) using
an inverted microscope, by counting the number of visible plaques per microscopic field
at X200 magnification. These counts were expressed as a mean (i SD) over 10 randomly
selected fields. After 13-day culture period, merozoites were harvested from all flasks by
rinsing the monolayer twice with medium and counting the number of merozoites using a
hemocytometer as described.

SDS-polyacrylamide gel electrophoresis (PAGE) and immunoblotting.
Proteins were extracted from an equal number of purified and un-purified merozoites by
suspending them in a lysing buffer containing 0.5 M Tris (pH 7.4) with 10% sodium
dodecyl sulfate (SDS), 20% glycerol and 5% E-mercaptoethanol and heating the mixture
at 95°C for 5 min. Solubilized proteins extracted from approximately 107 pure
merozoites were loaded onto a PAGE gel in 2, 5, and 10pl aliquots. For comparison, a
15 p1 of solubilized protein extracted from approximately 107 crude merozoites was also
loaded onto a lane of the same gel. The gel and a biotinylated molecular weight standard
(Bio-Rad, Hercules, CA) were separated electrophoretically on a 12—18% SDS—PAGE
gradient minigel at 150 V and transferred to Westran PVDF membrane (Bio-Rad). The
remainder of the immunoblotting procedure was performed as described previously
(Mansfield et al., 2001). Also, the protein extracts were visualized by silver staining as
follow; after SDS-PAGE, gels with the separated proteins were silver stained by fixation
for 30 min in 50% methanol-10% acetic acid followed by staining using the Silver Stain

Plus kit (Bio-Rad), and scanned using a Microtek ScanMaker III (Microtek Lab). Protein

52

bands from each lane were compared to size standards and patterns of each lane were
compared.

Immunofluorescence assay (IFA). To prepare IF A slides, serial dilutions of the
pure and un-pure merozoite stocks were placed in duplicate in wells of a 12 well slides
(20pl/well), air dried and fixed in acetone. Cerebrospinal fluid (CSF) from a horse
positive for S. neurona antibodies was placed on each well and incubated for 30 min at
37°C. The slide was washed several times with PBS and then incubated with Fluorescein
isothiocyanate- (FITC) - labeled goat anti-horse IgG- h+1 secondary antibodies in PBS
and 0.1% Evans blue and incubated at 37°C for an additional 30 min. The slide was
washed with PBS, cover-slipped and the optimum dilution of merozoites for IF A slides
determined. The stock merozoite solution was diluted with PBS to the optimum and
multiple slides prepared through the fixation step and stored frozen at -20°C. Positive
sera were applied at 1:2500 and 1:10,000 dilutions in PBS and incubated for 30 min at
37°C followed by washing as above. Serum from a knockout mouse previously infected
with S. neurona_was used as a positive control. FITC-labeled goat-anti-mouse IgG-h+1
secondary antibody (Kirkegaard & Perry Labs, Inc.) diluted 1:10 in PBS and 0.1% Evans
blue dye was applied. Stained merozoites were observed by epifluorescent microscopy
using the SPOT RT slider “F” mount camera (model No. 2.3.1, Diagnostic Instruments
Inc) and SPOT RT software V3.3.

Diagnostic sensitivity of conventional PCR for S. neurona. Crude merozoites
and merozoites purified using a PD-lO column were diluted and adjusted to equal
concentations of 1000 merozoites/200 pl 1x PBS. Ten-fold serial dilutions (1000, 100,

and 10 merozoites) from both samples were prepared to a final volume of 200p1 in 1x

53

PBS for DNA extraction using the QIAamp® DNA Blood Mini Kit (Qiagen). The
Qiagen® Blood and Body Fluid Spin Protocol was followed according to the
manufacturer’s recommendations to obtain a high DNA concentration from each dilution.
DNA concentration was assessed using a NanoDrop DN-1000 spectrophotometer and
NanoDrop software 2.5.4. (N anoDrop Technologies, Rockland, DE). The DNA extracts
were subjected to PCR using primers JNB25 and JD396 (Tanhauser et al., 1999). PCR
products were electrOphoresed through a 1.8% agarose gel for detection of a 334 bp band,
indicative of S. neurona.

Purification of B. darlingi tachyzoites. Besnoitia darlingi tachyzoites were
purified and analyzed from cultured BT cells using procedures identical to those
described for S. neurona merozoites. For each purification method used, recovery,
purity, and viability of B. darlingi tachyzoites were measured using methods identical to
those used for S. neurona merozoites.

Quantitative PCR assay design. We developed a new quantitative PCR assay
employing specific amplification of a small portion of nuclear 188 rDNA using primers
complementary to sequences conserved among coccidia but distinct from other taxa.
Additional selection criteria included minimal non-specific annealing, low GC content,
balanced primer length, and an amplification product which required fewer than 20
cycles in order to exceed threshold fluorescence (Ct) and dissociated over a narrow and
reproducible temperature range. After evaluating several candidate primer combinations,
we chose one pair that best met these criteria. Similar criteria were used in developing
quantitative assays for Besnoitia darlingi based on the Heat Shock Protein 70 gene

(hsp70) and for vertebrate 18S rDNA (Table 1).

54

Table 1. qPCR primers.

 

 

 

 

 

Target bp Forward Primer Reverse Primer

Coccidian 18$ rDNA 230 188 9F—GTTGGITTCTAGGACTGA 18$ lOR-AGCATGACGTTI‘TCCTATCTCTA
Vertebrate 185 rDNA 240 V1 BSF-CTCTTTCGAGGCCCTGTA V18SR-GGGACACTCAGCTAAGAGCA

3. darlingi 1151070 212 HSP7OF-GTGAGGCCGCAAARAAYGA lISP70R-ATGGCGGAGACYTCYTCGG

 

 

 

 

Brilliant® SYBR® Green QPCR Master Mix (Stratagene) was used to quantify
the accumulation of double-stranded DNA products in a Stratagene Mx300P Real Time
PCR System. Melting curves were performed to assure that the fluorescence was derived
from dye intercolating into a specific, homogeneous amplification product. Individual
50p1 reactions contained 25 p1 of SYBR® Master Mix, 19pl of ddH20, lpl of primer
mixture comprising lOOnM of each primer, and 5p] of template. Subsequent reactions
were performed using identical reagent concentrations in a total volume of 25 pl.

To establish that each assay for parasite DNA was quantitative, serial 5-fold
dilutions of the purified parasite extracts were amplified. Validation of the assay
targeting vertebrate rDNA made use of serial dilutions of crude preparations. Although
the absolute concentration of parasite DNA in our reference sample was initially
unknown, we deemed quantitative these assays whose linear regression of Ct against
relative concentration had an R2 value of >90 and an estimated efficiency between 90
and 110%. Assay reproducibility was determined through replication among extractions
and among triplicate qPCR assays of individual extracts.

A 10 min pre-incubation at 95°C was followed by 40 cycles of 30 s at 95°C, 1
min at 55°C, and 1 min at 72°C. The subsequent dissociation step included 1 min

incubation at 95°C, ramping down to 55°C at a rate of 0.2°C/sec, followed by 81 cycles

55

 

of incubation where the temperature was increased to 95° by 0.5°C/cycle every 305.
Fluorescence was quantified at the end of the annealing, extension, and dissociation steps
in each cycle. Each reaction was performed in triplicate to ensure reproducibility. Three
wells were used as negative controls; these lacked template but contained all other
reagents. The cycle at which fluorescence significantly exceed baseline levels, Ct, was
determined using the normalized fluorescence option of the instrument software.

To assess the affect of purification on the effective concentration of parasite DNA
targets using qPCR, we first assayed extracts containing 1000 merozoites /200pl, as
estimated by ocular haemocytometry. By normalizing a second set of pre- and post-
purification templates to a lng/ pl, as estimated by UV absorbance spectroscopy, we then
assessed how purification affected the relative proportion of total DNA attributable to
parasite and host sources. These estimates, when related to the total DNA concentration
estimated from UV absorbance spectroscopy, afforded the means to infer absolute
concentrations of host and parasite DNA concentrations before and after column
purification. Finally, triplicate assays were performed on mixtures of B. darlingi cultures
whose purified content comprised 25%, 50% and 75% of the total volume.

Statistical analysis. For determining recovery, parasite production, and plaque
forming efficiency, analysis of variance (AN OVA) procedure was performed using SAS
PC version 8 (SAS Institute, Cary, NC, USA). Statistical significance was assigned to P-
values <0.05. Charts were made using the graphical software Si gmaPlot version 8

(RockWare Inc., Golden, CO, USA).

56

RESULTS
We purified S. neurona merozoites and B. darlingi tachyzoites by eluting parasite
suspensions through PD-lO columns. Such purification was easy to perform and

successfully removed visible host cell contaminants from culture media (Figure 2).

 

Figure 2. Photomicrographs of S. neurona merozoites. (A) Crude merozoite preparations
before purification. Scale bar = 50 pl. (B) Merozoites (m) completely free of host cell
(hc) contaminants and extraneous debris alter purification using PD-10 column. Scale bar
= 25 pl.

By counting both merozoites/tachyzoites and debris particles, we estimated the
mean purity of S. neurona merozoites and B. darlingi tachyzoites at 97.5% and 96.5%,
respectively. We also purified the S. neurona_merozoite and B. darlingi tachyzoite
suspensions using cellulose filter pads and filter paper. The percentage purity was
estimated at 76% and 84% for S. neurona merozoite and 83%, 88% for B. darlingi
tachyzoites, respectively. Even though, the amount of host cell contamination observed
in the filter paper-purified parasite fraction was relatively small, purification by filter
pads and filter papers reduced the yield to 40% and 30% for S. neurona merozoites and

51% and 35%, for B. darlingi tachyzoites. In contrast, ~ 82% of merozoites and 94% of

57

 

”VJ

tachyzoites of the original suspensions applied to the columns were recovered (Figure 3).
Occasionally, merozoite yields were lower when the filter pads or filter papers were
washed with PBS instead of EMEM, perhaps because the serum in EMEM retarded or
prevented merozoite/tachyzoite attachment to the filter pads or filter papers while
washing. We investigated the effect of each purification method on the viability and rate
of in vitro development of parasites and by trypan blue dye exclusion assays and by
evaluating the rate of parasite production and plaque forming efficiency. Using the
trypan blue dye exclusion assay, the estimated viability of S. neurona merozoites and B.
darlingi tachyzoites before filtration was not less than 99%; only very minor reductions
in viability were observed in column-purified S. neurona (97.6% i 1.1) and column-
purified B. darlingi tachyzoites (97.2% i 0.8). In all methods, the purified organisms
were found to have variable abilities to penetrate BT cells and to reproduce. However,
the parasite production of column-purified parasites was significantly better than in
parasites purified by filter pad or filter paper (13.2 vs 7 and 5.5; P=0.003 and 0.001,
respectively, df=39 in case of S. neurona and 11.4 vs 5 and 6.2; P=0.007 and 0.008,
respectively, df=39 in case of B. darlingi). Additionally, the plaque forming ability of
column-purified parasites was significantly greater than that for parasites purified by
filter pad or filter paper (38 vs 22.7 and 15.6, P= 0.01 and 0.009; df=39 for S. neurona
and 54 vs 19 and 25; P: 0.002 and 0.006, respectively, df=39 for B. darlingi). On the
contrary, no significant difference was observed between the crude parasite and column-

purified parasites (P < 0.86).

58

 

- Recovery “/-

m Purity %

- Merozoite #

0
.00

 

 

 

..... - Plaquestt

m.dddddddddlddddddd0000000000

                                 

a Q Q

. .....O...°.O...O...C-
, .9 O O O b D D O D O tObOOOOOOOOObODOOOOODODQ

 
 

’9’
v
o
v
o
t
o
0' ...n.
v
o
v
.9
o

   

 
 
 

”warn. - .

  

       

 

   

11111111111

00000000000

   

          
 
 

 

 

120

100

0 0 0

8 6 4
:5 “3 :32

20

Column Pad Paper

Purification methods

Control

 

 

 

 

 

 

120

% a...
e
U % # .m
new“
w .H a. r
c u I... c
R P P M
_ m - — uuuuuuuuuu"uuuunumnuuuuununu"nun"ununuuuuuuunuuuuuuuuuu
EwwmmwwmumwmEmmawwmwfiflafi
m .u”aflawHanwfifiwfifiwwfi“Hummmummmmmmmflfl
III...-II..-III-IIIIIIIIIIIIIIIIII_
uuuuunuuuaunuuumunnuuuuunuuuuuuuuu_
0 0 0 0 0 0
0 8 6 4 2
1

Em i 53:

 

          
       

r
e
p
a
P;
.m
.m
.0
am...
Pm
m
Del.
s
u
CU
.IP
0
r.
t
n
o
C

filter papers compared to unpure (crude) merozoites. Number of parasites is expressed as
. All experiments were conducted in triplicate.

Figure 3. Purification of Sarcocystis neurona merozoites (A) and Besnoitia darlingi
x 106

tachyzoites (B) cultured in Bovine turbinate cells using PD-lO columns, filter pads, and

59

The efficacy of S. neurona merozoite purification using a PD-lO column was also
evaluated by SDS-PAGE and western immunoblotting and silver staining assays. Similar
but more intense electrophoretic profiles were obtained from proteins extracted from
column-purified merozoites as compared to crude S. neurona merozoite preparations
(Figure 4). Western blot analyses revealed ~30 and16 KDa antigens in all tested
preparations. The immuno-fluorescent staining (IFA) revealed a marked gain in

sensitivity and reduction in background in purified parasite preparations.

w-hOCD

N

 

Figure 4. Silver-stained SDS-PAGE gel of whole-cell protein preparations of
Sarcocystis neurona merozoites showing differences in the protein profiles of crude S.
neurona merozoites cultured in BT cells and PD-lO column-purified merozoites. M,
molecular weight markers (Broad Range Standards; Bio-Rad); Lanes 1 to 3 contain the
following amounts of S. neurona protein extract of column— purified merozoites: 1, 2 pl;
2, 5pl; 3, 10pl. Lane 4 contains 15 pl of S. neurona protein extract of un-pure merozoites.
The numbers on the lefi indicate the molecular masses of proteins in kilodaltons (KDa)
used as standard markers.

60

DNA extracts of column purified B. darlingi merozoites and S. neurona
tachyzoites absorbed markedly less light at 260nm than did extracts of crude parasite
preparations (Table 2). This absorbance reduction, if caused solely by a reduction in
DNA, would indicate a 75% and 90% reduction in the total DNA content of purified B.
darlingi and S. neurona extracts. In mixtures of purified and unpurified B. darlingi
preparations, the total DNA concentration declined proportionately with the contribution
of unpurified merozoites; purified and unpurified extracts from ~250 merozoites yielded
an estimated 218 and 790 ng of DNA, respectively (Table 3). These estimated reductions
in DNA be exaggerated if crude preparations are disproportionately contaminated with
other materials that also absorb light at 260 nm.

We preliminarily tested the effect of column purification on the detection of
parasite DNA by performing PCRs on extracts obtained from serial dilutions of purified
and unpurified S. neurona merozoites and attempting to visualize the resulting product on
ethidium bromide-stained agarose gels. PCR primers specific for S. neurona amplified
only the DNA isolated from the parasite and a diagnostic band of 334 bp was detected
from both purified and un-purified parasite templates. Even though the amount of DNA
obtained from 1000 crude S. neurona merozoites was estimated at ~1.2 times that
obtained from 1000 pure S. neurona merozoites, solely column-purified extracts
supported PCR amplification of a diagnostic marker when such extracts were diluted 10

and 100-fold (Figure 5).

61

Table 2. Total DNA concentration (ng/ pl) in purified and unpurified preparations of
DNA estimated to each contains 1000 merozoites/200p] estimated by UV absorbance

spectroscopy.

 

Before Purification

After Purification

 

Sarcocystis neurona

74.7 3: 17.93

4.5 :i: .50

 

 

Besnoitia darlingi

 

15.83i.12

 

4.37 i .31

 

Table 3. Total and parasite DNA content of defined admixtures of crude and column
purified preparations of B. darlingi estimated by UV absorbance and qPCR targetting

 

 

 

 

 

 

 

parasite rDNA.

Purernpure Observed Expected Observed Expected
ng/pl ng/pl* Ct Ct **

0:4 15.7-15.9 15.8 15.01-15.09 NA

1:3 12.0-12.7 12.94 14.63-14.85 14.86

2:2 9.4-12.3 10.08 14.07-14.47 14.69

3:1 6.0-7.6 7.22 14.00-14.40 14.54

4:0 4.1-4.7 4.36 13.83-14.22 14.40

 

 

 

 

 

*Assuming purified and unpurified contain 218 and 790 ng DNA, respectively.

"Assuming purified contain 1.57 times as much parasite DNA as crude

62

 

 

M12345678

Unpure Pure
1000 100 10 1000 100 10 +ve —ve

 

Figure 5. Analytical sensitivity of PCR analyses using the diagnostic primers

JNB25/J D396 on DNA obtained from crude Sarcocystis neurona merozoites cultured in
BT cells and PD-10 column-purified S. neurona merozoites. Lanes: M, molecular size
marker of a 100bp ladder; 1, 2, 3 unpure merozoites (1000, 100, 10, respectively); 4, 5, 6
column- purified merozoites (1000, 100, 10, respectively); 7, a positive control; 8 a
negative control. Numbers on the right and on the left of the gel are DNA fragment sizes
(in base pairs).

63

To more precisely quantify the affect of purification on the abundance and purity
of parasite DNA extracts, we developed three new quantitative real-time PCR assays.
Standard curves prepared from 5-fold serial template dilutions demonstrated an inverse
linear relationship between target DNA concentration and the number of amplification
cycles necessary to significantly exceed background fluorescence (Ct) (Figure 6A). For
assays targeting parasite rDNA and hsp70, this relationship remained linear even when
purified extracts were diluted to .00032 of their original concentration. For the assay
targeting vertebrate rDNA, the assay remained linear at dilutions as low as .0016 of that
present in the original crude extract. Repeated application of these assays resulted in
estimated efficiencies of94.6%-110.6%, and R—squared values of 957-999. Thus, each
assay supported template quantification over a broad concentration range.

We first employed qPCR assays to assess how column purification affected the
ability to detect best 188 rDNA. Colmnn purification reduced by more than 130-fold the
amount of host rDNA detectable from extracts of 1000 B. darlingi merozoites. A more
modest reduction in host DNA, of approximately 2-fold, was observed when crude and
column-purified extracts of S. neurona.were compared (Table 4). When total DNA was
first normalized to an estimated lng/ pl, purification reduced the amount of host rDNA in
B. darlingi preparations by an average of 55-fold (Figure 6B) but did not consistently
reduce the amount of host rDNA in S. neurona preparations.

Column purification consistently permitted earlier detection of paraSite rDNA
amplification products. A 0.79 and 0.65 reduction in Ct was observed for B. darlingi and
S. neurona, respectively, representing approximately 1.7 and 1.6 fold increases in the

effective concentration of each template. An independent assay targeted to the hsp70

64

gene of B. darlingi resulted in a similar detection improvement subsequent to column
purification (Table 5).

When the total DNA concentration was first normalized to lng/ul, significantly
earlier detection of parasite amplification products occurred in purified samples than in
crude samples, irrespective of the locus targeted (Figures 6C-D). The rDNA of purified
B._darlingi and S. neurona reached the critical fluorescence in such normalized templates
2.18 and 4.25 cycles earlier, representing approximately 4.5-fold and l9-fold increases in
the representation of parasite DNA (Table 6).

Finally, qPCR assays were performed on defined admixtures of purified and
unpurified B. darlingi templates in order to determine whether incremental improvements
in PCR sensitivity resulted when the proportion of purified preparations was increased.
Incremental increases in the relative proportion of purified parasites caused parasite DNA

targets to be detected earlier (Table 3).

65

Fluorescence (dR)
3 E

g

 

 

 

24081012141018202224202330323436380

 

Figure 6. Amplification plot (fluorescence vs. cycle number) in quantitative PCR
assays. (A) Proportional delay in the amplification of parasite rDNA as crude B. darlingi
are diluted, in serial 5-fold steps, over 4 orders of magnitude.

66

F Iuorescence (dR )
'8‘
O
\'

300.; ..... g ..... ..... g ..... g ..... g ..... g ..... g ..... g ..... g. ..... g

mo ..... ..... 1....5...E. ..... ..... ..... ..... ..... ..... .....

200 ..j ..... ; ..... ..... ..... j ..... j ..... ..... 3.

 

0 - :1 VMW“! W’ '_; ;_ _

 

24081012141618202224262030323430390
Cycles

 

Figure 6 (Cont’d). (B) After normalizing crude (red) and column-purified (blue) extracts
to a total DNA concentration of lng/ pl, amplification of host 18S rDNA is delayed.

67

 

E

5

  

Fluorescence (dR)

 

 

 

 

Cycles

 

Figure 6 (Cont’d). (C) Note amplification of B. darlingi 18S rDNA is accelerated.

68

F Iuorescenee (GR)

 

 

 

2408101214161820222420283032343638

Cycles

 

Figure 6 (Cont’d). (D) Note amplification of B. darlingi hsp70 is accelerated.

69

Table 4. Effect of column purification on the detection of host rDNA amplification in

preparations estimated to contain 1000 merozoites/200p].

 

 

 

 

Ct Ct ACt Fold
Before After difference*
Purification Purification
Sarcocystis neurona 21.74 :t .387 22.66 i .301 0.92 1.89
Besnoitia darlingi 22.81 i .270 29.88 :t .112 7.07 134.36

 

 

 

 

 

 

Table 5. Effect of column purification on the detection of parasite rDNA amplification in
preparations estimated to contain 1000 merozoites/200p].

 

 

 

 

Ct Ct ACt Fold
Before After difference*
Purification Purification
Sarcocystis neurona 16.15 :t .185 15.36 i: .318 0.79 1.73
Besnoitia darlingi 15.04 i .051 14.39 at .187 0.65 1.57

 

 

 

 

 

+Ct —— The critical threshold (Ct) value is the number of cycles required for fluorescence
to significantly exceed background.
*Assuming PCR product doubles with each cycle in the exponential phase.

Table 6. Effect of column purification on the detection of parasite rDNA in lng/ pl total

 

 

 

 

DNA.
Ct Ct ACt Fold
Before After Purification difference*
Purification

Sarcocystis neurona 21.42 :t .330 17.17 :t .164 4.25 19.02

Besnoitia darlingi 18.34 :l: .037 16.16 :i: .427 2.18 4.53

 

 

 

 

 

70

 

 

DISCUSSION

Accumulated host cell contaminants can hamper the study of intracellular
coccidia. Although purification methods for S. neurona merozoites have been previously
described (for example those using syringe filters, 3 pm polycarbonate filters, and
calcium ionophores) no highly efficient method suitable for routine processing has been
reported for the separation of S. neurona merozoites or B. darlingi tachyzoites fiom
tissue culture cells. We are aware of no previous attempts to quantify the affect of such
purification methods on the yield and purity of parasite DNA.

We have found a simple purification method to be suitable for separating S.
neurona merozoites and B. darlingi tachyzoites from cultured and have favorably
evaluated the method according to parasite (i) yield and purity, (ii) viability, (iii) SDS-
PAGE silver staining and immunoblotting, (iv) immune-fluorescent staining, and (v)
quantitative PCR.

Our newly developed quantitative PCR assays help evaluate the yield and purity
of such parasite preparations. The improved detection of parasite DNA in column-
purified preparations may have been caused by the removal of inhibitory substances, the
enrichment of parasite-derived DNA as a proportion of the total, or differences in the
accuracy of enumerating merozoites in purified and unpurified preparations. Whatever
the cause, parasite targets were more readily amplified after column purification. In the
case of B. darlingi preparations, contamination by host genomic DNA was markedly
reduced as a result of column purification.

The feasibility of using a PD-10 column for purification of coccidia from cultured

cells with a satisfactory recovery was initially reported by Hemphill et al., (1996)

71

working with N. caninum. In the present study, purification using a PD-10 column
proved to be a practical method that provided high yield and highly purified parasite
fractions for large—scale purification of parasites. The ability to obtain a pure
homogenous population of viable S. neurona merozoites and B. darlingi tachyzoites from
large-scale culture facilitates the biological, biochemical, immunological, and genetic
studies of these specific stages of these parasites life cycle. This technique appears to

have general applicability for purifying coccidia from various cultured cell lines.

72

REFERENCES

Dame JB, Greiner E, Ellison SP (2001) In vitro culture and synchronous release of
Sarcocystis neurona merozoites from host cells. Vet Parasitol 95:251—261

Dubey JP, Davis SW, Speer CA, Bowman DD, de Lahunta A, Granstrom DE, Topper
MJ, Hamir AN, Cummings JF, Suter MM (1991) Sarcocystis neurona n. sp. (protozoa:
apicomplexa), the etiologic agent of equine protozoal myeloencephalitis. J Parasitol
77:212—21 8

Dubey JP, Lindsay DS, Kerber CE, Kasai N, Pena HFJ, Gennari SM, Kwok OCH, Shen
SK, Rosenthal BM (2001 a) First isolation of Sarcocystis neurona from the South
American opossum, Didelphis albiventris, from Brazil. Vet Parasitol 95:295—304

Dubey JP, Lindsay DS, Saville WJA, Reed SM, Granstrom DE, Speer CA (2001b) A
review of Sarcocystis neurona and equine protozoal myeloencephalitis (EPM). Vet
Parasitol 95 :89—1 31

Dubey JP, Lindsay DS, Rosenthal BM, Sreekumar C, Hill DE, Shen SK, Kwok OCH,
Rickard LG, Black SS, Rashmir-Raven A (2002) Establishment of Besnoitia darlingi
from opossums (Didelphis virginiana) in experimental intermediate and definitive hosts,
propagation in cell culture, and description of ultrastructural and genetic characteristics.

Inter J Parasitol 32: 1053—1064

Elsheikha HM, Rosenthal BM, Mansfield LS (2004) Dexarnethasone treatment induces
susceptibility of outbred mice to Besnoitia darlingi infection. Parasitol Res (In Press)

Ginzinger DG (2002) Gene quantification using real-time quantitative PCR: An emerging
technology hits mainstream. Exp Hematol 30:503—512

Hemphill A, Gottstein B, Kaufinann H (1996) Adhesion and invasiOn of bovine
endothelial cells by Neospora caninum. Parasitology l 12: 183—197

Klein D (2002) Quantification using real-time PCR technology: applications and
limitations. Trends Mol Med 82257—260

Lindsay DS, Dubey JP (2001) Direct agglutination test for the detection of antibodies to
Sarcocystis neurona in experimentally infected animals. Vet Parasitol 95:179—186

Mansfield LS, Schott HC, Murphy AJ, Rossano MG, Tanhauser SM, Patterson J 8,
Nelson K, Ewart SL, Marteniuk JV, Bowman DD, Kaneene J B (2001) Comparison of
Sarcocystis neurona isolates derived from horse neural tissue. Vet Parasitol 95:167—178

Tanhauser SM, Yowell CA, Cutler TJ, Greiner EC, MacKay RJ, Dame J B (1999)
Multiple DNA markers differentiate Sarcocystis neurona and Sarcocystisfalcatula. J
Parasitol 85:221—228

73

CHAPTER 3
PHYLOGENETIC RELATIONSHIPS OF SARCOC YS T IS NE URONA ISOLATES
OF HORSE AND OPOSSUM TO OTHER CYST-FORMING COCCIDIA

DEDUCED FROM SSUrRNA GENE SEQUENCES

Submitted for publication as original paper:

Hany M. Elsheikha 1, David W. Lacher 2, and Linda S. Mansfield " 3 (2004) Phylogenetic
relationships of Sarcocystis neurona isolates of horse and opossum to other cyst-forming

coccidia deduced from SSUrRNA gene sequences. Parasitology Research (In Review).

I Department of Large Animal Clinical Sciences, College of Veterinary Medicine,
Michigan State University, East Lansing, MI 48824, USA.

2 Microbial Evolution Laboratory, National Food Safety and Toxicology Center,
Michigan State University, East Lansing, MI 48824, USA.

3 Department of Microbiology and Molecular Genetics, College of Veterinary Medicine

Michigan State University, East Lansing, MI 48824, USA.

74

ABSTRACT

Phylogenetic analyses based on sequences of the nuclear-encoded small subunit
rRNA (SSUrRNA) gene were performed to examine the origin, phylogeny, and
biogeographic relationships of S. neurona isolates from opossum and horse from the
State of Michigan, USA, in relation to other cyst-forming coccidia. A total of 31 taxa
representing all recognized subfamilies and genera of Sarcocystidae were included in the
analyses. Phylogenies obtained by four tree-building methods were consistent with
classical taxonomy. The “isosporid” coccidia Neospora, T oxoplasma, Besnoitia,
Isospora lacking stieda bodies, and Hyaloklossia formed a sister group to the Sarcocystis
spp. Sarcocystis species were divided into 3 main lineages; S. neurona isolates were
located in the second lineage and clustered with S. mucosa, S. dispersa, S. lacertae, S.
rodentifelis, S. murz's, and Frenkelia spp. Alignment of S. neurona SSUrRNA sequences
of Michigan opossum isolates (MIOP5, MIOP20) with S. neurona sequence of a
Michigan horse isolate (MHI8) showed 100% homology. Michigan isolates differed in
30/ 1085 (2.7%) bp from a Kentucky horse isolate (SNS). Additionally, S. neurona
isolates from horses and opossums were identical based on ultrastructural features and
PCR-RF LP analyses and formed a phylogenetically indistinct group. These findings
revealed the concordance between the morphological and molecular data and suggested

that S. neurona from opossums and horses generated from the same phylogenetic origin.

75

INTRODUCTION

Members of the genus Sarcocystis are among the most common and widespread
protozoan parasites of a wide variety of vertebrates worldwide. Studies of this genus may
be of economic importance since some of the species are important pathogens of
livestock and humans. Sarcocystis neurona is the principal agent of a serious
neurological disease in horses in the Americas known as equine protozoal
myeloencephalitis (EPM) (Dubey et al., 1991; MacKay, 1997; Dubey et al., 2001).
Horses acquire S. neurona by ingestion of the sporocyst stage in feces of opossums in
contaminated food or roughage (Fenger et al., 1997). Genetic heterogeneity among S.
neurona isolates from different animal hosts has been reported (Mansfield et al., 2001;
Rosenthal et al., 2001). This genetic diversity is matched by the wide range of hosts
infected with this parasite (Dubey et al., 2001). F urthermore, EPM has a wide range of
disease expression in horses. This range may affect the variable response to treatment
that is observed in horses infected with S. neurona and suggests that strain variation in
the organism may be a factor in the clinical management of EPM. It is likely that S.
neurona isolates from different hosts represent a single species based on a number of
characteristics, including parasite morphology, histopathology, antigenic reactivity, and
most notably the parasitic lifestyle. However, recognition of S. neurona as the single
causative agent of EPM in horses and other hosts is possibly an oversimplification.

Even though S. neurona displays considerable genetic, clinical, and ecological
diversity, this parasite has not been the subject of rigorous phylogenetic treatment of the
deeper relationships between isolates in relation to other cyst-forming coccidia. In spite

of the dramatic improvement in our understanding of the biology and lifecycle of S.

76

neurona (Dubey et al., 2001), both the phylogenetic position of S. neurona within the
genus Sarcocystis and the relationships between S. neurona isolates are ambiguous. The
monophyly of Sarcocystis spp. is uncertain partly because of difficulties in assessing the
homology of phenotypic characters. Therefore, molecular approaches might provide
alternative tools to compare, root, and clarify the phylogeny of these organisms.

Studies on the small-subunit ribosomal RNA (SSUrRNA) gene have proved
useful in elucidating cryptic diversity among parasites and in addressing phylogenetic
questions concerning many protozoa, including Sarcocystis spp. (Hasegawa et al., 1985;
Barta et al., 1991; Tenter et al., 1992; Ellis et al., 1994; Slapeta et al., 2003). Most of
these studies have questioned the monophyly of the genus Sarcocystis and provided a
strong correlation between the phylogenies of Sarcocystis species and those of their
definitive hosts. Additionally, the SSUrRNA gene is conserved among species and
complete or nearly complete sequences of this gene from many coccidia are available.
Consequently, the SSUrRNA gene was selected to examine relationships between S.
neurona isolates and other members of the family Sarcocystidae. Although our
taxonomic sampling is far from exhaustive, it is representative of the diversity within the
Sarcocystidae. We sequenced part of SSUrRNA gene for 3 selected S. neurona isolates
from a horse and two opossums from Michigan whose relationships had not been
previously examined with molecular data. Sequences from additional taxa representing
all subfamilies and genera of Sarcocystidae were included in the analyses to broaden the
context in which to interpret SSUrRNA diversity.

In this study, we tested the hypothesis of the monophyly of Sarcocystis, with

particular focus on S. neurona from the North American opossum (D. virginiana) and

77

horses, and among major Sarcocystidae lineages. Also, the hypothesis that phylogenies
of S. neurona and its hosts are consistent with a common history was tested. To
investigate these hypotheses we developed and differentiated between axenic cultures of
S. neurona isolates obtained from opossum sporocysts and from horse neural tissue using
ultrastructure and molecular characterization. Additionally, we explored the practical and
evolutionary consequences for molecular-based phylogenetic reconstruction of S.
neurona and related coccidia based on SSUrRNA sequence data by a wide range of tree-
building methods with reference to evolutionary model strategies and parasite host.
Results obtained provided new information regarding the interrelationships within S.
neurona isolates and the phylogenetic position of S. neurona isolates from opossums and

horses in relation to other coccidia.

78

MATERIALS AND METHODS

Taxa sources. A total of 31 operational taxonomic units were included in all
analyses. Sequences of 28 taxa, which span all of the currently recognized subfamilies
and genera of the family Sarcocystidae, were obtained from GenBank (Table 1). Three
taxa were axenic cultures of S. neurona collected in the current study from one horse and
2 opossums from Michigan as follows. One isolate recovered from brain tissue of a
naturally infected horse was cultured in equine dermal (ED) cells (ATCC CCL 57,
American Type Culture Collection, Rockville, Maryland, USA) as previously described
(Mansfield et al., 2001). Two feral opossums were trapped on local farms where horses
had died with confirmed EPM. The opossums were humanely euthanatized, necropsied,
and the intestines screened for Sarcocystis sporocysts. Sporocysts were recovered,
excysted, and cultured in ED cells as described (Murphy and Mansfield, 1999). In all
cases, inoculated cultured ED cells were examined every 24 hours for 7 days and weekly
thereafter for the presence of Sarcocystis merozoites by microscopy. Once significant
cell death was observed, initial cultures were passaged to fresh ED cells, and growth of
Sarcocystis was observed during long-term incubation.

Morphologic characterization. Microscopic evaluation of Sarcocystis
merozoites from ED cultures initially inoculated with infected horse neural tissues and
opossum sporocysts were performed using a stained cytospin technique to examine the
morphology of the schizonts as described (Elsheikha et al., 2003). Also, merozoites were
concentrated via low speed centrifugation (1000 xg), fixed in glutaraldehyde, and
routinely processed for electron microscopy (EM) by the MSU Electron Optics

Laboratory as described (Elsheikha et al., 2003).

79

Table 1. Details of Apicomplexan taxa included in the study.

 

 

80

Classification ' Taxa Accession Sequence Final Intermediate
No. Length Host Host
Ingroup
Sarcocystinae
Sarcocystis S. neurona AY518209 b 1085 bp Marsupiala Equidae
Sarcocystis S. neurona AY518210 b 1085 bp Marsupiala Equidae
Sarcocystis S. neurona AY518211 b 1085 bp Marsupiala Equidae
Sarcocystis S. neurona U07812 1803 bp Marsupiala Equidae
Sarcocystis S. mucosa AF 109679 1807 bp Marsupiala ?
Sarcocystis S. dispersa AF 1201 15 1610 bp Owl Muridae
Sarcocystis S. muris . M64244 1805 bp F elidae Muridae
Sarcocystis S. rodentzfelis AY0151 11 1593 bp Felidae Muridae
Sarcocystis S. Iacerate AY0151 13 1 802 bp Snake Lizard
Sarcocystis S. fusiformis U03071 1878 bp Felidae Bovidae
Sarcocystis S. buflalonis AF017121 1 884 bp Felidae Bovidae
Sarcocystis S. hirsute AF017122 1889 bp Felidae Bovidae
Sarcocystis S. hominis AF 006470 1505 bp Primates Bovidae
Sarcocystis S. sinensis AF 266959 1405 bp ? Bovidae
Sarcocystis S. arieticanis L243 82 1821 bp Canidae Bovidae
Sarcocystis S. cruzi AF 01 7120 1818 bp Canidae Bovidae
Sarcocystis S. singaporensis AF434054 1801 bp Snake Muridae
Frenkelia F. microti AF009244 1631 bp Birds Muridae
F renkelia F. glareoli AF 009245 1630 bp Birds Muridae
Toxoplasmatinae
T oxoplasma T. gondii M97703 1790 bp Felidae Mammals
Neospora N. caninum I 271354 1769 bp Canidae Bovidae
Besnoitia B. besnoiti AF 109678 1792 bp Felidae Bovidae
Besnoitia B. jellisoni AF 291426 1794 bp Felidae Muridae
Hammondia H. hammondia AF 096498 1747 bp Felidae Muridae
Hyaloklossia H.1ieberkuehni AF 298623 1564 bp Geen frog ?

Table 1 (Cont’ d).

Cystoisosporinae
Cystoisospora I. belli AF 1 06935 1784 bp Human Muridae
Cystoisospora I. ohioensis AF 029303 1392 bp Canidae Muridae
Cystoz'sospora I. suis ISU97523 1822 bp Pig Muridae
Outgroup
Eimeridae
Eimeria E. tenella U40264 1756 bp Avian ?

Cyclospora C. cayetanensis AF111183 1795 bp Human ?
Caryospora C. bigenetica AF060976 1751 bp Snake Mammals

a Family Sarcocystidae is subdivided into two subfamilies, the Sarcocystinae (2 genera)
and the Toxoplasmatinae (5 genera) (Smith and F renkel 2003).
b Present study.

DNA extraction, amplification, and sequencing. For the three Sarcocystis
isolates obtained in this study, merozoites from 3 heavily infected tissue culture flasks
were centrifuged and washed with Tris-EDTA buffer. DNA was extracted from these
culture-derived merozoites using the Qiagen DNeasy tissue kit. DNA concentration was
assessed using a NanoDrop DN-1000 Spectrophotometer and NanoDrop software version
2.5.4. (N anoDrop Technologies, Rockland, DE, USA). Then DNA was tested by S.
neurona-specific PCR-RFLP analysis for the presence of the parasite (Tanhauser et al.,
1999). Oligonucleotide primers for the SSUrRNA gene were designed from highly
Conserved regions detected by alignment of published sequences. The primers are
described in Table 2 and were used for both PCR and sequencing reactions.

PCR reactions were performed under the following conditions: the total volumes

Were 50 p1 containing 1.5 U AmpliTaq DNA Polymerase, 5 pl (10X) reaction buffer, 1.5

mM MgC12, 1 p1 dNTP (1 nM), 1 pl each primer (10 pM), 50 ng template DNA.

81

Amplification was implemented with denaturing at 94°C for 10 min, 35 cycles of
denaturing at 92°C for 1 min, annealing at 60°C for 1 min, and extension at 70°C for l
min, followed by extension at 72°C for 5 min in a FTC-100 thermocycler (MJ Research
Inc.). PCR products were evaluated on 1.5% agarose gels to check for appropriately sized
products. Gels were stained with ethidium bromide and visualized by UV

transilltunination.

Table 2. Names, locations, and DNA sequence of primers used for amplification and
sequencing of portion of SSUrRNA gene from Sarcocystis neurona isolates from horses
and opossums.

 

Primer name Orientation Position ' Primer sequence (5’-3’)

 

SN—SSUrRNA-Fl Forward 688-708 GAATTCTGGCATCCTCCTGA
SN-SSUrRNA -F2 Forward 1182-1201 CTGCGGCTTAATTTGACTC

SN-SSUrRNA-Rl Reverse 1285-1305 CATGCACCACCACCCATAGA
SN-SSUrRNA -R2 Reverse 1784-1804 CCTACGGAAACCTTGTTACG

 

3 Numbers indicate positions relative to the S. neurona, SSUrRNA sequence (GenBank
accession no. U07812).

The amplified products were purified from unincorporated primers and dNTPs
with the Qiagen QiAquick PCR Purification Kit and directly sequenced. Some samples
were sequenced twice to assure accuracy. Sequencing reactions were carried out on a
thermal cycler (FTC-100 thermocycler) with the Big Dye sequencing kit; the programs
and reaction were implemented according to the manufacturer’ recommendations, then
electrOphoresed in a Beckman CEQ 8000 sequencer (Beckman Coulter Inc., Fullerton,
CA, USA). PCR-RFLP assays and SSUrRNA sequencing were also used to distinguish

between axenic S. neurona merozoites obtained from horse neural tissues and sporocysts

82

isolated from opossums after long-term growth in tissue cultures. These procedures were
repeated six independent times for opossum isolates.

Sequence alignment and datablock creation. Sequences acquired from the
automated sequencer in the form of electropherograms were aligned, examined
thoroughly, and visually checked, and verified. Sequences of the 3 S. neurona isolates
obtained in the present study were aligned against 28 previously published SSUrRNA
nucleotide sequences from related cyst-forming coccidia obtained from GenBank (see
Table 1 for details). The total 31 sequences analyzed represented 7 genera and two
subfamilies and were a representative sample of members of the family Sarcocystidae
(e. g. species with felids, canids, and marsupials as definitive hosts). Inter-individual
variation was taken into account, by including multiple species of some genera and
multiple isolates of some species in the analyses. The former were included to resolve
relationships within certain genera and to aid in the correct placement of these genera.
The latter represented species that were originally sequenced several times due to
availability of multiple isolates and to provide confirmation of sequence data. This
strategy allows the examination of relative levels of divergence at different taxonomic
levels.

Sequences were aligned by Clustal X version 1.81 (Thompson et al., 1997),
multiple alignment parameters were as follows: gap open penalty 15, gap extension
penalty 6.66, transion weight 0.5, delay divergent sequence 30%; finally the alignment
was refined by eye. A datablock of the aligned sequences was created in Nexus format.
Additionally, the alignment was evaluated in MacClade 4.0b6 (Maddison and Maddison,

2000) with taxon names hidden. Since monophyly can only be assessed on a rooted tree,

83

outgroups were selected based on a higher level phylogeny of the Apicomplexa and
according to taxonomic and genetic criteria. Eimeridae represents the most closely
related family to the Sarcocystidae (Levine, 1985) and includes parasites such as Genus
Eimeria, and the recently identified genera Cyclospora and Caryospora. Therefore, one
sequence fi'om each of these 3 genera was chosen and these sequences were used as
outgroup taxa in all analyses. Because base compositional heterogeneity is known to
affect phylogenetic inference (Lockhart et al., 1994; Yang and Roberts, 1995), base
composition and departures from the average base-composition homogeneity across all
taxa were evaluated prior to phylogenetic analysis using a )8 analysis.

Phylogenetic analyses. Aligned sequences were analyzed using neighbor-joining
(NJ), maximum parsimony (MP), maximum likelihood (ML), and Bayesian approaches
of phylogenetic inference. All analyses were performed using the simple heuristic search
option in PAUP* Version 4.0b10 (PPC) (Phylogenetic Analysis Using Parsimony;
Swofford, 2002; Smithsonian Institute) unless otherwise stated. Molecular phylogenies
were reconstructed using the NJ method (Saitou and Nei, 1987). NJ tree was constructed
for all taxa from the distance matrix for each model tested. The relative distances among
different taxa were estimated using uncorrected “p”, F81 (Felsenstein, 1981), F84, Jukes-
Cantor (J ukes and Cantor, 1969), Kimura’s two-parameter (K2P) (Kimura, 1980),
Hasegawa-Kishino-Yano (HKY85) (Hasegawa et al., 1985), Tajima Nei, Tamura Nei
(Tamura and Nei, 1993), and General time reversible (GTR) (Yang, 1994) models. The
reliability of the NJ tree was assessed by the bootstrap method with 1,000 replications.
We used 95% as the statistically significant value (Efron et al., 1996); however, values

greater than 50% were reported since bootstrap values may be conservative estimates of

84

the reliability of groups.

In MP analysis, the single gaps and gaps that aligned unambiguously forming
indels were treated as “missing data”. Of the total 1993 character, 1646 uninformative
characters were excluded. Of the remaining 347 characters, character states were treated
as unordered and equally weighted and all were parsimony informative. An initial
heuristic search was performed on the aligned dataset (PAUP options including heuristic
search, starting tree (start) = stepwise addition, random addition of taxa sequence (10
replicates), branch swapping (swap) = TBR, accelerated transformation [ACCTRAN],
and MULTREES on). Bremer support values (Bremer, 1994) were calculated by
searching tree space for the shortest tree(s) and decay indices were calculated for each
node. Tree length (L), retention index (RI), consistency index (CI), rescaled consistency
index (RC), and homoplasy index (HI) were recorded. For evaluation of the robustness
of each branch, bootstrap values were calculated using 100 non-parametric
pseudoreplicates with heuristic searches employed within each replicate (with the same
PAUP options as the initial search and no more than 500 trees saved per replicate).
Nucleotide changes were used to determine branch lengths of the tree(s). An additional
MP analysis was performed identical to that described above, except that gaps were
treated as a fifth base. MP trees(s) were reconstructed as strict consensus and as 50%
maj ority-rule consensus tree(s).

Character evolution and character state defining clades were analyzed using
MacClade Version 4.0b6 (Maddison and Maddison, 2000). NJ and MP trees generated in
PAUP were imported to MacClade. All uninforrnative characters were excluded and

tree(s) L, CI, and RI were recorded. Character phylogeny was reconstructed and all

85

informative characters were traced to infer the evolutionary histories of the different
character states and to see how they are hypothesized to have evolved in the trees. Trees
were compared to each other and screened for any differences. Characters that supported
the monophyly of the Sarcocystidae and resolved the origin of S. neurona were examined
and compared in both trees. For these selected characters, the number of steps, CI and RI
were recorded.

Analyses under the maximum likelihood criterion were preceded by an evaluation
of alternative models of sequence evolution using a fixed tree generated NJ method
(Saitou and Nei, 1987) with J ukes—Cantor distances (J ukes and Cantor, 1969). The best-
fit model was the one for which additional parameters no longer significantly improved
the log-likelihood score, as determined with a likelihood-ratio test (Huelsenbeck and
Rannala, 1997). Evaluation of the best-fit model of nucleotide substitution was
determined by evaluating the likelihood of 56 alternative nucleotide substitution models
using the program Modeltest Version 3.06 (Posada and Crandall, 1998) and PAUP*. The
distribution of likelihood scores for all the models expressed as -ln likelihood was
obtained (unpublished data). Then, the output score file was imported to Modeltest to
compare models by likelihood ratio and Akaike Information Criterion (AIC) tests. A
series of likelihood ratio tests were performed for each successive pair of models to select
the least complex but most powerful model (Swofford et al., 1996; Sullivan and
Swofford, 1997).

The significance of increases in likelihood caused by addition of parameters (e.g.,
allowing for among-site variation in rates) was tested using the hierarchical likelihood

ratio test (hLRTs) statistic (—21nA = 2[ln Az—ln in], where M is the likelihood of the

86

restricted model (Felsenstein, 1988). This analysis selected the TrN + I + G model
(Tamura and Nei, 1993) with the following parameters (basefreqs = 0.2700, 0.1885,
0.2506, 0.2909; rate matrix = A—Czl, A—G:2.8408, A—Tzl, C—Gzl, C—T:3.9422, G—T:l;
proportion of invariable sites (I) = 0.5484, gamma distribution (F) = 0.4988) as the most
likely representation of the data. TrN + I + G model is a model with unequal rates for
transversions and transitions but with equal rates of all transversions and different rates
for each transition. The preliminary ML tree was significantly more likely under the TrN
+ I + G model than any other model (—1nL= 7745.2168, d.f. = 1, and P < 0.000001) using
hLRTs. Akaike Information Criterion test was also used. This test states that a model
that minimizes AIC= [—2.log (likelihood)] + [2(number of free parameters)] is considered
to be the most appropriate (Akaike, 1974). AIC selected the GTR + I + G as the best-fit
model with the following parameters (basefreqs = 0.2661, 0.1933, 0.2542, 0.2865; rate
matrix = A-C:0.8235, A—G:3.0090, A—T:1.3590, C—G:0.9l99, C—T:4.1126, G—Tzl;
proportion of invariable sites (I)=0.5516, gamma distribution (F) = 0.5076). The
preliminary ML tree was significantly more likely under the GTR + I +G model than any
other model (—lnL = 7741.5874, AIC = 15503.1748) using AIC test.

Subsequent to model evaluation and selection, the ML tree was determined using
a heuristic search in PAUP* without topological constraints in which the parameter
values under the best-fit model were fixed and a NJ tree was used as a starting point for
TBR branch swapping. The resulting tree topology and new parameter estimates were
used in a second round of branch swapping to provide the final ML tree; reported
parameter values were estimated on this tree. Bootstrap support for nodes in the ML tree

was evaluated for 100 pseudoreplicates using TBR branch swapping on starting trees

87

obtained by NJ. Log-likelihood scores for topologies of ML, NJ, and MP trees were
calculated using the estimated likelihood parameters and compared using Kishino-
Hasegawa test (Kishino and Hasegawa, 1989) in PAUP* with 1000 bootstrap replicates
for RELL.

To further assess the evolutionary relationships of the coccidia under
consideration, Bayesian analysis were performed using MrBayes program, Version 3.0b4
(Huelsenbeck and Ronquist, 2001) with general-time-reversible + gamma + invariants
(GTR + G + 1) model of sequence evolution and four heated Markov chain Monte Carlo
(MCMC) sampling and a default temperature set to 0.2. Parameters in MrBayes were as
follows: nst = 6, rates = gamma, basefreq = estimate. An initial 1000 generations were
run to estimate the time per generation. Then, a second run was done with ngn set to a
value that would result in an approximately 10 minute run. After the 10 minute run was
completed, the likelihood value converged to a stable value. Finally, runs were allowed
to proceed for 500,000 generations and trees were sampled every 100 generations. The
first 100,000 generations (1000 trees) were discarded as bum-in, and a 50% majority-rule
consensus tree was constructed of the remaining sampled trees using PAUP* to obtain

posterior probability estimates for each node on the tree.

88

RESULTS

Differentiation between opossum and horse isolates. The development of
axenic strains of S. neurona from horse and opossums was achieved. S. neurona isolates
from horse and opossum were identical based on ultrastructure features and by PCR-
RFLP and sequence analyses. S. neurona merozoites were harvested successfully from
ED cell cultures inoculated with opossum’ sporozoites. S. neurona ED cultures
inoculated with neural tissues of naturally infected horse also produced ample
merozoites. Parasites from both hosts underwent multiple cycles of schizogony in ED
cells for long-term. EM showed that the ultrastructure characteristics of merozoites
resulting from ED culture of opossum sporocysts were identical to axenic S. neurona

merozoites obtained from ED culture of horse neural tissues (Figure 1).

 

1 .9 V5

Figure 1. TEM of S. neurona merozoites cultured in equine dermal (ED) cells and
budding out of the residual body (RB). They had conoids (CO), micronemes (MN), and
dense granules (DG). Note the presence of dividing nucleus (N) and nucleolous (N O) and
the absence of rhoptries, which exist in S. falcatula. Scale bar = 1pm.

89

The S. neurona-specific PCR-RFLP assays produced positive tests on merozoites
obtained from opossum’ sporocysts and horse neural tissues cultured in ED cells, but did
not amplify the control S. falcatula DNA. Later testing with other S. neurona-specific
PCR-RF LP assays even after long-term grth of the organism in cultured ED cells also
confirmed identity of the horse and opossum isolates as S. neurona and not S. falcatula
(Tanhauser et al., 1999). We tested an aid for distinguishing whether more than one
species of Sarcocystis is present in a particular culture, which involved amplification and
sequence comparisons of SSUrRNA gene from 6 replicate samples from one opossum
isolate. We obtained 100% homology between all 6 replicate sequences from each
isolate (Unpublished data). SSUrRNA sequences of S. neurona isolates from opossums
and horses were aligned with the previously published sequences of other cyst-forming
coccidia(Table1). The SSUrRNA sequences of the two opossum isolates (MIOP5 and
MIOP20) and Michigan horse isolate (MIT-18) were identical, and they differed from the
sequence of the Kentucky horse isolate (SN5) only in 30 nucleotide positions.

Dataset and sequence attributes. The partial sequences of the SSUrRNA gene
of three selected S. neurona isolates from a horse and 2 opossums from Michigan were
determined (GenBank accession numbers AY518209, AY518210, AY518211). The final
SSUrRNA sequence aligrunent used in all analyses consisted of 1,993 bp including gaps
and missing data. Of these 466 (23.4%) were invariant, 1646 (82.6%) were parsimony
uninformative, and 347 (17.4%) were parsimony informative. All insertion—deletion
events were coded as missing data ("7") for purposes of phylogenetic analysis. The final
alignment contained sequences from 31 taxa, 28 for the ingroup including the three

Michigan isolates and three for the outgroup taxa (Table l). The average base frequencies

90

among all taxa were reasonably uniform, with a slight bias towards A and T (A, 0.230; T,
0.339; C, 0.226; G, 0.203). A chi-square test of homogeneity of base frequencies across
taxa showed that none of the taxa had a significant departure from expected base-
composition values (for all taxa chi-square (x2) = 200.858; df = 90; P = 0.0005). ()8 value
determined using the two-way test of independence implemented in PAUP“).

Phylogenetic inferences. Four tree-building methods were used. One of these
was a distance-based algorithmic method (NJ) and the other three methods were with
optimality criteria, MP, ML, and Bayesian. The PAUP* 4.0b10 computer program
(Swofford, 2002) was used for all methods but Bayesian, and employed heuristic
searches with global branch-swapping for those methods with an optimality criterion.
Representative trees for SSUrRNA gene based on NJ, MP, ML, and Bayesian are shown
(Figures 1—3). These trees represent the simplest analyses that resulted in the most
resolved trees.

NJ phylogenies of SSUrRNA were constructed for the three S. neurona isolates
sequenced in this study (MIOP5, MIOP20, and MIT—18) plus 28 published sequences.
Phylogenies were rooted using E. tenella, C. cayetanensis, and C. bigenetica as outgroup
taxa. NJ phylogenies were constructed using one distance-uncorrected and eight
distance-correction models described in Methods. NJ analyses identified 4 major groups
and high resolution was seen among all groups: (1) group I included Isospora spp.,
Besnoitia spp., N. caninum, T. gondii, H. hammondi, and H. lieberkuehm'. Besnoitia was
a sister group to a group containing T. gondii, H. hammondi, and N. caninum. Isospora
spp. were monophyletic and formed a sister group to the group that encompassed

Besnoitia spp., T. gondii, H. hammondi, and N. caninum. Hyaloklossia Iieberkuehni

91

(from European green frog) was the sister to this combined group I. (2) group 11 included
F renkelia spp. and non-ruminant Sarcocystis such as, S. neurona, S. mucosa, S. muris, S.
rodentifelis, S. dispersa, and S. lacertae and was strongly supported by bootstrap analysis
(93% of tress), (3) group 111 included ruminant Sarcocystis spp., and (4) group IV
contained S. singaporensis alone, which always formed a separate branch and was a sister
species to the carnivore-ruminant Sarcocystis spp. (group III). Taxa in group III were
further divided in agreement with their definitive host specificity into two subgroups: (A)
Sarcocystis species forming microcysts, with cats as definitive hosts (S. buffalom's, S.
hirsuta, S. fusiformis, S. sinensis, and S. hominis) and (B) Sarcocystis species forming
macrocysts, with dogs as definitive hosts (S. cruzi and S. arieticanis). In general, all
groups were well supported, for example, the Sarcocystis spp. groups 11 and III were
strongly supported by parsimony (93% and 100%, respectively) and NJ bootstrap
analyses (99% and 95%), respectively. The relationships within group II were not
consistent in all models used. The models F 81, Tamura Nei, Tajima Nei, and GTR gave
one tree topology. The J ukes-Cantor and uncorrected “P” distance models each gave a
different topology. The F84, K2P, and HKY methods gave a similar topology to each
other but different from other models. However, those groups that had strong bootstrap
support were found consistently using all models. The tree topology obtained with the
Tajami Nei model is shown in Figure 1, since more complex models gave substantially
similar results.

Unweighted MP analysis of 347 phylogenetically informative characters of the
SSU rRNA gene with gaps coded as missing data resulted in 117 most parsimonious trees

(MPTS), with a length of 809, a CI of 0.648, R1 of 0.812, RC of 0.526, and HI of 0.352.

92

The differences in the topologies of the equally parsimonious trees were primarily
associated with uncertainties in the relationships among taxa in group II, which contained
S. neurona. A strict consensus of these 117 most parsimonious trees (MPTS) is depicted
in Figure 1. The resulting MP trees showed identical topologies to that obtained in the NJ
analysis with the Tajami Nei model and hence only strict consensus MP tree is shown.
However, bootstrap values for the NJ analyses were shown on the tree (see Fig. 2
legend). When gaps were coded as a fifth character, the MP analysis yielded 3 MPTS (L
= 1310). A comparison of the topologies and bootstrap supports of groups obtained after
both procedures showed that resolution within all groups did not change but revealed
only one difference. When gaps were coded as a fifth character, MP trees showed a
different topology for taxa in group II which were further split into 2 distinct subgroups,
one subgroup contained 1. ohioensis, H. lieberkuehni, S. rodentifelis, S. dispersa, F.
microti, and F. glareoli and the second subgroup contained S. neurona, S. mucosa, S.

Iacertae, and S. muris.

93

79
92

98

 

d1

 

 

95

2 Besnoitia besnoiti (AF109678) \

Besnoitia jellisoni (AF 291426)
ammondia hammondi (AF 096498)

d(Is’bxoplasma gondii (M97703)

d2 Neospora caninum (A127l354)
dllsospora suis (U97523)

  
  
 

Group I

 

Hyaloklossia lieberkuehni (AF 29) J

s]:

 

d7

100

 

8
89
d13

- 10 changes

94

 

  

100 f Isospora ohioensis (AF029303)
d8
"’0 Isospora belli (AF10693 5)

 

Sarcocystis neurona, MIH8

98 fircocystis neurona, MIOPS
84 Sarcocystis neurona, MIOP20
Sarcocystis neurona (U07812)
- Sarcocystis mucosa (AF 109679)
— Sarcocystis lacertae (AY0151 13)
flfqugrcocystis mum's (M64244)
'00 Sarcocystis rodentifelis (AY015)
{dfrenkelia microti (AF 009244)

Frenkelia glareoli (AF009245)

‘ Sarcocystis dispersa (AF1201 15)
d7 100 dStarrcoqystis buflalonis (AF 0 l 7 120) D

:33 0° Sarcocystis hirsuta (AF017122)

Group II

 

 

 
    
 
  

   

 

Sarcocystis fusiformis (U 03071)
d 6Sarcocystis sinensis (AF266959)
13: ("30° Sarcocystis hominis (AF006470)
Sarcocystis cruzi (AF017120)
Sarcocystis arieticam's (L243 82) J

Sarcocystis singaporensis (AF43) Group IV

9 d4 Cyclospora cayetanensis (AFl l 1 l)
Eimeria tenella (U40264)
Caryospora bigenetica (AF 060976)

Outgroup

Group 111

Figure 2. Strict consensus of 117 most parsimonious trees inferred from analysis of 347
phylogenetically informative characters of the nuclear SSUrRNA gene of 17 Sarcocystis
spp. and 11 related taxa and rooted with three Outgroup taxa. Neighbor-joining (NJ) tree
topology based on Tajami Nei model was almost consistent with the MP trees. Values
above and below the nodes represent bootstrap resampling results (% of 100 replications
for MP analysis) and (% of 1000 replications for NJ algorithm), respectively. Numbers at
the nodes (d) indicate decay indices. Bootstrap values are reported only for clades
present in >50% of replicates. GenBank accession numbers of the SSUrRNA sequences
of the organisms are given in parentheses. To the right taxonomic affiliation is presented.
Branch lengths are proportional to distance. 50% majority role tree has identical

topology.

94

ML analyses were employed to allow a more complex model of nucleotide
evolution, including substitution rates and probabilities of nucleotide change, to be
developed and used to infer the most likely phylogenetic relationships among taxa. The
best-fitting models for the SSUrRNA sequences recovered by hLRT and AIC analyses
were found to be TrN and GTR, respectively, plus parameters to correct for invariant and
rate variation among sites. Identical trees of likelihood (—lnL) 7735.24808 and
7884.84486 were obtained by TrN + I + G and GTR + I + G, respectively. The resulting
tree from GTR +1 + G analysis was similar to that found in the strict consensus MP and
NJ analyses, with minimal topological differences in taxa in group II (Figure 3), but the
tree that resulted from TrN + I + G analysis was different from the trees that resulted
from NJ, MP and ML analysis using GTR + I + G in the relationships between taxa in the
second group. In the ML tree with TrN + I + G model, the taxa in group II were split into
two subgroups; one subgroup contained S. dispersa, S. mucosa, F renkeli microti, and F.
glareoli, while the second subgroup contained S. neurona, S. Iacertae, S. muris, and S.
rodentifelis. Log-likelihood scores of the trees were calculated using the estimated
likelihood (hLRT and AIC) parameters and compared by an Kishino-Hasegawa test
(Kishino and Hasegawa, 1989) as implemented in PAUP“ in order to test whether the
maximum likelihood tree topology was any better or worse than the other tree topologies
found in NJ and MP analyses. The likelihood scores for all trees with each best-fit model
used were obtained (Table 3). The Kishino-Hasegawa test indicated that none of these

trees are statistically significantly different from each other (P>0.05).

95

100 Besnoitia besnoiti (AF109678)
Besnoitiajellisoni (AF291426)
Hammondia hammondi (AF 096498)
Toxoplasma gondii (M97703 0)

Neospora caninum (AJ271354)

100 100 Isospora suis (U97523)

— . Isospora ohioensis (AF029303)

Isospora belli (AF 1 06935)

Hyaloklossia lieberkuehm' (AF29)

Sarcocystis neurona, MIH8

100

 

 

 

 
 
  
  
    
   
 
  

100 Sarcocystis neurona, MIOPS
Sarcocystis neurona, MIOP20
Sarcocystis neurona (U07812)
Sarcocystis mucosa (AF 109679)
00 F renkelia microti (A F009244)
Frenkelia glareoli (AF 009245)
Sarcocystis dispersa (AF 1201 15)
Sarcocystis lacertae (AY01 5 l 13)
Sarcocystis muris (M64244)
Sarcocystis rodentifelis (AY015)
100 Sarcocystis buflalonis (AF01712)

‘— , 10° Sarcocystis hirsuta (AF017122)

Sarcocystisfusiformis (U03071)
Sarcocystis sinensis (AF266959)
Sarcocystis hominis (AF006470)
Sarcocystis cruzi (AF 0 l 7 120)
Sarcocystis arieticanis (L243 82)

100

 

 

 
 
 
 
  

100

 

 

Sarcocystis singaporensis (AF43)
33 E: Cyclospora cayetanensis (AF 1 l l 1)
10° Eimeria tenella (U40264)

L— C aryospora bigenetica (AF060976)

 

— 0.01 substitutions/site

Figure 3. Maximum likelihood phylogeny of SSUrRNA among the Sarcocystidae,
rooted with three outgroup taxa constructed with a GTR + I + G model of substitution, (-
lnL: 7885) heuristic search, random stepwise addition, TBR, equal character weight, with
MULPARS in effect. The log likelihood score of the phylogeny is -7884.844. Numbers
above nodes represent percent bootstrap support of 100 non-parametric replications.
Branch lengths are proportional to the hypothesized amount of inferred evolutionary
change (ML distances), as shown by the scale bar. Sequences of the isolates in bold are
obtained in this study. Sequences of the isolates in bold are obtained in this study.

96

Table 3. The log likelihood scores (— 1n likelihood) of the topologies of ML, NJ, and
MP trees of the SSUrRNA under best-fit models of nucleotide evolution selected by
LRT and AIC testing using program modeltest. TrN = (Tamura and Nei, 1993),
GTR = (Yang, 1994).

 

 

Likelihood test Model of nucleotide Tree score

evolution ML MP NJ
Likelihood ratio TrN + I + G 7735.248 7736.325 7745.216
(LT)
Akaike Information GTR + I +G 7884.844 7885.570 7895.448
Criterion (AIC)

 

The pairwise number of nucleotide substitution differences detected among the aligned
SSUrRNA sequences from the 31 species of coccidia are shown in Table 3. There were
fewer nucleotide differences between the SSUrRNA genes of S. neurona and S. mucosa
than there were between S. mucosa and other Sarcocystis spp. analyzed. There were few
nucleotide differences between the sequences of T. gondii, N. caninum, and H. hammondi

thus explaining the polytomy observed between them.

97

 

:0 2.0 000 :0 00.0 3.0.0 00.0 3.0 00.0 who :0

«0.0 00.0 3.0 00.0 3.0 3.0 v0.0 30 00.0 00.0 000 $0

and N00 no.0 00.0 W00

3.0

00.0

8.3350300 . _ m

 

- 30 30.0 330 330 33.0 03.0 33.0 03.0 030 030 .00 30 30.0 300 300 00.0 30.0 30.0 000 000 00.0 30.0 30.0 00.0 00.0 000 00.0 30.0 300 0:05.: .m .03
- 3.0 330 30 030 03.0 33.0 33.0 33.0 33.0 30.0 00.0 300 300 00.0 000 000 30.0 030 030 03.0 30.0 300 300 000 :0 30.0 300 30.0 3330008030... 0 .03
- 33.0 30.0 30 3:0 3:0 300 30.0 30 33.0 30 330 330 330 33.0 33.0 03.0 03.0 03.0 03.0 20 20 0.0 :0 :0 :0 3.0 20 3003000.: .03 .33
- 33.0 33.0 _30 .30 .30 3.0 33.0 33.0 03.0 03.0 33.0 33.0 330 33.0 030 3.0 3.0 3.0 33.0 030 33.0 03.0 03.0 03.0 03.0 03.0 35.800003 .0103
- :0 33.0 03.0 33.0 03.0 03.0 3.0 33.0 3.0 33.0 33.0 03.0 33.0 33.0 33.0 330 33.0 33.0 03.0 33.0 000 0:0 0:0 000 03.0 332820.20 .0. .03
- 33.0 :0 33.0 03.0 330 33.0 3.0 33.0 3.0 33.0 3.0 33.0 :30 33.0 33.0 33.0 03.0 :0 03.0 :0 000 :0 00.0 03.0 RE... .0. .33
- 00.0 33.0 03.0 33.0 33.0 33.0 3.0 33.0 33.0 330 3.0 33.0 030 03.0 03.0 30 30 30 :0 30 3.0 :0 00.0 3.3.0500 .3 .03
- 33.0 33.0 3.0 3.0 33.0 3.0 330 03.0 03.0 3.0 3.0 3.0 3.0 3.0 30 $0 30 0.0 :0 :0 :0 03.0 035530133
- 00.0 00.0 3.0 3.0 33.0 33.0 33.0 33.0 03.0 03.0 .30 _3.0 _3.0 03.0 3:0 000 03.0 330 330 030 30 0300335013333
- _0.0 33.0 03.0 33.0 33.0 33.0 33.0 33.0 33.0 _30 .30 .30 :0 000 30.0 00 0:0 000 03.0 03.0 25.3 .0. :3
- 030 030 30 330 33.0 33.0 _30 _3.0 .30 .30 _30 000 30.0 :0 03.0 00.0 00 33.0 03.0 3.003%..00103

- 30.0 30.0 :0 0.0 2.0 30.0 00 30.0 30.0 30.0 03.0 _30 03.0 00 3.0 0.0 :0 3.0 02233.00 .0. .0_

- 30.0 :0 :0 :0 30.0 30.0 30.0 30.0 30.0 03.0 _30 03.0 3.0 :0 :0 :0 :0 08330. K .3.

- :0 0:0 3:0 00 30.0 00.0 00.0 00.0 03.0 33.0 03.0 :0 3.0 :0 :0 3.0 320.0: m .:

- 30.0 20 3:0 20 30.0 30.0 30.0 33.0 30 3.0 33.0 33.0 33.0 33.0 33.0 30020302 .0. 0_

- :0 20 0:0 :0 :0 :0 33.0 30 3.0 33.0 33.0 33.0 :30 30 3.250.012

- 0.0 2.0 :0 :0 :0 3.0 03.0 33.0 33.0 33.0 33.0 33.0 33.0 00:82 .m .3

- 00.0 30 :00 000 03.0 .30 _30 :0 :0 :0 :0 :0 3.805 .0. .3_

- 00.0 00.0 00.0 _30 33.0 33.0 03.0 03.0 03.0 00 0.0 33va 0:008: .013.

- 00.0 00.0 330 03.0 330 33.0 3.0 3.0 330 33.0 830920000000 .01:

- 00.0 33.0 03.0 33.0 33.0 30 30 33.0 33.0 300:): 0000.0: .3 .0.

- 330 03.0 33.0 33.0 30 30 33.0 33.0 33:50 Essa. .0. .0

- .00 30.0 00.0 0.0 00.0 00 2.0 .500 .3 .3

- 00.0 3.0 :0 :0 :0 30 3.200030 3.0

- :0 :0 0.0 0.0 :0 333.30

- 00.0 00.0 300 30.0 5:508 .2 .3

- 00.0 000 30.0 30203 g .3.

- 30.0 30.0 00090500 .33 .3

- 30.0 3:03.503 .33 .3

i QCNOEMNQ .m .—

..3 03 .03 .33 .33 .03 .33 .03 .33 .33 ..3 .03 .0_ .2 .: .0. .3. .3 .3. .3. .: .0. .0 .3 .3 .0 .3 .v .3 .3 ._ 532.380

 

05.5%? 00 3033053 00:80.6 0.30:0 08088035 .3 00583820

363:8wco .5300 0:0 .003 0.2038950. 00 38:30:33 <ZMeDmm apogee 38:08.06 Deuce—0:: 033E030 00 385.5qu .3. 23:38

98

The phylogeny found using Bayesian analysis (Figure 4) showed higher posterior
probabilities at most nodes in the tree than the bootstrap values that shown by other
methods. As before all Sarcocystis and F renkelia spp. formed a major monophyletic
group, and the internal arrangements of taxa in groups 111 and IV were stable and did not
differ from that found in other trees obtained by NJ, MP, and ML. However, taxa in
group II had different positions. Additionally, the non-ruminant Sarcocystis —S. murz‘s
and S. rodentifelis— which clustered together as before appeared as a sister group to
group containing ruminant Sarcocystis

Character evolution. For the MP tree, the tree length, CI, RI, and RC values
were, 809, 0.65, 0.81, and 0.53, respectively as inferred from MacClade. These values are
almost identical to the values of the same MP tree obtained by parsimony analyses in
PAUP. For NJ tree, the tree length, CI, RI, and RC values were, 797, 0.66, 0.82, and
0.54, respectively as inferred from MacClade. By comparing NJ and MP trees in
MacClade, minimal differences between the two tree types were observed. Indeed, there
were no consistent differences between the trees. Character 297 supported the
relationships of S. neurona isolates in both NJ and MP trees. The CIs, R13, and RCs of
this character were Is for both trees. Additionally, both trees had just two steps (i.e., two
evolutionary changes hypothesized for this character). Also, this character confirmed the
absolute similarity between S. neurona isolates from the Michigan opossums and the
Michigan horse, how they three isolates differed from the S. neurona isolate from a
Kentucky horse, and how S. neurona originated from the other Sarcocystis spp.
examined. Additionally, at least 1/3 of all included characters were consistent and

identical between the 2 trees and supported the monophyly of the Sarcocystidae.

99

100 Besnoitia besnoiti (AF 109678)
Besnoitiajellisoni (AF 29 I426)
Hammondia hammondi (AF 096498)
00 7790x0plasma gondii (M97703)
Neospora caninum (A1271 354)

76Isospora suis (U975230)
#Isospora ohioensis (AF 029303)
lsosvora belli (AF 106935)
Hyaloklossia lieberkuehni (AF 29)

71

94

 

 

 

Sarcocystis neurona, MIH8

100 Sarcocystis neurona, MIOPS
Sarcocystis neurona, MIOP20
Sarcocystis neurona (U07812)
Sarcocystis mucosa (AF 1 09679)
98Frenkelia microti (AF 0092440

F renkelia glareoli (AF 009245)
Sarcocystis lacertae (AY015 1 l 3)

54

 

192+»- Sarcocystis muris (M64244)
Sarcocystis rodentifelis (AY015)
100 Sarcocystis buflalonis (AF 017 12)
Sarcocystis hirsuta (A F0 1 71220)
Sarcocystisfusiformis (U03071)
Sarcocystis sinensis (AF266959)
Sarcocystis hominis (AF 1 09679)

100 100 Sarcocystis cruzi (AF 017120)
‘ { Sarcocystis arieticanis (L243 82)
Sarcocystis singaporensis (AF 43)
L Sarcocystis dispersa (AF 1201 15)

93

 

 
 
  

1.,

100 100

 

 

 

 

100 — Cyclospora cayetanensis (AF 1 1 1 1)
Eimeria tenella (U40264)

Catyospora bigenetica (AF060976)
— 10 changes

 

 

 

Figure 4. Phylogram based on SSUrRNA sequence alignment of 31 Sarcocystidae taxa.
Tree was constructed using the Bayesian inference and rooted on E. tenella, C.
cayetanensis, and C. bigenetica. Posterior probabilities are displayed at each node.
Sequences of the isolates in bold are obtained in this study.

100

DISCUSSION

The present study showed that axenic S. neurona isolates originally obtained from
opossum sporocysts can grow continuously in ED cells for long periods of time and have
identical morphological features and PCR—RFLP patterns that are compatible to axenic S.
neurona isolates obtained from horse neural tissues. At least 3 pathogenic Sarcocystis
spp.; including S. neurona, S. falcatula, and S. speeri; have been detected in the North
American opossums so far (Box et al., 1984; Dubey et al., 1998; Dubey and Lindsay,
1998, 1999). The presence of a mixture of any of these Sarcocystis spp. in opossums is
very common in nature (Elsheikha, unpublished observation). Thus, it was essential to
rule out the presence of any Sarcocystis spp. other than S. neurona in the tissue cultures
inoculated with opossum sporocysts. S. neurona isolates obtained in the present study
were maintained in tissue cultures for long periods whereas S. falcatula has been shown
to grow in ED cells for short time only before dying (Elsheikha et al., 2003).
Additionally, PCR-RFLP analyses of the three newly obtained S. neurona cultures were
negative for S. falcatula. S. speeri can be distinguished from S. neurona on the basis of
structure and biology (Dubey et al., 1998; Dubey and Lindsay, 1999). Unfortunately, no
samples of S. speeri from opossums were available for comparison in this study;
however, it is unlikely that it was present here because the Sarcocystis isolates from feral
opossums had morphological features very similar to those for defined S. neurona
isolates from horses based on published criteria (Dubey et al., 1998). Additionally, these
merozoites reproduced well in ED cells, unlike S. speeri, which does not reproduce in ED

cells (Dubey et al., 1998; Lindsay et al., 1999).

101

The 100% identity of SSUrRNA gene sequences of the 6 independent replicate
specimens from S. neurona isolates obtained from feral opossum sporocysts might
indicate that these isolates are clonal. The different isolates of the S. neurona species
analyzed from Michigan had the same sequence, and a very low level of polymorphism
was found between them and the S. neurona SNS isolate from the Kentucky horse. These
findings suggest that S. neurona isolates obtained from horse and opossums are similar;
they cannot be distinguished morphologically, by PCR-RF LP assays, or by SSUrRNA
gene sequence analyses. Moreover, the tree branches leading to the four S. neurona
isolates yield a polytomy since the genetic distances among them were close to zero. The
observations that S. neurona isolates from opossums and horses have a very similar
morphology and possessed identical SSUrRNA sequences were not surprising because S.
neurona cycles between opossums as definitive hosts and horses as aberrant intermediate
hosts (Dubey et al., 2001).

The family Sarcocystidae was divided into 2 subfamilies, Toxoplasmatinae and
Sarcocystinae based on morphological and biological characters (Smith, 1981; Smith and
Frenkel, 2003). Phylogenetic analyses by Carreno et al., (1998) and Ellis et al., (2000)
and the current study are congruent with this division. Eimeridae is considered to be the
sister group to the Sarcocystidae. Therefore, the approach used in the present analyses
was to use 3 sequences of Eimeridae as outgroup taxa for the alignment of SSUrRNA
sequences of representative members of the family Sarcocystidae (Barta et al., 2001).
Using this approach, phylogenetic analyses of SSUrRNA sequences by distance,
maximum likelihood, parsimony, and Bayesian methods gave trees having consistent

relationships among the taxa and split Sarcocystidae into four major monophyletic groups

102

in all analyses, in agreement with previous findings (Ellis et al., 2000; Slapeta et al.,
2002). The overall topology was similar and highly robust with all tree-building methods
and nodes above subspecies level were supported by high bootstrap and posterior
probability values (>90%). Few topological differences were found in MP analysis with
gaps coded as a fifih base, in ML analysis using the TrN + I + G model; in the Bayesian
method only in the positions and relationships of the taxa in group II.

In the present study, we examined the relationship between S. neurona isolates
and other Sarcocystis spp. and the question of whether Sarcocystis species having non-
ruminants as intermediate hosts form a monophyletic group. The previous phylogenetic
analyses of one S. neurona isolate (SN5) from a horse in Kentucky based on SSUrRNA
gene sequences indicated close but incompletely resolved phylogenetic position of S.
neurona (Fenger et al., 1994). These authors found that S. neurona isolate SN5 clustered
with T oxoplasma gondii. In this study, S. neurona isolates from opossums and horses
occupied different positions in the SSUrRN A phylogenetic tree rather than previously
reported.

Previous studies questioned the monophyly of the genus Sarcocystis based on
morphological criteria alone. The species of this genus clustered into three distinct sister
groups, one of which (group II) shared a common ancestor with F renkelia spp. The close
relationship of S. neurona, S. mucosa, S. muris, S. dispersa, S. Iacertae, and S.
rodentifelis with two F renkelia spp. supported the conclusion that Sarcocystis is
paraphyletic, in agreement with results obtained by Jenkins et al.,(l999) and Mugridge et
al., (1999). There are biological and epidemiological differences between Sarcocystis

spp. and F renkelia spp. and their designation as Sarcocystis, which encyst in muscles, or

103

F renkelia, which encyst in the central nervous system, suggest retaining the practically
useful generic distinctions. To date, the numbers of experimental and phylogenetic
studies of members of both genera are limited, therefore, deferring taxonomic changes as
suggested by Smith and Frenkel, (2003).

Phylogenetic analysis provides one of numerous frameworks for the interpretation
of biological data. Group II of the Sarcocystidae, which contains S. neurona, consistently
included species with a diverse range of hosts. This might explain partly the
inconsistency in the positions of taxa in this group. The complete life cycle of S. neurona
is almost known with opossums as the definitive host and horses and other vertebrates as
aberrant intermediate hosts (Dubey et al., 2001). Other Sarcocystis spp. that clustered
closely with S. neurona are known to cycle between cats, marsupials, lizards, birds, and
mice. In all analyses, S. neurona isolates clustered closely with S. mucosa, S. muris, S.
dispersa, S. Iacertae, and S. rodentifelis, forming a monophyletic group. Similar
relationships between these Sarcocystis spp. based on SSUrDNA alignment were also
noted by Jenkins et al., (1999). They mentioned that the group that contained S. neurona,
S. mucosa, S. muris, F renkelia, and Sarcocystis sp. (from a rattlesnake) was less robust
and the position of S. muris and Sarcocystis sp., in particular, was unstable. A more
recent study by Slapeta et al., (2003) gave also similar results. The weight of evidence
from parasite biology indicates that the grouping of S. neurona isolates with S. mucosa, S.
muris, S. dispersa, S. Iacertae, and S. rodentzfelis described here (i.e. a monophyletic
group radiating from a common ancestor) might be a reasonable conclusion.
Incongruence in co-evolution, as shown by some lizard parasites, may reflect the

existence of a common ancestor with low host specificity. However, more data from

104

representative taxa are needed, especially from Sarcocystis spp. that infect reptiles and
marsupials to resolve the relationships among the branches in this region of the tree. The
lack of resolution in that part of the tree containing F renkelia, S. neurona, S. mucosa, S.
muris, S. dispersa, S. Iacertae, and S. rodentz'felis might also indicate slow rates of
evolution. The fourth group contained only S. singaporensis, which has a snake-rodent
life cycle (Dolezel et al., 1999; Jenkins et al., 1999). S. singaporensis always formed a
separate branch and arose as a sister species to the carnivore-ruminant Sarcocystis spp. in
all analyses. A similar observation was made by Slapeta et al., (2002).

Trees obtained by NJ and MP analyses were topologically identical. However,
for a complete comparison the changes and behavior of some characters in both trees
were also compared. CI, RI, and RC are statistics describing how a particular character
behaves within a particular tree. Based on the values of CI, RI, and RC in both trees (= 1
in all cases), the evolution of character 297 in the NJ and MP trees was completely
consistent and gave us a preliminary clue about the origin of S. neurona. Additionally,
both sets of characters in each tree sharing many homologous characters with very few
different characters. Character 297 and the similar characters between the two trees
produced trees that have a reasonable biological meaning and reflected the evolutionary
relationships of S. neurona and support the hypotheses that S. neurona evolved as a
subset fi'om other Sarcocystis spp., and the monophyly of Toxoplasmatinae and
Sarcocystinae.

The Modeltest program gave two different results under two different criteria,
hLRTs and AIC. Two models of nucleotide substitutions were used, in order to test the

sensitivity of the analysis to the source of variability. As expected, the most complex

105

model GTR + I + G obtained by the AIC test was better than the TrN + I + G model
obtained by LRT. Even so, the ML tree generated by the TrN + I + G model had a higher
likelihood score (= lowest -In L) than the ML tree generated by the GTR + I + G; the
latter tree was completely consistent topologically with trees obtained by NJ and MP with
gaps coded as missing; furthermore, it agreed with the conventional taxonomy, parasite
biology, and previous studies. Trees obtained by the ML plus TrN + I + G model, the MP
tree with gaps treated as a fifth base, and the Bayesian tree disagreed with all other
methods in the arrangement and relationships between taxa in group 11.

Many studies have favored co-evolution of Sarcocystis spp. with the final rather
than the intermediate host (Doleiel et al., 1999; Slapeta et al., 2001; Slapeta et al., 2002).
Our results are consistent with this scenario. The exact geographic origin of S. neurona is
unknown, but it may not be indigenous to the New World. The similarity of S. neurona
and S. mucosa suggests that S. neurona and S. mucosa have a common evolutionary
history. The genetic distance between S. neurona and S. mucosa was found to be 4.3%
(Table 3). Escalante and Ayala, (1995) used a general evolutionary rate of 0.85%
sequence divergence/ 100 millions of years (Myr) for divergence of the Apicomplexan
parasites in general for the entire SSUrDNA and 2-4% sequence divergence/ 100 Myr for
the hypervariable regions. These findings indicate that these Sarcocystis spp. diverged
from each other at least 100 Myr ago. A slower rate of nucleotide-sequence divergence
in marsupials of 0.31% per 100 Myr was proposed by Westennan et al., (1990), which
was based on an earlier estimate of 103-128 Myr before present for the separation of
Didelphid (American opossum) and Australian marsupials (Richardsons, 1988). The fact

the divergence time of S. neurona and S. mucosa and the divergence time of their hosts

106

are similar is consistent with the hypothesis that these species co-evolved with their
definitive hosts, carnivorous marsupials (J akes, 1998; Rosenthal et al., 2001).

Finally, it is thought that the continuous land connections between South America
and Australia started to separate 140 Myr ago (Veevers, 1991). Recently, S. neurona has
been shown to be transmitted from South to North America by opossums through the
Panamanian isthmus (Rosenthal et al. 2001). Therefore, it is possible that S. neurona
may have been derived from S. mucosa as suggested by J akes (1998). The geographic
separation between Americas and Australian continents might have precluded gene flow
between S. neurona and S. mucosa, which might cause alleles to be distributed unevenly
among organisms. This isolation might have followed by founder effect and genetic drift
that enabled S. neurona to be genetically distinct from S. mucosa. Further phylogenetic
analyses of the SSUrRNA and additional polymorphic genes from a large number of
these two organisms and other Sarcocystis from the three continents are needed to

address these issues more deeply.

107

 

REFERENCES

Akaike H (1974) Information theory and an extension of the maximum likelihood
principle. In Proceedings of the 2nd International Symposium on Information Theory (ed.
Petrov, B.N. & Csaki, F.), pp. 267—281. Akademia Kiado, Budapest, Hungary

Barta JR, Jenkins MC, Danforth HD (1991) Evolutionary relationships of avian Eimeria
species among other apicomplexan protozoa: monophyly of the Apicomplexa is
supported. Mol Biol Evo 82345—355

Barta JR, Martin DS, Carreno RA, Siddall ME, Profous-Juchelkat H, Hozza M, Powles
MA, Sundermann C (2001) Molecular phylogeny of the other tissue coccidia:
Lankesterella and Caryospora. J Parasitol 87:121—127

Box ED, Meier J L, Smith JH (1984) Description of Sarcocystisfalcatula Stiles, 1893, a
parasite of birds and opossums. J Protozool 31 :521—524

Bremer K (1994) Branch support and tree stability. Cladistics 10:295—304

Carreno RA, Barta JR (1999) An eimeriid origin of isosporid coccidian with stieda bodies
as shown by phylogenetic analysis of small subunit ribosomal RNA gene sequences. J
Parasitol 85:77—83 '

Carreno RA, Schnitzler BE, J effries AC, Tenter AM, Johnson AM, Barta JR (1998)
Phylogenetic analysis of coccidia based on 188 rDNA sequence comparison indicates

that Isospora is most closely related to T oxoplasma and Neospora. J Eukaryot Microbiol
45: 184—1 88

Dolezel D, Koudela B, J irku M, Hypsa V, Obomik M, Votypka J, Modry D, Slapeta JR,
Lukes J (1999) Phylogenetic analysis of Sarcocystis spp. of mammals and reptiles

supports the coevolution of Sarcocystis spp. with their final hosts. Inter J Parasitol
29:795—798

Dubey JP, Dabis SW, Speer CA, Bowman DD, de Lahunta A, Granstrom DE, Topper
MJ, Hamir AN, Cummings J F, Suter MM (1991) Sarcocystis neurona n. sp. (Protozoa:

Apicomplexa), the etiologic agent of equine protozoal myeloencephalitis. J Parasitol
77:212—218

Dubey JP, Lindsay DS (1998) Isolation in immuodeficient mice of Sarcocystis neurona
from opossum (Didelphis virginiana) faeces and its differentiation from Sarcocystis

falcatula. J Parasitol 29:1823-1828

Dubey JP, Lindsay DS (1999) Sarcocystis speeri n. sp. (Protozoa: Sarcocystidae) from
the opossum (Didelphis virginiana). J Parasitol 852903—909

108

Dubey JP, Lindsay DS, Saville WJ, Reed SM, Granstrom DE, Speer CA (2001) A review
of Sarcocystis neurona and equine protozoal myeloencephalitis (EPM). Vet Parasitol
95 :89—1 3 1

Dubey JP, Speer CA, Lindsay DS (1998) Isolation of a third species of Sarcocystis in
immunodeficient mice fed feces from opossums (Didelphis virginiana) and its
differentiation from Sarcocystisfalcatula and Sarcocystis neurona. J Parasitol 84:1158—
1 164

Efron B, Halloran E, Holmes S (1996) Bootstrap confidence levels for phylogenetic trees.
Proc Natl Acad Sci U S A. 93:13429—13434

Ellis JT, Holmdahl OJ, Ryce C, Njenga J M, Harper PA, Morrison DA (2000) Molecular
phylogeny of Besnoitia and the genetic relationships among Besnoitia of cattle,
wildebeest and goats. Protist 151 :329—336

Elsheikha HM, Saeed MA, Fitzgerald SD, Murphy AJ, Mansfield LS (2003) Effects of
temperature and host cell type on the in vitro growth and development of Sarcocystis
falcatula. Parasitol Res 91 :22—26

Escalante AA, Ayala FJ (1995) Evolutionary origin of Plasmodium and other
Apicomplexa based on rRNA genes. Proc Natl Acad Sci U S A 92:5793-5797

F enger CK, Granstrom DE, Langemeier J L, Gajadhar A, Cothran G, Tramontin RR,
Stamper S, Dubey JP (1994) Phylogenetic relationship of Sarcocystis neurona to other
members of the family Sarcocystidae based on small subunit ribosomal RNA gene
sequence. J Parasitol 80:966-975

Fenger CK, Granstrom DE, Gajadhar AA, Williams NM, McCrillis SA, Stamper S,
Langemeier J L, Dubey JP (1997) Experimental induction of equine protozoal
myeloencephalitis in horses using Sarcocystis sp. Sporocysts from opossums. Vet
Parasitol 68: 199—2 1 3

Felsenstein J (1981) Evolutionary trees from DNA sequences: a maximum likelihood
approach. J Mol Evo 172368—376

Felsenstein J (1988) Phylogenies from molecular sequences: Inference and reliability.
Annu Rev Genet 22:521—565

Frenkel JK (1977) Besnoitia wallacei of cats and rodents: with a reclassification of other
cyst-forming isosporoid coccidia. J Parasitol 63:611—628

F renkel JK, Dubey JP (1972) Rodents as vectors for feline coccidia, Isosporafelis and
Isospora rivolta. J Infect Dis 125:69—72

109

Hasegawa M, Iida Y, Yano T, Takaiwa F, Iwabuchi M (1985) Phylogenetic relationships
among eukaryotic kingdoms inferred from ribosomal RNA sequences. J Mol Evo 22:32—
38

Hasegawa M, Kishino H, Yano T (1985) Dating of the human-ape splitting by a
molecular clock of mitochondrial DNA. J Mol Evo 21 :160—174

Huelsenbeck JP, Rannala B (1997) Phylogenetic methods come of age: testing
hypotheses in an evolutionary context. Science 276:227—232

Huelsenbeck JP, Ronquist F (2001) MrBayes: Bayesian inference of phylogeny.
Biometrics 17754—17755.

Kimura M (1980) A simple method for estimating evolutionary rate of base substitutions
through comparative studies of nucleotide sequences. J Mol Evo 16:111—120

Kishino H, Hasegawa M (1989) Evaluation of the maximum likelihood estimate of the
evolutionary tree topologies from DNA sequence data, and the branching order in
Hominoidea. J Mol Evo 29: 1 70—179

J akes KA (1998) Sarcocystis mucosa in Bennetts wallabies and pademelons from
Tasmania. J Wildl Dis 34:594—599

Jenkins MC, Ellis JT, Liddell S, Ryce C, Munday BL, Morrison DA, Dubey JP (1999)
The relationship of Hammondia hammondi and Sarcocystis mucosa to other heteroxenous
cyst-forming coccidia as inferred by phylogenetic analysis of the 18S SSU ribosomal
DNA sequence. Parasitology 119:135—142

J ukes TH, Cantor CR (1969) Evolution of protein molecules. In: H.N. Munro, Editor,
Mammalian Protein Metabolism vol. 3, Academic Press, New York, pp. 21—132

Levine ND (1985) Phylum II. Apicomplexa Levine 1970. In: An Illustrated Guide to the
Protozoa (Edited by Lee, J .J ., Hutner, S.H. & Bovee, E.C.), pp. 322—374. Society of
Protozoologists, Lawrence

Lindsay DS, Dubey JP, Horton KM, Bowman DD (1999) Development of Sarcocystis
falcatula in cell cultures demonstrates that it is different from Sarcocystis neurona.
Parasitology 1 1 8: 227—233

Lockhart PJ, Steel MA, Hendy MD, Penny D (1994) Recovering evolutionary trees under
a more realistic model of sequence evolution. Mol Biol Evo 12:605—612

MacKay RJ (1997) Equine protozoal myeloencephalitis. Vet Clin North Am Equine Pract
13:79—96

Maddison WP, Maddison DR (2000) MacClade: Analysis of Phylogeny and Character
Evolution, version 4.0. Sunderland, Massachusetts: Sinauer Associates

110

Mansfield LS, Schott HC 2nd, Murphy AJ, Rossano MG, Tanhauser SM, Patterson J S,
Nelson K, Ewart SL, Marteniuk JV, Bowman DD, Kaneene JB (2001) Comparison of
Sarcocystis neurona isolates derived from horse neural tissue. Vet Parasitol 95:167—178

Mugridge NB, Monison DA, Johnson AM, Luton K, Dubey JP, Votypka J, Tenter AM
(1999) Phylogenetic relationships of the genus Frenkelia: a review of its history and new

knowledge gained from comparison of large subunit ribosomal ribonucleic acid gene
sequences. Inter J Parasit0129z957-972

Murphy AJ, Mansfield LS (1999) Simplified technique for isolation, excystation, and
culture of Sarcocystis spp. from opossums. J Parasitol 85:97 9—981

Posada D, Crandall KA (1998) MODELTEST: testing the model of DNA substitution.
Bioinforrnatics 14:817—81 8

Richardson BJ (1988) A new view of the relationships of Australian and American
marsupials. Austr Mammal 11:71—73

Rosenthal BM, Lindsay DS, Dubey JP (2001) Relationships among Sarcocystis species
transmitted by New World opossums (Didelphis spp.). Vet Parasitol 95:133—142

Slapeta JR, Modry D, Votypka J, J irku M, Koudela B, Lukes J (2001) Multiple origin of
the dihomoxenous life cycle in sarcosporidia. Inter J Parasitol 31 :413—417

Slapeta JR, Kyselova 1, Richardson AO, Modry D, Lukes J (2002) Phylogeny and
sequence variability of the Sarcocystis singaporensis Zaman and Colley, (1975) 1976
serNA. Parasitol Res 88:8 10—8 15

Slapeta JR, Modry D, Votypka J, J irku M, Lukes J, Koudela B (2003) Evolutionary
relationships among cyst-forming coccidia Sarcocystis spp. (Alveolata: Apicomplexa:

Coccidea) in endemic African tree vipers and perspective for evolution of heteroxenous
life cycle. Mol Phylogenet Evo 272464—475

Smith DD (1981) The Sarcocystidae: Sarcocystis, F renkelia, T oxoplasma, Besnoitia,
Hammondia, and Cystoisospora. J Protozool 28 262—266

Smith DD, Frenkel JK (2003) Determination of the genera of cyst-forming coccidia.
Parasitol Res 91 :384—389

Saitou N, Nei M (1987) The neighbor-joining method: a new method for reconstructing
phylogenetic trees. Mol Biol Evo 41406—425

Swofford DL, Olsen GL, Waddell PJ, Hillis DM (1996) Phylogenetic inference.

Molecular Systematics, 2nd edn (ed. By Hillis, D.M., Moritz, C. & Mable, B.K.), pp.
407—514. Sinauer Associates, Sunderland, Massachusetts

111

Swofford DL (2002) PAUP“. Phylogenetic Analysis Using Parsimony (* and Other
Methods). Version 4. Sinauer Associates, Sunderland, Massachusetts

Tamura K, Nei M (1993) Estimation of the number of nucleotide substitutions in the

control region of mitochondrial DNA in humans and chimpanzees. Mol Biol Evo
1 0:5 12-526

Tanhauser SM, Yowell CA, Cutler TJ, Greiner EC, MacKay RJ, Dame JB (1999)
Multiple DNA markers differentiate Sarcocystis neurona and Sarcocystisfalcatula. J
Parasitol 85:221—228

Tenter AM, Baverstock PR, Johnson AM (1992) Phylogenetic relationships of
Sarcocystis species from sheep, goats, cattle and mice based on ribosomal RNA
sequences. Inter J Parasitol 22:503—5 13

Thompson JD, Gibson TJ, Plewniak F, J eanmougin F, Higgins DG (1997) The ClustalX
windows interface: flexible strategies for multiple sequence alignment aided by quality
analysis tools. Nucleic Acids Res 25:4876—4882

Veevers JJ (1991) Phanerozoic Australia in the changing configuration of Proto-Pangea
through Gondwanaland and Pangea to the present dispersed continents. Austr Sys
Botany 4: 1—1 1

Westennan M, J anczewski DN, O’Brien SJ (1990) DNA-DNA hybridisation studies and
marsupial phylogeny. In ‘Mammals from Pouches and Eggs: Genetics, Breeding and
Evolution of Marsupials and Monotremes’. (Eds Marshall Graves, J .A., Hope, RM. and
Cooper, D.W.) pp. 173—181

Yang Z (1994) Maximum likelihood phylogenetic estimation from DNA sequences with
variable rates over sites: Approximate methods. J Mol Evo 39: 1 05—1 11

Yang Z, Roberts D (1995) On the use of nucleic acid sequences to infer early branchings
in the tree of life. Mol Biol Evo 12:451—458

112

CHAPTER 4

SARCOCYSTIS NE URONA MAJOR SURFACE ANTIGEN GENE 1 (SAGI)
SHOWS EVIDENCE OF HAVING EVOLVED UNDER POSITIVE SELECTION

PRESSURE

Published as original paper:
Hany M. Elsheikha 1 and Linda S. Mansfield 1' 2 (2004) Sarcocystis neurona major
surface antigen gene 1 (SA G1) shows evidence of having evolved under positive selection

pressure. Parasitology Research 94: 452—459.

I Department of Large Animal Clinical Sciences, College of Veterinary Medicine,
Michigan State University, East Lansing, MI 48824, USA.
2 Department of Microbiology and Molecular Genetics, College of Veterinary Medicine,

Michigan State University, East Lansing, MI 48824, USA.

113

ABSTRACT

The major surface antigen gene 1 (SA G1) is conserved among members of
Sarcocystidae and may play an important role in parasite pathogenesis. Additionally,
generation and selection of different antigenic variants of SA GI has the potential for
inclusion in a subunit vaccine or in the development of a diagnostic assay. In this study,
patterns of nucleotide polymorphism were used to test the hypothesis that natural
selection promotes diversity in different parts of SA G1 of Sarcocystis neurona.
Nucleotide and amino acid sequence analysis of SA GI from multiple S. neurona isolates
identified two alleles. Sequences were identical intra-allele and highly divergent inter-
alleles. Also, phylogenetic reconstruction showed sequences clustering into two clades.
Tajima 's and F u and Li '3 neutrality tests indicated that selection is more likely to be
acting on SA GI . Moreover, a sliding window analysis based on the ratio of silent
substitutions to amino acid replacements provided strong evidence that two short
segments in the central and 3 ' domain of SA G1 have been under positive selection in the
divergence of the two alleles, suggesting that it may be important for the evasion of host

immune responses and would be a suitable target for vaccine deve10pment.

114

INTRODUCTION

Sarcocystis neurona, the major causative agent of equine protozoal
myeloencephalitis (EPM) in horses, is widespread throughout the New World (Dubey
et al., 2001). EPM is a neurological disease that imposes a serious burden on the
horse industry in the United States, for which there is no effective vaccine.
Seroprevalence rates of S. neurona among horses can reach very high levels, such as
60% in Michigan (Rossano et al., 2003) and 89.2% in Oklahoma (Bentz et al., 2003).
Even though a large percentage of horses are exposed to this parasite, the actual
incidence of EPM in horses is very low ~1% (MacKay et al., 2000). The design of an
effective vaccine is complicated by the fact that field isolates differ in
immunogenicity (Mansfield et al., 2001; Marsh et al., 2001, 2002). Additionally,
genetic information for the distinction of S. neurona isolates is lacking.

S. neurona has an immunodominant surface antigen gene 1 (SA G1), which
appears to occur as a single copy and encodes a major antigenic surface protein of
about 29 kDa (Ellison et al., 2002). SAG] has been suggested as a diagnostic marker
for phylogenetically related Sarcocystis species from different geographical areas
(Hyun et al., 2003). The genetically closely related organism T oxoplasma gondii has
an immunodominant major surface antigen P30, which has been extensively studied
with a view to the development of a vaccine (Bonenfant et al., 2001; Letscher-Bru et
aL,2003)

Positive selection and sequence polymorphism are among the major factors
that determine how the gene has evolved and, consequently, may be important for the

development of a vaccine. Synonymous (silent) mutations are largely invisible to

115

natural selection, while nonsynonymous (amino acid-altering) mutations can be under
strong selection pressure (Akashi, 1995). The rates for nonsynonymous (p3) and
synonymous (pg) mutations were defined as the numbers of substitutions per site.
Comparison of the nonsynonymous/synonymous substitution rate ratio (w=dN/dg)
provides a powerful tool for understanding the effect of natural selection on molecular
sequence evolution and can be used as an indicator of selection pressure at the protein
level (Ohta, 1993). In general, amino acid sites in a protein are expected to be under
different selection pressures and have different underlying wratios, where w=1
indicates neutral mutations, w<1 means purifying selection, and w>1 indicates
positive selection.

There is a paucity of published data on the extent of sequence polymorphism
and evolution of the outer membrane protein genes, including SA GI of S. neurona.
Therefore, the objectives of this study were to examine the molecular basis of genetic
variation and adaptive molecular evolution of SA G1 of S. neurona and to identify the
amino acid sites of this gene under diversifying selection. The main findings of this
study are that: (1) parts of S. neurona SA GI have evolved under positive selection,

and (2) SA GI has two alleles in the limited number of S. neurona isolates examined.

116

MATERIALS AND METHODS

Nucleotide sequences. All SA GI sequences used in the study were obtained
from GenBank. Nine SA G1 sequences of S. neurona (accession nos. AF401682,
AF397896, AY032845, AY245695, AY170900, AF480854, AY170620, AF480853,
AY245696) were used as in—groups. One T. gondii SAGI sequence (accession no.
AY217784) was used as an out-group to root the phylogenetic tree(s).

Sequence alignment and phylogenetic analyses. Nucleotide sequences were
aligned using the multiple sequence alignment program Clustal X (Thompson et al.,
1997) with the default parameters and checked by eye. Base frequencies and pairwise
sequence divergences were obtained using the program PAUP. version 4.0b10
(Swofford, 2002). All insertion-deletion events were coded as missing data (“17") for
the purpose of phylogenetic analysis. Phylogenetic trees were sought with both
maximum parsimony (MP) and neighbor-joining (NJ) using MEGA (Molecular
Evolutionary Genetics Analysis, version 1.01) software (Kumar et al., 1994). To find
the shortest (MP) tree, a branch-and-bound search was performed with unknown
initial upper bound and ‘furthest’ taxa addition sequence. NJ searches were
performed to assess the influence of alternative phylogenetic algorithms on the tree
topology using the Kimura 2-parameter (K2P) distance model of sequence evolution.
Relative bootstrap support for nodes in resulting trees (both by MP and NJ) was
evaluated using 1,000 pseudoreplicates (Felsenstein, 1985). The maximum-likelihood
(ML) tree was obtained using a heuristic search in PAUP* without topological
constraints in which the parameter values under the best-fit model were fixed and a

NJ tree was used as a starting point for TBR branch swapping. The resulting tree

117

topology and new parameter estimates were used in a second round of branch
swapping to provide the final ML tree. Bootstrap support for nodes in the ML tree
was evaluated for 100 pseudoreplicates using TBR branch swapping on starting trees
obtained by neighbor joining. Bayesian phylogenetic analysis implementing the
GTR+ F model was conducted using the software program MrBayes version 3.0
(Huelsenbeck and Ronquist, 2001). Four (nchains=4) heated (temp=0.5) Markov
chains were calculated simultaneously and sampled every 100 generations
(samplefreq=100) for 2,000,000 generations. The first 2,001 (of 20,001) trees were
discarded (bumin=2,001). The posterior probabilities of each branch were calculated
by counting the occurrence in trees that were visited during the course of the Markov
Chain Monte Carlo (MCMC) analysis.

Statistical tests of neutrality and sequence polymorphism. The numbers of
polymorphic sites in the nucleotide and amino acid sequences of each identified allele
were computed using the programs PSFIND (version 2) and PAP IND (version 1.1),
respectively. These statistics were then used to draw the similarity plot (Figure 1A, B)
using Happlot program (version 1.1). To examine how the level of selective
constraint varies along different parts of SA GI, nucleotide site differences were
calculated by the method of Nei and Gojobori, (1986) and tabulated in a sliding-
window analysis of 30 codons for the length of the protein-coding region of the gene
(Figure 2) using the program PSWIN (version 1.1). The difference pijs is a measure
of the degree of selective constraint: the more negative the value, the less the
contribution of replacement substitutions and the greater the contribution of

synonymous substitutions. The zero-difference line indicates selectively neutral

118

variation, where the per-site rates of synonymous and nonsynonymous substitutions
are equal. A positive difference, where amino acid replacements exceed the silent
substitutions, suggests the action of diversifying (positive) selection. The PSF IND,
PAF IND, and PSWIN programs were written and kindly provided by Dr. T.S.
Whittam (Microbial Evolution Laboratory, National Food Safety and Toxicology
Center, Michigan State University, East Lansing, MI 48824, USA)

The Approximate Likelihood Ratio (ARS) analysis software (Pond and Frost,
Antiviral Research Center, University of California, San Diego;
http://www.datamonkey.org) was used to evaluate selection pressures on SA GI gene.
This software uses a maximum-likelihood approach with codon-based models to
estimate the ratio (at) of dN, the rate of nonsynonymous substitutions per
nonsynonymous site, to d5, the ratio of synonymous substitutions per synonymous site
(Figure 5).

To determine if the sequence diversity observed in SA G1 is a consequence of
diversifying selection, which might suggest that it is a target of protective immune
response, additional measures of DNA sequence polymorphism, such as the number
of segregating (polymorphic) sites, number of haplotypes (NHap), haplotype diversity
(HapDiv), nucleotide diversity (0), and F—statistics, were computed using the DnaSP
program version 3.52 (Rozas and Rozas, 1999). The genetic polymorphism was also
estimated by the statistic 7r, which is the average number of nucleotide differences
per site between two sequences. Because selection may have played a role in the
evolution of SA GI of S. neurona, and divergence between isolates (given the putative

short divergence time), polymorphism data for each allele were tested by Tajima

119

(Tajima, 1989), and Fu and Li (Fu and Li, 1993) neutrality tests. Tajima ’s test
compares two estimates of genetic diversity: the mutation parameter (0) and the
difference between pairs of isolates (71'). Fu and Li '5 test is related to Tajima 's and
compares the number of singletons to that expected by the neutral theory. It was
performed without an out-group. Gaps and a few sites with unreliable information
were coded as missing and ignored. In general, mutations occur at intermediate
frequencies under positive selection. Thus, positive selection results in a small
number of segregating sites relative to the average number of pairwise nucleotide
differences between sequences (77), which is reflected by positive values of Tajima’s
test (Tajima, 1989) and also results in decreased mutation rates in external branches
of the genealogy, which is reflected by positive values of Fu and Li ’5 test statistics
(Fu and Li, 1993).

To investigate the role of intagenic recombination in SA GI sequence
polymorphism, the degree of linkage disequilibrium (or nonrandom association
between variants of different polymorphic sites) was estimated using the DnaSP

program, with the following parameters: D, D', and R. Both the two-tailed Fisher '3

X . . .
exact test and the 2-test were computed to detenmne whether the assocratrons

between polymorphic sites were, or were not, significant.

120

RESULTS

Nucleotide sequence diversity. Aligned sequences used for phylogeny
reconstruction were 1,685 bp in length. Of these, 852 (51%) were variable and 208 (12%)

were parsimony informative. Mean base proportions were 20.9% A, 22.9% T, 28.2% C,

.X . . .
and 27.9% G. 2-test of homogeneity of base frequenc1es across taxa showed sllght non-

significant bias to the nucleotides G and C; 334.36 A, 366.65 T; 451.30 C, and 447.69 G;

x2=23.766 (d.f.=24), P=0.4749. Sequences were clustered into two grOUpS based on the

uncorrected pairwise sequence divergence (p) model, and the degree of divergence
between the two groups was large but minimal within each group (Table 1). Nucleotide
similarity analysis revealed that within each monophyletic group of S. neurona isolates,
sequences of SAG] were 100% identical while sequence similarity between the two
groups was 71%. The fact that sequences within each cluster were identical and very
different between the two clusters clearly indicates that there are two SA GI alleles,

which, hereafter, we refer to as H1 and H2.

121

Table 1. DNA sequence differences of SAG] of Sarcocystis neurona isolates based on
pairwise comparisons using an uncorrected (“p") distance matrix.

 

 

 

 

 

.-.-...- ... ._—v—‘ ——

 

--.... -....- ..---- -.
l
:

l3. Sn-mucat2l(AY245695) 10.000000001- l
l4.UCDl {(AF397896) 0.00010000100001 l
35.1mm l(AF40l682) 100001000010.00010000i- l l
36.SN3 i(AYO32845) 1000010001(00000001100011- l g l
l
l

 

22. Sn-mucat2 l(AY170900) [0.00014 1

 

...—3...." ...—....q

 

 

llsolate iAccession no.l1 ,2 i3 l4 35 6 l7____l8 "l9
{1.CAT2 1(AF480854) [4 . l T” W l
l

l

l

l

[.

i7.MU-1 I(AF480853) [0.243l0.237|0.243l0.238l0.238l0.238 1
l8. Sn-MUl [(AY245696) [024210.241 [0.243102421024210243 {0.0001-
9. Sn-Mul §(AY170620) 10.255 [0393 1024010335 [0.402 [0.240|0.00910.000 (-

 

 

 

 

Phylogenetic inferences. The evolutionary relationships between the two SA G1
alleles were investigated using a phylogenetic analysis. A branch-and-bound search with
equally weighted data produced a single tree, which was rooted by the T. gondii
sequence. The SA GI tree (Figure 1) indicates clearly that there are two well-defmed
clusters: cluster I consisted of six nucleotide sequences representing three S. neurona
isolates; the SN3 and UCDl isolates from horses from Panama and California,
respectively, and the CAT2 isolate from a cat from Missouri, and cluster 11 included three
SA GI sequences of the isolate Sn-MUl from a horse from Missouri. The SAG]
genealogy was also estimated using the NJ method with the K2P model since more
complex models gave substantially similar results (data not shown). The tree obtained by
the NJ method had identical branching patterns and very close bootstrap support to the
tree found by the MP method (Figurel). Bootstrap values for NJ and MP methods

obtained from 1,000 replications were robust with 100% support at the basal node of each

122

clade. Maximum likelihood and Bayesian phylogenetic analyses of the nucleotide
sequences also consistently yielded trees with almost identical topology (data not shown),

indicating that the inferred phylogenetic relationships are robust to different methods of

tree estimation.

S. neurona (AF397896. Horse, California)

S. neurona (AF 401682, Horse, California)

52/51 S. neurona (AY032845, Horse, Panama)

 

77/ 75 S. neurona (AY170900, Cat, Missouri)

100/100'3. neurona (AY245695, Cat, Missouri)

S. neurona (AF480854, Cat, Missouri)

 

S. neurona (AF480853, Horse, Missouri)

 

100/100I
S. neurona (AY245696, Horse, Missouri)
99/96

 

S. neurona (AY170620. Horse. Missouri)

 

T. gondii (AY217784)

 

|———l
0.1 substitution

Figure 1. Maximum parsimonious phylogenetic tree from analysis of 208
phylogenetically informative characters of SA G] showing relationships among the
Sarcocystis neurona isolates (rooted to Toxoplasma gondii). The first of the two
numbers at the nodes represent bootstrap resampling results based on maximum
parsimony analysis (MP, % of 1,000 replicates). The second number represents the
bootstrap support using neighbor-joining algorithm (NJ, % of 1,000 replicates).
Numbers in parentheses after taxon names refer to GenBank accession nos., followed
by the host and origin of isolate. Tree statistics are length (L)=290, consistency index
(CI)=O.997, retention index (RI)=0.998 excluding uninforrnative characters

123

Evidence for positive selection of amino acid changes. The degree of
nucleotide and amino acid polymorphism and divergence between H1 and H2 alleles
was remarkable. Variations in sequences were distributed almost across the whole of
SA GI (Figure 2A—B). Nucleotide diversity (0) was also remarkably high (0.16545).
The ratio of nonsynonymous to synonymous mutations in the divergence of the whole
gene between the H1 and H2 alleles was 0.73:0.05, a ratio that would be predicted
for unconstrained, neutral variation. Three exceptions were noticed. First, the region
encoding the signal peptide sequence (Figure 3) showed the least polymorphism;
amino acid replacements were highly constrained in this part of the gene. Secondly,
there were two large segments where the per-site nonsynonymous-substitution rate
peaked upward (Figure 4). These peaked regions of the gene showed the most
polymorphism as shown in Figure 2A, B.

Thirdly, SA GI showed much less constraint on amino acid-altering mutations
in particular regions among alleles. Of the whole open reading frame (980 bp) of
SA GI, only two stretches, located in the central and the 3' regions of the gene, were
identified where there was a positive difference between the nonsynonymous- and
synonymous-substitution rates. The longest stretch with dN/ds>1 was ~51 codons in
length and ran from position 114 to position 165. In this region, the rate of
substitution per 100 sites (corrected for multiple hits) was dN=1922t16 and
ds=165i21. The second region extended from codon 222 to codon 259. In this region,
the rate of substitution was dN=2962t15 and ds=255:19. Thus, a moderately positive

selection pressure is acting on these regions of SA GI .

124

Approximate Likelihood Ratio (ARS) analysis also identified a positive
difference between the nonsynonymous- and synonymous-substitution rates in the
same two areas located in the central and the 5’ regions of the gene (Figure 5).

There were two haplotypes, and haplotype diversity (expected allele
heterozygosity) was 0.667 (variance 0.04167, SD 0.204). The number of polymorphic
(segregating) sites was 204. This dataset had a nucleotide diversity of 0.16545 (
variance 0.0025663, SD 0.05066). A number of statistical tests were used to
determine departure from neutrality at the SA GI locus. Tajima '5 test is the most
powerful for testing alternative hypotheses involving selective sweeps, population
bottlenecks, and population subdivision (Simonsen et al., 1995). For the SA GI
sequence data of S. neurona, all values of neutrality tests were significant, indicating
an excess of intermediate frequency variants and balancing selection (i.e., rejected the
neutral model of molecular evolution). Tajima 's D test was 2.33106 with a P<0.01.
The Fu and Li statistics, D and F had values of 2.33108 and 2.50897, respectively,
both with P<0.02. Fu ‘5 Fs statistic was 12.990 (P>0.001). No linkage disequilibrium
tests were significant (P<0.333) and there was no evidence for any recombination

events.

125

 

A

 

 

 

5-prime Intron
30.5% 40%

3-prime
24%

 

Domains

 

 

 

 

 

 

S. neurona
I A F401 682)

 

 

 

 

 

 

S. neurona J

 

 

 

 

(AY170620)

 

B

 

 

 

l

Domains I I

 

 

 

 

 

 

S-prime

 

 

 

 

 

Ell ll|||||||l|ll|l|llClll|

 

 

 

 

 

 

 

 

 

 

 

 

3-prime

 

 

 

S. neurona
(AF401682) -

S. neurona
(AY170620)

 

I II I
lllllllll ll Illllllllllllll

11111]

 

[I

 

 

 

Figure 2. Diversity in SA GI gene of Sarcocystis neurona. Plot of the variable
nucleotide sites (A) and amino acid sites (B) in pairwise comparisons of the two alleles of
SA G1 gene with percentage of the nucleotide polymorphism shown above the domains.
General three-domain structure model of Sarcocystis neurona SA GI gene is drawn above
with the intron region shaded in grey. The arrow marks the signal peptide sequence
region. Each breakpoint in the domains denotes the location of an alignment gap. Each
vertical line marks the location of a nucleotide or amino acid difference between the two

alleles sequence.

126

Domain = S-prime
S. neurona AF401682 ATGACGACGGCGGTGCTGCTGACGTTTCTGACACTCTGCTCCGCCAGAGTGTCCCTTGTG 60
S. neurona AY170620 ............ A ........... A . C .......................... AT . C . . .

 

S. neurona AF401682 AGGGCCGGAGCGCCGCCTCAAGCAACGTGCGCCAATGGCGAAACGACTGTTACTAAGCTC 120
SneuronaAYl70620 .AC ..... G.G.G.C.G ...... G..A....AAG.G...C ......... C ...... A...

S. neurona AF401682 GGCAGCTCTGGCGCACTACGAATCCACTGCCCAAATAATTTTCGACTC---GCGCCCCGG 180
S.neur0naAYl70620.AG.A.C ...... T ..... A.C.A..A ..... GGC.CGG.A..A...GAATC..G.T.CC

S. neurona AF401682 GCTGGGAATGACGCCGGTCAGATGCAGGTCTATGCAACTGCGGTTGCTGAGAATCCTGTA 2 4 O

S. neurona AY170620. .C.C.G.C.C.AA. .AAGGT .......... T.A. . . .C. . . .C. .TC.G. . . .G. . . .G
S. neurona AF401682 AACATACGAGACGTCCTGCCCGGCGCATCTTACCTCTCTGTACAGAACGTCCCG ------ 300
S. neurona AY170620 GCGC.T.A. .GT ..... C ......... A ..... G.TCG. .CG.CA. .T.GAG.TGGAGCT

S. neurona AF401682 ACCCTCACCGTCCCGCAATTGCCCGCCAAAGCTACGAGCGTCTTTTTTCACTGCCAGCAG 360
SneuronaAYl70620..A..G...A ....... GC ..... C.A..G...G....T..G...A....A ........ A

S. neurona AF401682 CAA---CCCGACAACCAATGCTTCATCCAGGTAGAAGTAGCGCCGGCTCCGCGCCTAGGT 420
S. neurona AY170620 . . . GGA . AAC . GGGA . . G ......... G ..... C ........ GG . T ..............

Domain = intron
S. neurona AF401682 AGGTTTATATGCATTAGGAGTAGGTTCATATGCATTAGGATGCGTACTCGTGGCTCGAAT 480
S. neuronaAYl70620 .A. . .CTC ------------ . .T. .ATG. . . .TG.--.TTCC.T.CG.T. .A. . . .T. ..

S. neurona AF401682 GTTCCACTGCAGCGCGGTGAGATTGGCGGCTAACCGGGTAAATGTGCGTCTTTTTTTGTT 540
S.neuronaAYl70620.GAG ......... T.C...C..C....A ....... C..GT.TC....ACT ..........

S. neurona AF401682 ACGCAGGT
S. neurona AY170620 . . .T. . .-

Domain = 3-prime
S. neurona AF401682 CCGAATACCTGCGCGGCGCTGCAGTCCACGATCGCCTTCGAAGTTCAACAAGCGAATGAA 610
S. neurona AY170620 .................. G . A ....... G ..... AT ..... GA . .AG . GC . CA .......

S. neurona AF401682 ACAGCAGTCTTCAGCTGCGGCGAGGGACTTGCTGTGTTCCCGCAAGGTAGCAAAGCGTTG 670
S. neurona AY170620 G ....... G ............. C ..... GG. . .CC . . .T .....................

S. neurona AF401682 GATGAAGCCTGCTCCAAAGAGCAGGCCCTACCCAGTGGCGCCGCTTTAGCTCCAAAGGAT 730
S. neurona AY170620 ..... T ........ G ........ AT .......... C. . . .TG ..................

S. neurona AF401682 GGTGGG---CTCCACCTTGGTTTTCCTCAGCTTCCTCAGCAGGCTATGAAGATTTGCTAT 790
S.neuronaAYl70620.C...TTCGT....G ..... C ..... G ..... C ...... A.CC..C.A ..... A..T...

S. neurona AF401682 ATTTGTACGAATGGTGGTGTGCAGGCAGAGGCGGCC---CAACGGTGTGAGGTTCGCATC 850
SneuronaAYl70620 ..... C...G.A ..... CCA..G..T...T ..... GGAT...A ....... A....AT...

S. neurona AF401682 TCCGTCGCAGCGAACCCAGACGGAAGCGTTCCAGGGGCTAACGGAGCCGCCTCTCTAGGA 910
S.neuronaAYl70620 G....T....G.GC.GA..C....G..CC.A ..... C...C ..... G...T...G.G..C

S. neurona AF401682 GCTGCCGCACGCAGCGCCTCTGCGTTAGGGTTGGCTCTCGTTGCAGGCGCTTTCTTGCAC 970
S.neuronaAYl70620 C....T ................. A....TCC ..... G ......... C..G ...... C..T

S. neurona AF401682 TTTTGCTAA 980
S. neurona AY170620 ..... G . G .

Figure 3. Nucleotide alignment of 2 SA G] sequences representing the 2 alleles of S.
neurona SA G] that shows predicted cleavage site for a signal peptide (underlined) as
identified by Hyun et al., (2003). Dots indicate sequence identity and dashes indicate
alignment gaps.

127

 

Pn—Ps

Proportion of sites (x100) _.
e a a '3 o s a s s s

 

Codon pos'tion

 

 

 

Figure 4. Plot of the number of substitutions per 100 sites for synonymous (pg) and
nonsynonymous (p3) sites between H1 and H2 SAG] alleles in a 30-codon sliding
window. The difference (Ml—pg) is a measure of the level of selective constraint on
various parts of the molecule.

128

3.3.3.338 Amm<v 200m 08.0.8.5 8082033503 mafia 05w Nb Em. 05 0o 30:3 .808 0353 .80.. mason 323.3323. 033.02% .m 959%

 

Ill

.33.... 03.0... o. A. ..0 30. us 3.0... o .30. v3 3.0.. .02. o.

 

 

 

3.20.32— ..0—EU
mam 03h mmm ovm mmm own map 000 000 000 09 09 wow 00 B. 00 mv on 3 0
fl. .. . - . . lit . . . Firii-. . . ..LlLilL.|0le.- Ill .. .1. L . . ;-LI-..r...-iL.1!F . -r-. .I .r. 1.. . .. L . r .-.». -3 l»; . 3 .. . . ml
. NNN ‘
o 0.3 0: 00 o 3.3-
foam 0 mm _
_ . 003 _
v 03 cm? a N-
. o o o o o 003 _
o o o o o o o . 3.3-
o 3.3 03. _
333 303 303 .33. 3: .
3.. .7...
o o o o . T
o o o o o o _ ...—.37.
0.0..
_ oo
.
_ o
o o fin 0
o

 

129

DISCUSSION

A better understanding of host-parasite interactions in nature is essential for
the development and deployment of a successful vaccine against S. neurona infection
in horses. One of the leading strategies for developing a vaccine against S. neurona is
based on one of the major surface antigen genes, which are accessible to the host ’3
immune system. Immunity conferred by such vaccines might have the ability to block
the invasion of the host cells by the parasite. However, surface antigens are always
known for their antigenic variation, which is on the main mechanisms used by some
parasites to evade host immune responses. Additionally, an immune response
generated against one form of an antigen might not be effective in controlling
infection with parasites of the same species expressing different forms of the antigen,
and this represents a major challenge for the development of an effective vaccine.

In view of the lack of information on the genetic diversity in S. neurona,
assessing the genetic diversity of isolates from horses and other hosts at the species or
subspecies level is a preliminary step towards the understanding of their potential
evolutionary relationships in connection with both their wide host range and their
particular reproductive and parasitic strategies. In this context, investigation of the
extent of the genetic polymorphism in SA G] of the protozoan S. neurona, and the
processes that maintain the observed diversity, is a matter of epidemiological and
evolutionary significance. Previous study has addressed the diversity of SA GI and
reported 73—100% sequence homology among S. neurona isolates (Hyun et al.,
2003). In the present study, similar nucleotide diversity was identified in SA G1 and

nucleotide sequence variation was observed in almost all of the gene, particularly in

130

the intron region. Two hypotheses can be suggested to explain the high variability of
SA G] of S. neurona observed in our study. First, high levels of intragenic
recombination are occurring between different lineages. Recombination could be a
major source of variation that might have created the new SA GI allele and generated
putative antigenic diversity. In this study, a low and non-significant level of linkage
disequilibrium was detected, which provides weak support for the possibility of
frequent SA GI recombination among these lineages. However, studying
recombination and linkage disequilibrium in a gene under positive selection as sole
evidence for sexual recombination might be inappropriate. A better understanding of
the nature and extent of genetic recombination in SA G] can be securely accomplished
by examining sequences from multiple isolates from different geographical areas
using numerous statistical and evolutionary tests. The second hypothesis, for which
there is no current supporting evidence, is that this gene might undergo “
unconstrained evolution” through the rapid fixation of neutral mutations by genetic
drift. In either case, description of the role of natural selection is problematic: in some
parts of the gene, it can be a diversifying force promoting protein polymorphism and
rapid evolution, whereas elsewhere it operates as a conservative force against amino
acid change.

Phylogenetic trees constructed with SA GI sequences showed no indication of
clustering of sequence alleles from the same geographic origin (Figure 1). In this
study, isolates genetically identical at the SA GI locus were widely distributed
throughout the United States, without any particular geographic segregation or any

association with opossum distribution. S. neurona can be considered a generalist

131

parasite because of its long list of intermediate host species. In addition to opossums,
these intermediate hosts and other vectors, particularly long—distance travelers such
as migrating birds and cats accompanying people, can play an important role in the
distribution of S. neurona and promoting their diversity across the United States.
However, it is unknown whether or not the alternating life cycle of S. neurona that
involves the passage of the parasite through opossums and different intermediate
hosts could have any impact on genetic diversity. Potential selection pressures could
stem from differences in host immune responses to infection or to differences in the
physical host environment. However, it is unclear if the genetic differences among S.
neurona isolates are due to selection pressure exerted from the host species or from
the off-host environment; both factors are likely to play a role. Horse is an accidental
intermediate host and infection in horses consists only of the asexual stage of the
parasite. It is conceivable that there will be no selective input of the horse immune
system because of its role as a “ dead-end” host. The chances of passing this
organism from infected horses to other hosts are very low and even if selective
pressure was generated within the horse, it will not contribute to gene flow/change
within the population.

Positive selection is very important for the evolution of a new or improved
protein function, as it offers fitness advantages to the protein and increases the
probability that an advantageous mutation becomes established in a population. On
the contrary, purifying selection eliminates mutations due to the deleterious effect that
they might have on the performance of the encoded protein. During SA GI evolution,

different codons might have undergone different ratios of nonsilent to silent base

132

substitutions (dN/ds). Previous studies have shown that genes on which positive
selection may operate can be identified by comparing synonymous substitution rates
(pg) and nonsynonymous substitution rates (pN). To test this hypothesis, the
proportions of pN and pg polymorphic sites of the protein-coding region of SA GI were
calculated and compared across SA GI . These analyses identified two segments of

SA G] that had an excess of nonsynonymous substitutions and thus are a likely target
for diversifying selection, where natural selection has favored the substitution of
amino acids, and suggesting a departure from neutral evolution. These findings were
also supported by results obtained by the approximate likelihood ratio (ARS)
analysis. Because these short segments lie within the major antigenic surface protein
of SA G], the evolutionary benefit to new variants may be the enhanced ability to
escape host immunity and colonize new hosts. These two segments might mediate the
interactions of the parasite with host cells and codify the organism virulence traits.
The reason for the presence of selective pressure at these particular parts of the gene
remains unclear. However, it is possible that positive selection pressure does occur at
a larger part of the gene but that conformational constraint on the SA GI molecule act
to reduce the value of dN/ds below the critical threshold, but there is currently no
theory that allows such an effect to be quantified.

Tajima ‘s D and Fu and Li '5 D and F neutrality tests also detected significant
evidence of positive selection in SA G1 . All values obtained were high and significant.
These results indicated that the polymorphism found in SA GI alleles might be
maintained by positive natural selection. Neutrality tests are known to be conservative

when applied to sequences that are subject to recombination (Tajima, 1989; Fu and

133

Li, 1993). However, sequences analyzed in this study showed no evidence of
recombination. These findings indicate that it is important to perform many tests to
accurately evaluate the presence/absence of positive selection acting on a particular
gene (Conway and Polley, 2002). Although our findings are not completely
conclusive, considering the strengths and drawbacks of each test used, they do
indicate clearly that diversifying selection is acting on two regions in SA GI . These
results are further reinforced by the remarkably high number of polymorphic sites and
genetic diversity in SA G1 alleles, which indicates the presence of strong diversifying
selection acting on this gene.

Positive selection analysis of genes coding for cell surface proteins of the
malaria parasite, Plasmodium spp., have identified gene regions that have
nonsynonymous substitution rates that are significantly higher than the synonymous
rates (Ohta, 1991). Since host antibodies target these protein regions, it has been
suggested that this phenomenon is evidence of positive selection-based diversification
of the protein to maintain an advantage over the host immune system. Thus, given the
fact that the SAG] sequence is highly polymorphic and SAGl antigen is serologically
immunodominant (Ellison et al., 2002) and involved in the attachment and invasion of
host cells by S. neurona (HM Elsheikha et al. unpublished data), this protein possibly
is important in allowing the parasite to evade the host immune response and the
selection pressure is presumably the surveillance of the host immune system. In this
event, the support for positive selection pressure on SA GI is consistent with the
biological filnction of this gene and suggests that the SAGl protein could be a good

candidate antigen for vaccination against S. neurona infection.

134

In spite of the limited set of sequences analyzed in the present study, this
analysis provides statistically significant support for the existence of sites in SA G] of
S. neurona under diversifying selection pressure and presents SA G] as an example of
adaptive evolution. These findings have major implications for the design and
interpretation of population genetic studies of selection on surface antigen genes of S.
neurona and other coccidia.

Finally, this approach should be used in further analyses of SA G1 sequences
from multiple isolates from different geographic locations, which will increase the
probability of the identification of more divergent alleles in order to precisely infer
the history and processes involved in the evolution of the observed polymorphism in

SAG].

135

REFERENCES

Akashi H (1995) Inferring weak selection from patterns of polymorphism and divergence
at “silent” sites in Drosophila DNA. Genetics 139:1067—1076

Bentz BG, Ealey KA, Morrow J, Claypool PL, Saliki JT (2003) Seroprevalence of
antibodies to Sarcocystis neurona in equids residing in Oklahoma. J Vet Diagn Invest
62597—600

Bonenfant C, Dimier-Poisson I, Velge-Roussel F, Buzoni-Gatel D, Del Giudice G,
Rappuoli R, Bout D (2001) Intranasal immunization with SAGl and nontoxic mutant
heat-labile enterotoxins protects mice against T oxoplasma gondii. Infect Immun
69:1605—1612

Conway DJ, Polley SD (2002) Measuring immune selection. Parasitology 125
[Suppl]:S3—Sl6

Dubey JP, Lindsay DS, Saville WJ, Reed SM, Granstrom DE, Speer CA (2001) A review
of Sarcocystis neurona and equine protozoal myeloencephalitis (EPM). Vet Parasitol
95289—131

Ellison SP, Omara-Opyene AL, Yowell CA, Marsh AB, Dame JB (2002) Molecular
characterisation of a major 29 kDa surface antigen of Sarcocystis neurona. Int J Parasitol
32:217—225

Felsenstein J (1985) Confidence limits on phylogenies: an approach using the bootstrap.
Evolution 39:783—791

Fu YX, Li WH (1993) Statistical tests of neutrality of mutations. Genetics 133:693—709

Hyun C, Gupta GD, Marsh AE (2003) Sequence comparison of Sarcocystis neurona
surface antigen from multiple isolates. Vet Parasitol 112211—20

Huelsenbeck JP, Ronquist F (2001) MRBAYES, Bayesian inference of phylogenetic
trees. Bioinformatics 17: 754—755

Kumar S, Tamura K, Nei M (1994) MEGA: Molecular Evolutionary Genetics Analysis
software for microcomputers. Comput Appl Biosci 10:189—191

Letscher—Bru V, Pfaff AW, Abou-Bacar A, Filisetti D, Antoni E, Villard O, Klein JP,
Candolfi E (2003) Vaccination with T oxoplasma gondii SAG-l protein is protective
against congenital toxoplasmosis in BALB/c mice but not in CBA/J mice. Infect Immun
71 :6615—6619

MacKay RJ, Granstrom DE, Saville WJ, Reed SM (2000) Equine protozoal
myeloencephalitis. Vet Clin North Am Equine Pract 16: 405—425

136

Mansfield LS, Schott HC2nd, Murphy AJ, Rossano MG, Tanhauser SM, Patterson J S,
Nelson K, Ewart SL, Marteniuk JV, Bowman DD, Kaneene JB (2001) Comparison of
Sarcocystis neurona isolates derived from horse neural tissue. Vet Parasitol 95:167—178

Marsh AE, Johnson PJ, Ramos-Vara J, Johnson GC (2001) Characterization of a

Sarcocystis neurona isolate from a Missouri horse with equine protozoal
myeloencephalitis. Vet Parasitol 95 : 143—1 54

Marsh AE, Hyun C, Barr BC, Tindall R, Lakritz J (2002) Characterization of monoclonal
antibodies developed against Sarcocystis neurona. Parasitol Res 882501—506

Nei M, Gojobori T (1986) Simple methods for estimating the numbers of synonymous
and nonsynonymous nucleotide substitutions. Mol Biol Evol 3:418—426

Ohta T (1991) Cirumsporozoite protein genes of malaria parasites (Plasmodium spp.):
evidence for positive selection on immunogenic regions. Genetics 127:345—353

Ohta T (1993) The nearly neutral theory of molecular evolution. Annu Rev Ecol Syst 23:
263—286

Rossano MG, Kaneene JB, Marteniuk JV, Banks BD, Schott HC2nd, Mansfield LS
(2003) A herd-level analysis of risk factors for antibodies to Sarcocystis neurona in
Michigan equids. Prev Vet Med 57:7—13

Rozas J, Rozas R (1999) DnaSP version 3: an integrated program for molecular
population genetics and molecular evolution analysis. Bioinformatics 15:174—175

Simonsen KL, Churchill GA, Aquadro CF (1995) Properties of statistical tests of
neutrality for DNA polymorphism data. Genetics 141: 413—429

Swofford DL (2002) PAUP*. Phylogenetic Analysis Using Parsimony (* and Other
Methods). Version 4. Sinauer, Sunderland, Mass.

Tajima F (1989) Statistical method for testing the neutral mutation hypothesis by DNA
polymorphism. Genetics 123:585-595

Thompson JD, Gibson TJ, Plewniak F, J eanmougin F, Higgins DG (1997) The ClustalX

windows interface: flexible strategies for multiple sequence alignment aided by quality
analysis tools. Nucleic Acids Res 25:4876—4882

137

CHAPTER 5
PHYLOGENETIC CONGRUENCE OF SARCOCYSTIS NE URONA ISOLATES IN
THE UNITED STATES BASED ON SEQUENCE ANALYSIS AND

RESTRICTION FRAGMENT LENGTH POLYMORPHISM

Submitted for publication as original paper:

Hany M. Elsheikha 1, Alice J. Murphy 2, and Linda S. Mansfield 1' 3 (2004) Phylogenetic
congruence of Sarcocystis neurona (Apicomplexa: Sarcocystidae) in the United States
based on sequence analysis and restriction fragment length polymorphism (RFLP).

Systematic Parasitology (In Press).

I Department of Large Animal Clinical Sciences, College of Veterinary Medicine,
Michigan State University, East Lansing, MI 48824, USA.

2 Diagnostic Center for Population and Animal Health, Michigan State University, East
Lansing, MI 48824, USA.

3 Department of Microbiology and Molecular Genetics, College of Veterinary Medicine,

Michigan State University, East Lansing, MI 48824, USA.

138

ABSTRACT

The objectives of the present study were to assess the genetic diversity,
phylogeny, and phylogeographic relationships of S. neurona isolates from different
localities in the United States. DNA was extracted from 13 isolates from different hosts.
All 13 isolates were subjected to PCR-RFLP analyses using two DNA markers (25/396
and 33/54). Additionally, the 334 bp region of the 25/396 marker of all 13 Sarcocystis
isolates and Besnoitia darlingi, B. bennetti, T oxoplasma gondii, and Neospora caninum
was sequenced. Phylogenetic analysis was performed using neighbor-joining (NJ),
maximum parsimony (MP), and minimum evolution (ME) methods based on the
sequences of the 25/396 marker of the 13 Sarcocystis isolates obtained in this study and
sequences of 10 related isolates from GenBank. Phylogenetic trees revealed a close
relatedness among S. neurona isolates in the US (nucleotide sequence diversity <5.0%).
US isolates formed a monophyletic group and appeared more closely related to each
other than to the South American isolates, which formed a separate lineage. NJ and ME
trees with Kimura 2-parameter model separated S. neurona in two separate groups: a
northern US group and a southern US group. These findings suggest a correlation
between grouping of the isolates and geographic segregation and were consistent with a
genetic bottleneck hypothesis during opossum colonization of North America. Finally,
our DNA data do not support either the view of S. neurona as a single super-species or its

division into multiple sub-species.

139

INTRODUCTION
Sarcocystis neurona is the main causative agent of equine protozoal
myeloencephalitis (EPM) on the American continent. It infects the brain and spinal cord

of horses (Equus asinus), whereas sporocysts are formed in the intestine of the definitive
hosts, opossums of the genus Didelphis (Cutler et al., 2001; Dubey et al., 2001; Rosenthal
et al., 2001). S. neurona has been reported in various domestic and wild animal
intermediate hosts (Dubey et al., 2001). Despite the development of many therapeutic
agents and a vaccine against this pathogen (Kruttlin et al., 2001; Lindsay and Dubey,
2001; Franklin et al., 2003; Witonsky et al., 2004), EPM continues to be a common
neurological disease associated with significant morbidity in horses (Dubey et al., 2001).

Considering the serious economic impact caused by S. neurona infection in horses
in the United States, a clear understanding of the role of genetic polymorphism in the
epidemiology of this parasite is needed. S. neurona is a good candidate for phylogenetic
analysis based on DNA sequence data because of the limited number of phylogenetically
informative morphological characters. However, a limited number of genes have been
identified and a few phylogenetically informative DNA sequences of this organism have
been reported so far compared to other coccidia such as T oxoplasma gondii
(www.toxodb.org)

Molecular marker techniques are an essential and integral part of bio-systematic
and epidemiological studies. Tanhauser et al., (1999) described some RAPD-derived
markers, which can be used for identification and differentiation of S. neurona and S.
falcatula. Previous phylogenetic analyses of Sarcocystis species transmitted by opossums

in the New World has demonstrated the usefulness of these DNA markers to help

140

understand the relationship among closely related isolates of these organisms (Rosenthal
et al., 2001). To our knowledge, only two studies have focused on the sequence
polymorphism and phylogenetic relationships in a limited number of S. neurona isolates
(Rosenthal et al., 2001; Hyun et al., 2003). Additionally, information about the genetic
variability of this parasite and its correlation with ecoepidemiological distribution of
EPM disease is lacking.

The present study was conducted to test the hypothesis that S. neurona has a
clonal population structure and to extend our knowledge of the genetic diversity and
geographic distribution of S. neurona isolates from different hosts in the state of
Michigan in relation to other isolates from other regions in the United States of America
and South America. Sequences of the 25/396 marker of 13 Sarcocystis isolates
originating from 4 horses, 4 cowbirds (Molothrus ater), 3 Virginian opossums (D.
virginiana), 1 cat (F elis domesticus), and 1 raccoon (Procyon lotor) were obtained and
compared to similar sequences of other related isolates in GenBank from other regions in
the US and from South America. This analysis revealed homogeneity among all S.
neurona isolates from different hosts in the US particularly among isolates from
Michigan. A unique 2-bp insertion (TA) was observed in 10 out of 11 S. neurona isolates
originating from Michigan. Sequence data were concordant with restriction analyses of 2
previously described DNA markers from the Michigan isolates. The observed genetic
variability and the relationships among the isolates were the basis for investigating
whether the genetic differences in S. neurona isolates reflected distinct eco-
epidemiological features of S. neurona infection, thus bringing to light potential

fundamental information for future control programs of EPM infection in horses.

141

MATERIALS AND METHODS

Clinical and demographic details of the isolates. Isolates (n = 23) were used in
the sequence and phylogenetic analyses and the sources of origin are listed in Table 1.
Twenty-two Sarcocystis isolates were used as ingroup taxa. This included 18 Sarcocystis
isolates (15 S. neurona and 3 S. falcatula) from different host species from 5 different
states in the USA: Michigan, Ohio, Virginia, Florida, Mississippi, and California and 4
Sarcocystis isolates (2 S. neurona and 2 S. falcatula) from the South American opossum
(D. albiventris). One Sarcocystis spp. isolated from a bird from South America was
included in the analyses as an outgroup to root the phylogenetic tree (8). Representative
taxa of other cyst-forming coccidia such as Besnoitia darlingi, B. bennetti, T oxoplasma
gondii, and Neospora caninum were also included and their sequences were compared to
sequences of Sarcocystis isolates. B. bennetti bradyzoites were obtained from skin lesions
containing numerous cysts from a donkey (Elsheikha et al., 2004b). B. darlingi was
obtained from cysts isolated from opossum (D. virginiana) from the mid-Michigan area
(Elsheikha et al., 20040). T. gondii tachyzoites were obtained by removing them from
glass slides (Toxo-ImmunoFA slides) that are commercially used in immuno diagnostic
assay (GenBio, San Diego, CA). S. neurona isolates from a raccoon from Ohio and from
a cat fed with sporocysts from opossum from Virginia were provided by Dr. JP Dubey
(U SDA, Animal Parasite Diseases Laboratory, Beltsville, Maryland). All S. neurona
isolates were maintained in tissue cultures using previously described procedures
(Mansfield et al., 2001). N. caninum NC-l strain was kindly provided by Dr. David S.

Lindsay (University of Maryland, USA).

142

 

38332... 83 :83 ..a a 35583 :53 592332 33283 $2533: .3 2 28a: .3
33333353. Now 2va ...3 8 35533. Sow «ESE 333.3825 33:253.: .Q N. _ _ 3833: .3.
33338.».4. :3 388 .._3 8 32.9.3 Sow oEO 3.3.3883 5088 33-633380 EooooEEODm—z 3:88: .3.
«83333: :3 88313.5 .83 253.5 33283 .8 33.6388 332.5232 228.. .3
39.33382 83 n ._3 8 22.3352 333— 533222 333.3803 €33 .35 25300 chEODmE £53353.
amvmnmc>< mmm a ._3 8 22.3352 333— 533522 333308035 363 .3: 95300 mmUEODmZ 253.3: .3.
anwwnmc>< $3. a .138 23:35.2 33— cmeozz 333330.535 333 $5 933.50 vaEODmE 3:93»: .3.
30V33~c>< Sm a .E B 223ng 3333 533532 mumxuocoam A83 .3; 25300 mmUEODmZ 36.33: .3.
«9.33335. 3.. a .3 8 23223 383 53222 35283 25533.: .3 3330:). 22%: .3
«3.33382 333 2833 ..m s 223222 33.3. 53222 $3283 2253...; .Q EOE. 28%: .3
3933.854. 333 :83 ..3 “o 22.33533. 83. 533222 mamxoeoam 25333.33 .Q 50:2 56.33: .3.
3N¢33Nc>< 333. :83 .3 8 23:35.2 Son 233222 3863832 A3358 ..3 38: 22:2 3058: .3.
33333.5.3 mmm 28$ ..3 8 22.3532 ooom 59:22 3263202 3:53 .3 038: SE: 26.53: .3.
«ovwn~c>< N33 :88 ._3 8 Seaman—2 333— 533522 3263232 A3353 ..3 85: 33:2 38.53: .3.
nomwhmc>< mmm COOS ..3 3 20:35.2 333— 5.3522 32.53982 33536 ..3 03.5: .52 38.33: .3.
9.23-:—
.e= :23339< 35 53>

25550 53:9.— uuctouom 5:29.: 3:39.— 335 33:: cozaaufiua 33.8—

 

.3353 :8on 2: E 333353 328% 3.3.3.0833. 2: no mzflom .3 03.3.3.

143

$33282. .3222 2.2.2283 223233.... 22.22 .3 £332.85... 2228

2.823382 ANmmR©><V .5363. 25323883 3333 3:: E £3335 822.com E 33: 033 083 3822.33 mama/0:8 25.
222.22: a

.383 :8on 05 E 358% moocozcom a

 

2282... EN 282.22 .2233. .8m .22.. 3.28.2.3 22222 .2 8. 2m 23.22823

955-50
3332“? N8 :83 .22 .2283. 58 2.222 5.8.2.3 2.22.222 .Q :23 2.28.2 .3
8833"? N8 :83 ...... 2 .2283. 58. .225 32522.3 2.22.222 .2 .3 222823
8232... 22 283.22 .2233. 8.2. ...Em 23283 22222 .Q 2 “2.8. 22.22 .3
832.2 SN :8. .22 .2233. 58 ...Em 2322.3 2.22.222 .2 m 22.8. 2822 .3
8:83... 3% .82....2222252 58 2.2.... 23 28:2 2.2 2.2238 2.283 .3
33:83... 323 88: ...2382252 .88 2.22.8 28 .223 22.5. 3.8.2: .-ao: 28.22 .3
3833:. SN 283.22.22.58”. :8 2.2: 23 8.2.2 .22 8222.83 3.0-23 282... .3

 

.8 .208 _ 2...;

144

DNA extraction, PCR amplification, and sequencing. Total genomic DNA
was extracted from culture-derived merozoites, tachyzoites, bradyzoites, or sporocysts
using the DNeasy Tissue kit (Qiagen, Valencia, CA) according to the manufacturer’s
instructions with one modification. Sporocysts were subjected to at least 6 alternating
cycles of freeze-thawing in liquid nitrogen and warm water just prior to the DNA
extraction. Quantity and purity of the extracted DNA was evaluated on a NanoDrop DN-
1000 Spectrophotometer and NanoDrop software 2.5.4. (NanoDrop Technologies,
Rockland, DE, USA). DNA was then used as template for 2 polymerase chain reactions
(PCR). In one PCR, the JNB25 (S’- CACACA AA CA CTGAAAGTCACGTACTT) and
J D396 (5’- CCTGCCTCACTTCGACACAT) primers, which amplify 334 bp of a
random DNA marker 25/396 specific for S. neurona and S. falcatula were used. In a
second PCR the JNB33 (5’-CGAACAGAGATGAGGAAAAT) and JNBS4 (5’-
GTTGTGGTGTTGCGTGAGTC) primer pair, which amplifies 1100 bp marker region
from the first internal transcribed spacer region (ITS-1) were used (Tanhauser et al.,
1999). PCRs were performed using conditions described previously for the 2 primer sets
(Tanhauser et al., 1999). Nested PCR was performed on all specimens to amplify
sufficient amount of DNA fragments from each of the two genetic markers for further
analyses using the same primer sets with 4 pl of the first PCR products as a template for
the nested (second) PCR. The nested PCR was used to increase sensitivity of the reaction
and to conserve the limited amounts of samples because the initial PCR amplification
attempts produced little product from many of the templates.

The 334 bp and 1100 bp amplicons were purified using spin columns (Qiagen).

The purified 334 bp PCR products obtained from 13 S. neurona isolates from Michigan,

145

Ohio, and Virginia besides the following isolates; B. darlingi, B. bennetti, T. gondii, and
N. caninum were sequenced. The nucleotide sequences of the amplicons were determined
directly by using the Bi gDye termination cycle sequencing technique (Applied
Biosystems Inc., Foster City, CA) according to the manufacturer’s instructions.
Sequences were analyzed with a 3100 GenAmp automatic sequencer (Applied
Biosystems) at the Michigan State University Genomic Technologies support facility.
Sequences were obtained for both strands of each DNA amplicon for confirmation of the
nucleotide sequence.

Phylogenetic Analysis. Phylogenetic analysis was based on sequences of the
25/396 MPD marker obtained fi'om 13 isolates in addition to sequences of 10
Sarcocystis isolates obtained from GenBank (Table 1). Sequences of 22 Sarcocystis
isolates were used as the ingroup. Sequence of 1 Sarcocystis sp. isolate was used as an
outgroup to root the phylogenetic tree (3). Editing and alignment of the nucleotide
sequence data were performed with ClustalX software Version 1.18 for Windows
(Thompson et al. 1997). The computer program Seqweb (Version 9.1, Genetic Computer
Group, Madison, Wis., USA) was used to generate a multiple nucleotide sequence
alignment and to detect any discrepancies between the 2 strands of each PCR product by
careful examination of the positive- and negative-sequence strands. The prototype
sequence (GenBank accession no. AF 093158) of Tanhauser et al., (1999) was used as a
reference sequence.

Phylogenetic analysis was carried out on the multiple sequence alignment from 23
isolates by using the neighbor-joining (NJ) method (Saitou and Nei, 1987) using the

Kimura-2 parameter (K-2P) model or the General Time Reversible (GTR) model of

146

sequence evolution as implemented in the computer program PAUP software Version
4.0b4a for Macintosh (Swofford, 2002). Bootstrap support (Felsenstein, 1985) for the
results of the NJ analysis was based on 1000 pseudoreplicate datasets generated from the
original multiple sequence alignment. Maximum parsimony (MP) analysis of the aligned
sequences was also conducted by PAUP. MP analysis included unordered and
unweighted character states. MP trees were constructed by running the heuristic search
algorithm with tree-bisection reconnection (TBR) branch swapping, the MULPARS
option, ACCTRAN default, and 100 random additions of sequence replications. Single
gaps and gaps that aligned unambiguously forming large indels were scored as “missing
data”. All uninformative characters were excluded. The relative strength of each node
within the phylogeny was estimated by using non-parametric bootstrap analyses
(Felsenstein, 1985) with 1000 pseudoreplicates. A minimum evolution (ME) tree was
also constructed from the aligned sequences, using the K-2P method for estimating
pairwise genetic distances as implemented in MEGA software Version 2.1 for Windows
(Kumar et al., 2001). An initial tree was obtained using the NJ method, and the minimum
evolution method was used to search for the tree, which minimizes the sums of the
branch length estimates, by branch swapping using closest neighbor interchange
(Takahashi and Nei, 2000). Statistical support for the tree nodes was evaluated by
examining their percentage of recovery in 1,000 resampled trees using the bootstrap test
(Sitnikova et al., 1995).

Restriction digestion analyses. Thirteen S. neurona isolates obtained from
Michigan, Ohio, and Virginia were analyzed by restriction fragment length

polymorphism (RF LP) of the amplified and purified PCR products. The 334-bp PCR

147

products obtained by the primers JNB25 and JD396 were digested independently with the
restriction endonuclease enzymes HindIII and Hinfl(1nvitrogen Life Technologies,
Carlsbad, CA), and the 1100 bp amplicons obtained by the primers JNB33 and INBS4
were digested separately with Hian and DraI, as described by Tanhauser et al., (1999).
Ten pl of each PCR product and restriction digestion product were electrOphoresed on a
1.8% agarose gel containing a 0.5-ug/ml ethidium bromide for 1 h at a constant voltage
(100 V). The 100 bp DNA ladder (New England Biolabs, Beverly, MA) was included as
a molecular size marker. The gels were visualized on an UV transilluminator (Multilmage
Light Cabinet). All gel images were captured with a charge-coupled device camera to the
Alphalmager computer program (AlphaImager gel documentation and Analysis system;
Alpha Innotech Corp., San Leandro, CA). Also, images of the gels were printed onto
thermal imaging paper (UPP-l 10S; Sony) by using Sony digital graphic printer for
Windows Version 1.1 (UP-D890) for visual evaluation of the RFLP patterns for detection
of restricted fragment bands characteristic of S. neurona or S. falcatula. Bands in each
lane were identified by their molecular size compared to those of the bands in the
molecular size standard. Isolates were typed by visual assessment as S. neurona or S.
falcatula or a combination of both species based on the presence of bands characteristic

for each species following restriction digestion with the above enzymes.

148

RESULTS

Epidemiological characteristics of the Sarcocystis isolates. The relevant
characteristics of the 23 Sarcocystis isolates included in this study, including isolation
date, genotype, source, and geographic origin are presented in Table 1. Among 13
isolates collected in this study from Michigan, Ohio, and Virginia, 12 isolates were
identified as S. neurona. Ten isolates were from Michigan (4 were from horses that died
with confirmed EPM, 3 were from cowbird-fed opossums, and 3 were from opossums), 1
was from a raccoon from Ohio, and 1 was originally from a cat infected with opossum
sporocysts from Virginia. One S. falcatula isolate was identified from a cowbird from
Michigan. Sequence (MIH12) obtained from sarcocysts from a horse from Michigan has
raised questions concerning the authenticity of horses as intermediate host. However, the
sequence determined for the 25/396 marker for this case was identical to that of three
isolates (MIHl, MIH7, MIHlO) from confirmed EPM horses in the same locality,
suggesting that this isolate indeed represents a true S. neurona isolate. Additionally,
sequences were obtained from GenBank for 5 previously characterized isolates
representing the Southern US as follow; 4 S. neurona isolates (1 opossum from Florida, 1
opossum from Mississippi, 1 sea otter (Enhydra lutris nereis Linnaeus) from Florida, and
1 horse from California) and 1 S. falcatula isolate from a cowbird from Florida. The
frequency and distribution of S. neurona isolates in the US by host and by state is
illustrated in Figure 1. Finally, sequences of 2 S. neurona and 2 S. falcatula isolates
obtained from South American opossums and the sequence of 1 Sarcocystis sp. isolate
obtained from a bird from South America available in GenBank were also included in the

analyses.

149

4 horses 4 COWbil‘dS
3 opossums

   

l opossum

1 horse 1 sea otter

l cowbird

Figure 1. Map of the United States illustrating the number and distribution of S. neurona
isolates analyzed from different animal hosts in different states in the present study.

150

Molecular analyses. Two approaches were used in order to assess genotypic
variability in S. neurona. 1) Sequencing and phylogenetic analyses of the 25/396 genetic
marker and 2) RF LP restriction analysis of the 25/396 and 33/54 markers of 13
Sarcocystis isolates (10 from Michigan, 1 from Ohio, and 1 from Virginia) for which we
had available genomic DNA. DNA sequences of the 25/396 marker were obtained from
13 Sarcocystis isolates from Michigan (4 horses, 4 cowbird, 3 opossums), 1 isolate
originally from a raccoon from Ohio, and 1 isolate originally from a cat infected with
sporocysts from an opossum from Virginia. Also, sequence of the 25/396 marker were
obtained from other coccidia such as B. darlingi, B. bennetti, T. gondii, and N. caninum
and compared to sequences of Sarcocystis isolates. These newly reported sequences
were submitted to the GenBank database and have been assigned accession numbers
(Table 1). Finally, 10 sequences of the 25/396 marker were retrieved from GenBank and
were included in the sequence and phylogenetic analysis. Sequence and RFLP results
obtained in this study are summarized in Figures 2-7.

Characterization of sequence polymorphism. The total number of base pairs
sequenced from each isolate is shown in Table 1. A sequence alignment of the
Sarcocystis isolates and sequences of other coccidia (B. darlingi, B. bennetti, T. gondii,
and N. caninum) are shown in Figure 2. The sequence alignment of the 25/396 DNA
fragment from 27 isolates of different coccidian genera demonstrated that the region
amplified by JNB25 and JD396 primers was highly conserved and therefore suitable for
inference of genetic relatedness among Sarcocystis isolates. The sequence of S. neurona
(UCD1 isolate) was selected as the prototype, and all sequences were compared with this

sequence (Figure 2). Based on nucleotide base differences, sequences could be divided

151

into 3 groups. The sequence conservation noted for S. neurona isolates within each of
these groups was remarkable with 299% nucleotide similarity. Group I contained 12 S.
neurona isolates from Michigan, Ohio, and Virginia. Nucleotide similarity revealed that 8
ofthe Michigan isolates; MIHl, MIH7, MIHlO, MIHIZ, MIOP7, MIOP20,
MSUOP/CB4, and MSUOP/CBS, and the Virginia isolate; MSUOP/cat2 shared 100%
sequence similarity to each other. The Ohio isolate; MSUOP/raccoonl and a cowbird
isolate; MSUOP/CB3 from Michigan had 99.5% sequence homology to this group
because they differed only in 1 nucleotide (Figure 2). One isolate from Michigan
opossum; MIOPl had 99% sequence homology to that group because it had 2 different
nucleotide bases. Group 11 included 4 S. neurona isolates from Florida, Mississippi, and
California. This group had 100% sequence similarity within the group. Group 111
included 2 S. neurona isolates from South America and had within group 100% sequence
homology. These findings suggest the existence of three distinct groups of S. neurona
isolates (Northern US, Southern US, and South America), with a high degree of
homology within each group and lesser homology between groups.

Similarly, S. falcatula showed geographic structuring where 2 S. falcatula isolates
from cowbird [1 from Florida and 1 from Michigan, (MSUOP/CB6)] had 100% sequence
homology to each other. Two S. falcatula isolates from South American opossums also
shared 100% sequence homology to each other. Ten S. neurona isolates from Michigan
but 1 S. falcatula MSUOP/CB6 isolate obtained from a cowbird had a unique 2-bp
insertion (TA) at nucleotide positions 270 and 271. This unique feature was absent in all

other isolates. Of interest is that N. caninum also had this unique insertion.

152

OOOOOOOOOOOOOOOOOOOOOOOOOOOOOOOOOOOOOOOOOOOOOOOOOOOOOOOOOOOOOOOOOOOOOOOOOOOOOOOOOOOOOOOOOOOOOOOOOOOO

OOOOOOOOOOOOOOOOOOOOOOOOOOOOOOOOOOOOOOOOOOOOOOOOOOOOOOOOOOOOOOOOOOOOOOOOOOOOOOOOOOOOOOOOOOOOOOOOOOOO

OOOOOOOOOOOOOOOOOOOOOOOOOOOOOOOOOOOOOOOOOOOOOOOOOOOOOOOOOOOOOOOOOOOOOOOOOOOOOOOOOOOOOOOOOOOOOOOOOOOO
OOOOOOOOOOOOOOOOOOOOOOOOOOOOOOOOOOOOOOOOOOOOOOOOOOOOOOOOOOOOOOOOOOOOOOOOOOOOOOOOOOOOOOOOOOOOOOOOOOOO
OOOOOOOOOOOOOOOOOOOOOOOOOOOOOOOOOOOOOOOOOOOOOOOOOOOOOOOOOOOOOOOOOOOOOOOOOOOOOOOOOOOOOOOOOOOOOOOOOOOO

OOOOOOOOOOOOOOOOOOOOOOOOOOOOOOOOOOOOOOOOOOOOOOOOOOOOOOOOOOOOOOOOOOOOOOOOOOOOOOOOOOOOOOOOOOOOOOOOOOOO

¢UOBOthdwgfioéwwgdeogumudOOBoowomwhmu/wwdwduwIIdHUBUdUdDBUUOUGOUBUUOBUBHdOUUU§0<U¢BUUUUUBDOH¢UBDBBBO

:m w em
a m

..méén .m
23223 .M
.5328 K
825m
_-<n_m

mm mzmxooemm
h 838—

M 0:28—
_-._.O-Zm

E 820$
N: 838—
N80E03m2
_ cocoofiEODmE
cmOEODmE
£58~u\ ..m.
$3328 .2
#530292
vaEODm—z
mmUEODmE
SEE
o:.=2
2.52

3.32
ommOHE
30:2
30:2
.003

dogs—cg 38.3»: a. Emmi—0:2 no 2 mo So 2 E

28 53558 6.88.82 E CR 25 chm 30:88 «a <.C 883 3:85 m we oocomoa 05 20 Z .9505 8a 8:268 2:88.:— 2: mo 89: 80¢
“:20me 8a 65 v.83 can .5qu .3 @8365 0.3 883 wEmflE £859 .3 @8865 8a 8:258 090905 05 mo 80.: 8 .8352 2a 85
88am ._&_8_8_2 use «ESE Eat 882880 80¢ 358:0 885238 8553 mo gym—.2 05 8 3802?: 29$ moocoscom .823 53288 new
dam .3389an 5:8 .23 @8888 886$ S53»: mtgoeuuum no 885358 e338 canm 05 mo “co—SEE 8:268 02820: Z .N unsur—

153

o o

o o

e c

oooooooooooooooooooooooooooooooooooooooooooooooooooooooooooooooooooooooooooooooooooooooooooooooooooo

oooooooooooooooooooooooooooooooooooooooooooooooooooooooooooooooooooooooooooooooooooooooooooooooooooo

oooooooooooooooooooooooooooooooooooooooooooooooooooooooooooooooooooooooooooooooooooooooooooooooooooo

oooooooooooooooooooooooooooooooooooooooooooooooooooooooooooooooooooooooooooooooooooooooooooooooooooo

oooooooooooooooooooooooooooooooooooooooooooooooooooooooooooooooooooooooooooooooooooooooooooooooooooo

....................................................................................................

oooooooooooooooooooooooooooooooooooooooooooooooooooooooooooooooooooooooooooooooooooooooooooooooooooo

oooooooooooooooooooooooooooooooooooooooooooooooooooooooooooooooooooooooooooooooooooooooooooooooooooo

oooooooooooooooooooooooooooooooooooooooooooooooooooooooooooooooooooooooooooooooooooooooooooooooooooo

oooooooooooooooooooooooooooooooooooooooooooooooooooooooooooooooooooooooooooooooooooooooooooooooooo

oooooooooooooooooooooooooooooooooooooooooooooooooooooooooooooooooooooooooooooooooooooooooooooooooo

oooooooooooooooooooooooooooooooooooooooooooooooooooooooooooooooooooooooooooooooooooooooooooooooooooo

oooooooooooooooooooooooooooooooooooooooooooooooooooooooooooooooooooooooooooooooooooooooooooooooooooo

oooooooooooooooooooooooooooooooooooooooooooooooooooooooooooooooooooooooooooooooooooooooooooooooooo

dBQUBUBBBUBUUBBBdUdBOdUgOU UGdUUU0090U§0d0480§900¢04<980¢00UHUUUUdUdUOBéUHHUO<OO<UUUflUUUBBdUHBHUd

a;

wit.

852% .m
.5830: .m
.538 K
8:5:
T<mm

8 $2083:
5 8£0$

m 820$
790-7%

E 820$
N: 820$
90030332
2000008303m2
omUEODmE
550.02% .m.
8:550 .2
#:030sz
$525332
3530292
25:2
95:2
2:2

3:2
8:032
nm0=2
50:2
2003

.00 .2600 N 28E

154

Molecular Phylogenetics. In addition to the sequences obtained from the 13
isolates above, sequences of 10 related Sarcocystis isolates obtained from GenBank
(National Center for Biotechnology Information, Bethesda, Md., USA) were also
included in the analyses. Phylogenetic trees were generated from sequence data of 17 S.
neurona isolates and 2 S. falcatula isolates from different states in the USA and 2 S.
neurona and 2 S. falcatula isolates from South America as the ingroup, plus 1 non-
neurona Sarcocystis sp. as an outgroup, by neighbor-joining (NJ), maximum parsimony
(MP) and minimum evolution (ME) analysis with node support values generated by 1000
bootstrap replications (Figure 3-6). All tree—building methods yielded similar tree
topologies with some differences. The NJ tree with General Time Reversible (GTR)
model and the NIP tree showed 2 main lineages: Lineage 1 included all S. neurona
isolates from the different states of the US including isolates from Michigan, Ohio,
Virginia, Florida, Mississippi, and California from horses, opossums, cowbirds, cat,
raccoon, and sea otter. Lineage 2 included S. neurona isolates from South America only.
The NJ and ME trees with K-2P model of sequence evolution showed clustering of all S.
neurona isolates in one monophyletic group with bootstrap values of 65% for NJ and
69% for ME. Within this group, 4 subgroups of S. neurona isolates were identified;
Group (I) included 2 S. neurona isolates (1 from an opossum and 1 from a cowbird) from
Michigan, Group (II) included 10 S. neurona isolates: 8 isolates fi‘om Michigan (4 from
horses, 2 from cowbirds, and 2 from opossums), l isolate from a raccoon from Ohio, and
1 isolate from a cat from Virginia, Group (III) encompassed 4 S. neurona isolates: 1 from
a horse from California, 1 from an opossum from Mississippi, and 2 from an opossum

and a sea otter from Florida, and Group (VI) contained 2 S. neurona isolates from South

155

American opossums. The tree topology showed that S. neurona isolates from the South
American opossums (D. albiventris) were always distinct from US isolates and that two
distinct lineages were identified within the US isolates. These were designated as
Northern-United States (N-US) and Southern United States (S-US). N-US encompassed
isolates from Michigan and Ohio and S-US included isolates from Florida, and
Mississippi. Based on the sequence similarities, the Virginia isolate was grouped with
the northern US lineage and California isolate was grouped with the Southern US lineage.
Of interest is that S. falcatula isolates from each continent also were grouped together.
For instance, 2 S. falcatula isolates from a cowbird from each of Michigan and Florida
clustered together, forming a monophyletic group with 100% support value. Two S.
falcatula isolates from South American opossums also grouped together in one group

with 100% bootstrap support.

156

l S. falcatula, MSUOP/CB6, Ml

 

100 I I S. falcatula, FL

 

 

I l S. falcatula, SFA-l, Argentina

 

I Sarcocystis sp., Brazil
S. neurona, MlOPl, M1

64
"I s. neurona, MSUOP/CB3, Ml

S. neurona, MSUOP/CB4, Ml

S. neurona, MSUOP/CBS, Ml

S. neurona, MSUOP/raccoonl, 0H
S. neurona, MIOP7,MI

S. neurona, MIOP7, M]

S. neurona, MlHlo, MI

S. neurona, SN-OT-l, FL

 

S. neurona, 16, MS

79 S. neurona, 3, Brazil

S. neurona, 7, Brazil

S. neurona, MlHl, Ml
S. neurona, UCD-l, CA
S. neurona, MIOP20, MI

S. neurona, MSUOP/catz, VA

— S. neurona, MIHIZ, MI

 

 

S. neurona, 1112, FL

L—Sarcocystis sp., Bird100, Brazil

— 0.05 substitutions/site

Figure 3. Neighbor-joining (NJ) phylogenetic tree of Sarcocystis neurona isolates from
various hosts and different localities in the US inferred from sequences of the 25/239
marker using General Time Reversible (GTR) model. One Sarcocystis sp. isolate
(GenBank Accession no. AF 323943) was used as outgroup. Bootstrap confidence values
were based on 1,000 replicates; only those values that exceed 60% are shown above the
node. Branches without bootstrap values represent polytomies that could not be resolved
because the sequences were almost identical. The scale bar represents the calculated
distance value. Branch lengths are drawn proportional to the amount of genetic change.
The isolate designation and locality follows taxon. For details of isolate designation

please refer to Table 1.

157

 

100 ‘ s.fa1cam1a, MSUOP/CB6, MI
I s. falcatula, FL

63 . S. neurona, MIOP], MI
“'I, s. neurona, MSUOP/CB3, M1

S. neurona, MIOP7, M]

S. neurona, MIOP7, MI

S. neurona, M11112, Ml

S. neurona, MIOP20, MI

S. neurona, MSUOP/CB4, Ml
S. neurona, MSUOP/CBS, MI

78 S. neurona, M11110, MI

 

S. neurona, M1111, M]

S. neurona, 16, MS

S. neurona, 1112, FL

S. neurona, MSUOP/catz, VA

S. neurona, MSUOP/raccoonl, OH
S. neurona, UCD-l, CA

S. neurona, SN-OT-l, FL
. S. neurona, 3, Brazil
I. S. neurona, 7, Brazil

96 I S. falcatula, SFA-l, Argentina

 

 

 

I. Sarcocystis sp., Brazil

 

_ Sarcocystis sp., Bird100, Brazil

-— 1 change

Figure 4. Maximum parsimony (MP) phylogenetic tree of Sarcocystis neurona isolates
from various hosts and different localities in the US inferred from sequences of the
25/239 marker.

158

89 r‘ S. falcatula, MSUOP/C B6, Ml

 

L S. falcatula, FL

ELI—S. neurona, MlOPl, Ml
S. neurona, MSUOP/CB3, Ml
PS. neurona, M1117, M1

n S. neurona, MIOP7, Ml

- S. neurona, M11112, M1

I-S. neurona, MIOP20, Ml

. S. neurona, MSUOP/CBS, Ml
' S. neurona, M [1110, M1

F. - S. neurona, M1111, Ml

-S. neurona, MSUOP/catz, VA

— S. neurona, MSUOP/raccoonl, OH

 

- S. neurona, MSUOP/CB4, MI

65 -— S. neurona, SN-OT-l, FL

 

h S. neurona, 16, MS

- S. neurona, 1112, FL

85 r S. neurona, 3, Brazil

 

]- S. neurona, 7, Brazil

 

' S. neurona, UCD-l, CA

 

92 I S. falcatula, SFA-l, Argentina

 

lSarcocystis sp., Brazil

 

 

Sarcocystis sp., Bird100, Brazil

— 0.005 substitutions/site

Figure 5. Neighbor-joining (NJ) phylogenetic tree of Sarcocystis neurona isolates from
various hosts and different localities in the US inferred from sequences of the 25/239
marker using Kimura two-parameter distance model.

159

66 S. neurona, MlOPl, Ml

S. neurona, MSUOP/CBB, Ml

S. neurona, MSUOP/raccoonl, OH
S. neurona, MIH7, M1
S. neurona, MIOP7, MI
S. neurona, MSUOP/CB4, M1
S. neurona, MSUOP/CBS, Ml
S. neurona, M11110, MI
S. neurona, M1111, MI
S. neurona, MIOP20, Ml

S. neurona , M11112, Ml

 

S. neurona, MSUOP/catZ, VA

88 IS. neurona, 3, Brazil

 

IS. neurona, 7, Brazil

S. neurona, SN-OT-l, FL

S. neurona, 16, MS

69
S. neurona, 1112, FL

 

S. neurona, UCD-l, CA

 

96 I Sarcocystis sp., Brazil

 

IS. falcatula, SFl, Argentina
100 I'— S. falcatula, MSUOP/CB6, MI
I S. falcatula, FL

Sarcocystis sp., Bird100, Brazil

 

0.005

Figure 6. Phylogenetic tree of Sarcocystis neurona isolates obtained by minimum

evolution (ME) method.

160

PCR-RFLP analyses. To supplement and confirm the sequencing analyses,
PCR-RFLP analysis of the 334 bp and 1100 bp amplicons was used to characterize the
isolates where live material was available. We analyzed 11 isolates from Michigan (4
horses, 4 cowbirds, and 3 opossums), 1 isolate from a raccoon from Ohio, and l isolate
from a cat from Virginia. Two restriction fragment patterns of genotypic distribution
were found in these Sarcocystis isolates. Pattern (I) included isolates obtained from
horses, opossums, raccoon, and cat, which produced one indistinguishable pattern
consistent with S. neurona. Pattern (11) included isolates obtained from cowbird isolates,
which gave restriction profiles indicative of a mixture of S. neurona and S. falcatula.
Restriction digestion of the PCR products of the 25/396 marker with Hian and HindIII
enzymes exhibited a restriction profile compatible with S. neurona for MIHl, MIH7,
MII-I 10, M11112, MIOP], MIOP7, MIOP20, MSUOP/raccoonl, and MSUOP/cat2, where
334bp amplicons were not digested with HindIII but yielded 180bp and 154bp fragments
with Hinfl (Figure 7A). Isolates obtained from cowbirds from Michigan MSUOP/CB3,
MSUOP/CB4, MSUOP/CBS, and MSUOP/CB6 gave very weak 334bp undigested bands
with HindIII similar to S. neurona besides another stronger band similar to that of S.
falcatula. With Hinfl enzymes cowbird isolates gave RFLP profiles consistent with S.
neurona. Similar findings were obtained with restriction digestion of the 1100 bp PCR
products with Hinfl and DraI (Figure 78). All isolates gave restriction profiles consistent
with S. neurona whereas cowbird isolates gave restriction profiles combining S. neurona
and S. falcatula. PCR-RFLP analysis of the 334bp and 1100bp of representative isolates

are shown in Figure 7.

161

1100bp 1100bp

300bp . - '- --.-- - . 300bp

 

B M1 2 3 4 5 6 7 8 Y 910111213141516V1718192021222324M

  

. n
- '
1100bp “he"! won“. “mag”! "WP

. -
300bp , - 30°”

Figure 7. Diversity revealed by RFLP analyses of PCR products of representative
isolates of S. neurona from Michigan. (A) RFLP pattern obtained by restriction digestion
of the 25/396 marker (334 bp amplicons). Lanes 1 to 8, 334 bp undigested amplicons;
lanes 9 to 16, samples treated with Hinjl; lanes 17 to 24, samples treated with HindIII.
(B) RFLP pattern obtained by restriction digestion of the 33/54 marker (1 100 bp
amplicons). Lanes 1 to 8, 1100 bp undigested amplicons; lanes 9 to 16, Hinfl restriction
profile; lanes 17 to 24, DraI restriction profile. Lanes M, molecular size standards are
shown to the leftmost and rightmost of the figure in 100 bp-sized bands; (V) denote
empty lanes. Sources of isolates of the amplicons and restriction analyses: MIH7 (lanes
1, 9, and 17), MIHlO (lanes 2, 10, and 18), MIOP] (lanes 3, 11, and 19), MIOP7 (lanes 4,
12, and 20), MSUOP/CBS (lanes 5, 13, and 21), MSUOP/CB6 (lanes 6, l4, and 22),
MSUOP/cat2 (lanes 7, 15, and 23), MSUOP/Raccoonl (lanes 8, l6, and 24).

162

 

 

DISCUSSION

Thirteen years from its original identification and recognition (Dubey et al.,
1991), S. neurona remains the main causative agent of one of the most serious
neurological diseases in horses in the Americas (Dubey et al., 2001). Workers in the field
generally concur that this organism represents a biologically diverse group of strains
showing intraspecies variability that complicates their taxonomic classification as well as
clinical and epidemiological studies. Hence, a clear understanding of the nature and
extent of genetic variability present among S. neurona isolated from several host species
and from different geographical regions is essential to obtaining an understanding of the
natural history of this organism and the potential risks to horses.

Results in this study suggest that the current recognition of S. neurona as a
single species has to be reconceived. The present study showed a high probability for the
existence of two sub-species of S. neurona in the US and an additional subspecies of S.
neurona in South America. Phylogenetic tree (5) generated by NJ with the GTR model
and MP analyses showed a geographic structuring of the sequences of S. neurona isolates,
where trees showed well-supported 2 groups of S. neurona isolates; one group included
all isolates from the US and a second group included South American isolates. Within
the US, NJ with K-2P model and ME trees showed that isolates were firrther split
according to the geographic locality. Our results indicated that, on the basis of sequence
divergence, Michigan, Ohio, and Virginia isolates were most closely related to each
other, formed one sub- group designated as Northern US and were distantly related to
isolates from the southern US (Florida, Mississippi, and California), which clustered

together forming a separate sub- group. South American isolates also clustered together.

163

Sequences were very similar intra- group and less so inter-group. These findings provide
evidence for the presence of two closely related genetic variants of S. neurona existing
within the US, one in the Northern part and one in the Southern part. Therefore, this
work represents to our knowledge the first study able to discriminate between S. neurona
based on geographic origin of the isolate and the first to suggest that S. neurona isolates
present in each of these locations might have evolved independently.

Nucleotide sequence data in the present study revealed an absolute genetic
similarity among S. neurona isolates from Michigan (100% sequence homology), even
from different host species that lived in the same or close locality collected for over a 5
years period. This indicates an epidemiological relationship but less so compared to the
distantly related isolates from the Southern United States and South America. These
findings were supported by RFLP analysis of 334 bp and 1100 bp amplicons of the
25/396 and 33/54 DNA markers, respectively. The finding of indistinguishable
sequences in Michigan isolates that were isolated in the same area 5 years apart from the
same and different hosts suggests that these isolates shared a common ancestry. Shared
genotypes among epidemiologically linked isolates fi'om Michigan supports a hypothesis
of spread of parasite clones rather than frequent recombination events for this organism
in nature, at least at the local level. These findings also imply that the ancestor isolate
introduced into this particular locality and spread out to infect different hosts and the
isolate genotype remained stable for several years leading to such very high
homogeneity. The stability of this genotype adapting to some environmental pressures

through years in the same area supports the aforementioned statement.

164

Two major factors can predispose pathogens such as S. neurona to infect multiple
hosts: 1) high levels of genetic diversity and 2) frequent opportunities for cross-species
transmission (Woolhouse et al., 2001). Thus, extent of host adaptation might be linked to
or limited by the genetic variability. Pathogens that contain more accumulating mutations
should produce more genetic variants and are more likely to be generalists. The
identification of a unique TA insertion mutation in the sequences of the 25/396 marker,
which was present in 10/11 of S. neurona isolates from the state of Michigan might have
an as-yet-unknown important epidemiological or evolutionary implication. An
alternative explanation to the extent of host adaptation in S. neurona is to consider that
the infection of a given host is only related to opportunity; therefore, S. neurona that
infect several different hosts are more likely to be isolated in different transmission
cycles. Indeed, genotypic homogeneity among Michigan isolates may be promoted by
the dispersal capacity of the organism. The sarrre dispersal capacity can also facilitate the
spread of genetic mutation such as the TA insertion mentioned above.

The application of systematic and molecular tools herein enabled us to explore
the interface of population genetics and epidemiology of the S. neurona organism and
allowed us to better determine the genetic diversity of various S. neurona isolates from
different hosts and various geographic regions. Sequence deviation and the differences in
the 25/396 DNA sequences among S. neurona isolates appeared to be informative in this
study and could be used as a marker to describe species boundaries in conjunction with
other testing. The identical 25/396 marker sequence similarity between N. caninum and
S. neurona isolates could be due to the fact that the 25/396 marker is a more stable

genetic locus or intrinsically highly conserved between these two species.

165

This might be a problematic in inferring the phylogenetic relationships between the two
species based on 25/396 marker sequence only. This finding emphasizes the importance
of using more than one gene or DNA marker for a robust phylogenetic analysis.

However, for long-term epidemiological studies, greater isolate discrimination could be
obtained by sampling variation at multiple genomic loci. In this regard, amplified —.
fragment length polymorphism (AF LP) markers may provide a very useful tool for better I
understanding of the genetic diversity, population structure, and gene flow in the natural I

population of S. neurona from different geographic areas. This information is more

 

likely to provide a clear understanding of the genetic differences among isolates, which
may be associated with the organism’s virulence and the severity of the EPM disease it

causes in horses.

166

REFERENCES

Cutler TJ, MacKay RJ, Ginn PE, Gillis K, Tanhauser SM, LeRay EV, Dame J B, Greiner
EC (2001) Immunoconversion against Sarcocystis neurona in normal and
dexamethasone-treated horses challenged with S. neurona sporocysts. Vet Parasitol 95:
197—210

Dubey JP (2000) Prevalence of Sarcocystis species sporocysts in wild-caught opossums
(Didelphis virginiana). J Parasitol 86:705—7 1 O

Dubey JP, Davis SW, Speer CA, Bowman DD, de Lahunta A, Granstrom DE, Topper
MJ, Hamir AN, Cummings JF, Suter MM (1991) Sarcocystis neurona n. sp. (Protozoa:
Apicomplexa), the etiologic agent of equine protozoal myeloencephalitis. J Parasitol 77:
212-218

Dubey JP, Lindsay DS, Saville WJ, Reed SM, Granstrom DE, Speer CA (2001). A
review of Sarcocystis neurona and equine protozoal myeloencephalitis (EPM). Vet
Parasitol 95: 89—131

Elsheikha, H.M., Lacher, D.W., Whittam, T.S., Mansfield, L.S. (2004a). Phylogenetic
relationships of Sarcocystis neurona isolates of horse and opossum to other cyst-forming
coccidia deduced from SSUrRNA gene sequences. Parasitology (In Review)

Elsheikha HM, Mackenzie CD, Rosenthal BM, Marteniuk JV, Steficek B, Windsor S,
Saeed AM, Mansfield LS (2004b) An outbreak of besnoitiosis in miniature donkeys. J
Parasitol (In Press)

Elsheikha HM, Rosenthal BM, Mansfield LS (2004c) Dexamethasone teatment induces
susceptibility of outbred Webster mice to experimental infection with Besnoitia darlingi
isolated from opossums (Didelphis virginiana). Parasitol Res (In Press).

Felsenstein J (1985) Confidence limits on phylogenies: An approach using the bootstrap.
Evolution 39:783—791

Franklin RP, MacKay RJ, Gillis KD, Tanhauser SM, Ginn PE, Kennedy TJ (2003) Effect
of a single dose of ponazuril on neural infection and clinical disease in Sarcocystis
neurona-challenged interferon-gamma knockout mice. Vet Parasitol 114:123-130

Kimura M (1981) Estimation of evolutionary distances between homologous nucleotide
sequences. Proc Nat Acad Sci USA, 78: 454—458

Kruttlin E, Vrable R, Murphy AJ, Rossano MG, Mehler S, Mansfield LS (2001) The
effects of pyrantel tartrate on Sarcocystis neurona merozoite viability. Vet Therap 2:268-
276

Kumar S, Tamura K, J akobsen IB, Nei M (2001) MEGA2: Molecular Evolutionary
Genetics Analysis software. Bioinformatics 17:1244—1245

167

Lindsay DS, Dubey JP (2001) Determination of the activity of pyrantel tartrate against
Sarcocystis neurona in gamma-interferon gene knockout mice. Vet Parasitol 97: 141—144

Mansfield LS, Schott HC2nd, Murphy AJ, Rossano MG, Tanhauser SM, Patterson J S,
Nelson K, Ewart SL, Marteniuk JV, Bowman DD, Kaneene JB (2001) Comparison of
Sarcocystis neurona isolates derived from horse neural tissue. Vet Parasitol 95:167—178

Rosenthal BM, Lindsay DS, Dubey JP (2001) Relationships among Sarcocystis species
transmitted by New World opossums (Didelphis spp.). Vet Parasitol 95:133—142

Saitou N, Nei M (1987) The neighbor-joining method: a new method for reconstructing
phylogenetic trees. Mol Biol Evo 4:406—425

Sitnikova T, thetsky A, Nei M (1995) Interior-branch and bootstrap tests of
phylogenetic trees. Mol Biol Evo 12:319—333

Stanek JF, Dubey JP, Oglesbee MJ, Reed SM, Lindsay DS, Capitini LA, Njoku CJ,
Vittitow KL, Saville WJ (2002) Life cycle of Sarcocystis neurona in its natural
intermediate host, the raccoon, Procyon lotor. J Parasitol 88:1151-1158

Swofford DL (2002) PAUP*. Phylogenetic Analysis Using Parsimony (*and Other
Methods). Version 4. , Sinauer Associates, Sunderland, Massachusetts

Takahashi K, Nei M (2000) Efficiencies of fast algorithms of phylogenetic inference
under the criteria of maximum parsimony, minimum evolution, and maximum likelihood
when a large number of sequences are used. Mol Biol Evo 17: 125 1—1258

Tanhauser SM, Yowell CA, Cutler TJ, Greiner EC, MacKay RJ, Dame JB (1999)
Multiple DNA markers differentiate Sarcocystis neurona and Sarcocystisfalcatula. J
Parasitol 85:221—228

Thompson JD, Gibson TJ, Plewniak F, J eanmougin F, Higgins DG (1997) The ClustalX
windows interface: flexible strategies for multiple sequence alignment aided by quality
analysis tools. Nucleic Acids Res 25: 4876—4882

Witonsky S, Morrow JK, Leger C, Dascanio J, Buechner-Maxwell V, Palmer W, Kline
K, Cook A (2004) Sarcocystis neurona-specific immunoglobulin G in the serum and
cerebrospinal fluid of horses administered S. neurona vaccine. J Vet Inter Med 18: 98—
103

Woolhouse ME, Taylor LH, Haydon DT (2001) Population biology of multihost
pathogens. Science 292:1109—1112

168

 

CHAPTER 6
GENETIC VARIATIONS AMONG SARCOCYST IS NE URONA AS REVEALED
BY AFLP MARKERS
ABSTRACT
The genetic diversity of Sarcocystis neurona was evaluated using amplified

fragment length polymorphism (AF LP) method. F ifieen S. neurona taxa fi'om different
regions in the State of Michigan collected over the last 10 years were used; six strains
were from clinically diseased horses, eight strains were from wild-caught opossums
(Didelphis virginiana), and one isolate was from a cowbird (Molothrus ater).
Additionally, four outgroup taxa were also fingerprinted. Nine primers were used to
generate AF LP patterns, with a total number of amplified fragments ranging from 30 to
60, depending on the strain and primers tested. Based on presence/absence of amplified
AF LP fragments and pairwise similarity values, all the S. neurona strains tested were
clustered in one monophyletic group. No significant correlation could be found between
genomic similarity and host origin of the S. neurona strains. AF LP revealed significant
intraspecific genetic variations and S. neurona appeared as a highly variable species.
Furthermore, Linkage disequilibrium analysis suggested that S. neurona populations
within Michigan have an intermediate type of population structure that included
characteristics of both clonal and panamictic population structure. AF LP is a reliable
molecular technique that provided one of the most informative approaches to ascertain
phylogenetic relationships in S. neurona, and its closest relatives, allowing them to be
clustered by relative similarity using band matching and unweighted pair group method

with arithmetic mean analysis, which may also be applicable to other related species.

169

 

INTRODUCTION

Sarcocystis neurona is the most important cause of equine protozoal
myeloencephalitis (EPM) in horses and other vertebrates. EPM is the most commonly
diagnosed neurological disease of horses in the Americas (Dubey et al., 2001). As an
initial and important step towards an appropriate strategy for management of EPM
infections, the genetic diversity of the S. neurona populations must be assessed.
Understanding the clinical role of S. neurona genetic variants in causing EPM is greatly
hampered by the difficulty involved in their isolation, cultivation and identification.
However, the molecular identification and typing of S. neurona associated with equine
diseases and that isolated from opossums using current published methods are hampered
by a lack of resolution at the inter- or intraspecific level. There have been a few studies
that used molecular methods for the identification of S. neurona. These include use of
random amplified polymorphic DNA (RAPD), sequencing analysis of SSUrRNA and
SAGl genes, and restriction analysis of PCR amplicons of a diagnostic marker
(Tanhauser et al., 1999; Elsheikha and Mansfield, 2004; Elsheikha et al., 2004a). These
methods differ in their taxonomic range, discriminatory power, reproducibility, and ease
of interpretation. Additionally, in all of the above-mentioned typing methods, only a very
limited part of the parasite genome is covered through highly specific molecular targeting
of one or more DNA markers.

Accurate measurement of genetic relationships requires a neutral marker system
that distributes randomly throughout the parasite genome such as amplified fragment
length polymorphism (AF LP). The AF LP method is based on (i) digestion of whole-

genomic DNA with two restriction enzymes, (ii) ligation of double-stranded

170

Oligonucleotide adapters to the restriction half sites, and (iii) selective amplification of the
modified restriction fragments with adapter-specific primers that have an extension of one
or more bases at their 3' end (V 05 et al., 1995; Janssen et al., 1996). AFLP analysis is a
relatively new and sensitive whole-genome coverage fingerprinting technique, which
generates highly reproducible results and has superior discriminatory power (Savelkoul et
al., 1999). Also, AF LP provides an unbiased estimate of total genomic variance and is
sufficiently abundant to minimize errors due to sampling variance (Spooner et al., 1996).
A recent review by Masiga et al., (2000) highlighted the value of this technique for

parasite genetic studies. Although AF LP analysis has been widely used for taxonomic,

 

diagnostic, epidemiological, and source-tracking applications (Savelkoul et al., 1999), the
feasibility of using AF LP fingerprinting to genotype S. neurona isolates or any other cyst-
forming coccidians has not been investigated.

The present study was carried out to test the hypothesis that AF LP markers reveal
the phylogenetic relationships between S. neurona strains at the species and subspecies
level. AF LP was used to analyze 15 S. neurona strains and four outgroup taxa. We
demonstrate here the potential of AF LP technique for inferring phylogenetic relationships
and providing information on infraspecific genetic variation in S. neurona strains mainly
from horses and opossums. This information provides a robust methodology and

advances understanding of S. neurona population structure.

171

MATERIALS AND METHODS

Parasite taxa. A total of 15 S. neurona taxa were collected from six naturally
infected horses, eight opossums, and one brown-headed cowbird from various geographic
locations within the State of Michigan. Four closely related taxa were also included in the
study as outgroup species. The parasite taxa, their geographical origin, year of isolation,
and host sources are shown in Table 1. Methods for isolation and cultivation of S.
neurona strains from horses were previously described (Mansfield et al., 2001).
Environmental samples of S. neurona oocysts/sporocysts were collected from small
intestines of naturally infected opossums (D. virginiana) and purified by centrifugation of
the mucosal scrapings on potassium bromide discontinuous density gradient as described
(Elsheikha et al., 2003). Sporocysts were processed and introduced into cell cultures as

described (Murphy and Mansfield, 1999). When this method failed, sporocysts were

bioassayed in y-IFN-KO mice as described (Elsheikha and Mansfield, 2004). Parasite

growth and tissue culture conditions were described elsewhere (Mansfield et al., 2001;
Elsheikha et al., 2003) to obtain viable isolates in the asexual stage of the life cycle.
Briefly, Equine dermal (ED) cells were inoculated with parasites and incubated until cell
death due to parasite proliferation was evident and merozoites were present in high
numbers, at which time merozoites were harvested and cultures passaged to fresh ED
cells. Harvest and purification of the culture-derived parasites were performed using PD-
10 columns as described previously (Elsheikha et al., 2004b). This method produced high

quantities of purified parasites almost free of host cell contamination.

172

Table 1. Origin, source, year of isolation of taxa used in this study.

 

 

Taxa Designation State Year Host Source
, a

S. neurona MII-12 Michigan 1997 Horse (Equs “51"”) Mansfield et al.
, a

S. neurona MIHS Michigan 1997 Horse (54"5 05mm) Mansfield et a1.
, a

5. neurona MIH6 Michigan 1998 Horse (Equs 05W“) Mansfield etal.
, a

S. neurona MIH7 Michigan 1998 Horse (Equs asmus) Mansfield et al.
, a

5. neurona MIH8 Michigan 1993 Horse (50” 05mm) Mansfield et a1.
_ a

s. neurona M1119 Michigan 1998 HOYSCWIUS 03mm) Mansfield etal.
. . . b

S. neurona MIOPl Michigan 1996 Opossum (D' v1rgzmana) Mansfield et al.
. . . b

S. neurona MIOP4 Michigan 1997 Opossum (D' v1rg1mana) Mansfield et al.
. . . b

S. neurona MIOP7 Michigan 1998 Opossum (D' vzrglmana) Mansfield et al.
. . . b

s. neurona MIOP17 Michigan 2002 Opossum (D- Virginian“) Elsheikha et al.
. . . b

s. neurona MIOP19 Michigan 2002 Opossum “1 V'rgmlana) Elsheikha et al.
. . . b

s. neurona MIOP20 Michigan 2002 Opossum (D. Virginian“) Elsheikha et al.
. . . b

S. neurona MIOP29 Michigan 2002 Opossum (D- W’gmmna) Elsheikha et al.
. . . b

s. neurona MIOP30 Michigan 2002 Opossum (D- Virginian”) Elsheikha et al.

S. neurona MSUOP/CBS Michigan 1999 Cowbird(M010thrus ater) Mansfield et al.

S. falcatula st New York 2002 Cowbird (Molothrus ater) ATCC

Toxoplasma gondii RH USA Homo sapiens Lindsay, DS

Neospora caninum Nc-l USA Dog Lindsay, DS

Besnoitia darlingi MIBDl Michigan 2002 Opossum Elsheikha et al.

 

a Horse samples all came from horses with neurological signs of EPM.

b Opossums were picked up as road kill or donated from animal control officers and
humanely euthanatized in compliance to institutional animal care and use committee.

173

DNA extraction and AFLP reactions. Genomic DNA was isolated according to
the manufacturer’s recommendations using a DNeasy tissue kit (Qiagen). AF LP analysis
was performed using the Perkin Ehner AF LP Plant Mapping Kit (PE Applied
Biosystems, Foster City, CA, USA). Details regarding the adaptor and primer sequences
are available in the kit documentation. The restriction and ligation steps of the AF LP
reaction were performed simultaneously in a total reaction volume of 11 pl by adding
200-400 ng of DNA, 1 U of MseI, 5 U of EcoRI, and 1 U of T4 DNA ligase (all enzymes,
New England Biolabs, Beverly, MA, USA) to a reaction mixture containing: 1XT4 DNA
ligase buffer [50 mM Tris-HCl, 10 mM MgC12, 10 mM dithiotheritol, 1 mM ATP, and 25
pg ml'1 bovine serum albumin (BSA, New England Biolabs)], 0.55 M NaCl, 0.2 pM
EcoRI adaptor, and 2 pM MseI adaptor (PE Applied Biosystems). The reactions were
incubated in the dark for overnight at room temperature and room air, and then diluted by
adding 189 pl of TEOJ buffer (20 mM Tris-HCl, 0.1 mM EDTA, pH 8.0).

Preselective amplification reactions were performed following the manufacturer’s
protocol. Successful amplification was verified by running 10 pl of the PCR products on
a 1.5% agarose gel stained with ethidium bromide and visualized using UV illumination.
The remaining product from the preselective amplification (10 pl) was then diluted with
190 pl TE“ buffer and used for subsequent steps. For the selective amplifications, 64
primer combinations were initially tested and 9 primer sets were selected for use in this
work. The selected primer pairs produced the most polymorphic and reproducible
fingerprints. In each case, one EcoRI primer, labeled with a fluorescent dye, was paired
with one MseI primer. The AF LP patterns obtained using primers of the AF LP Plant

Mapping Kit (PE applied Biosystems) were few. Therefore, primers from the AF LP

174

Microbial Fingerprinting Kit (PE Applied Biosystems) were also tested. These primers
were identical to those designed for the plant samples, but contained a fewer selective
nucleotides at the 3’ end of the primer. Selective amplification was performed as directed
in the kit documentation in a total reaction volume of 20 pl. The selective primer pairs
designed for this study are shown in Table 2.

AF LP technique optimization. To test the reproducibility of both the DNA
extraction method and AF LP methodology, we conducted repeated measures of AF LP on
two well-characterized strains (S. neurona MII-Il strain and S. falcatula). DNA
extraction and restriction—ligation and AF LP reactions were performed on aliquots of ~1
X 107 merozoites of both strain on three separate occasions. Additionally, 64 primer
combinations from the AF LP Microbial Fingerprinting Kit and the AF LP small genome
Plant Mapping Kit were screened. Primer combinations producing patterns comprising
30—60 amplified fiagments throughout the size range of 50—500 bp were selected to test
all of the strains. An iterative process was used to select the primer combinations. When
the AF LP patterns were too complex, the complexity was reduced by the addition of an
additional selective nucleotide to the primers. However, if the selective nucleotides were
too numerous, the AF LP patterns were comprised of only a few low molecular weight
fragments. Therefore, to obtain a greater number of amplified fi'agments for detection of
polymorphism, stringency was reduced by using primers with fewer numbers of selective
nucleotides. Since a pre-amplification step could preferentially enrich heterosite
fragments (fragments with different adapters at each end) following adapter ligation (Vos

et al., 1995), the 2-step PCR amplification procedure was chosen: a pre-selective reaction

175

(without any selective nucleotide on primer), followed by selective amplification
reaction.

AFLP fragment detection and analysis. F luorescent-dye labeled amplification
products were separated using a 5% (wt/vol) Long Ranger gel (J. T. Baker, Toronto,
Ontario, Canada) for 3.5 h and an ABI 3100 DNA sequencer (Applied Biosystems).
Detection mixtures consists of 2 pl of PCR products diluted 1:20 in sterile water, 2 pl of F1
deionized formamide, 0.3 pl of internal—lane size standard labeled with TAMRA dye ‘

(GeneScan 500 TAMRA; Applied Biosystems, Foster City, Calif), and 0.7 pl of loading

buffer (supplied with the size standard). Data collection, processing, fragment sizing, and

 

pattern analysis were done by using ABI PRISMTM GeneScan 2.0.2 software (PE Applied V
Biosystems) to obtain fragment sizes and their fluorescent intensity. For the purpose of
numerical analysis, the background level was subtracted from the raw AF LP data with
GeneScan. Data outputs were obtained both as electropherograms and tabular data. The
number of polymorphic bands (scored on the basis of presence/absence) was determined
for the fingerprints of all of the taxa. Only fragments in the range from 50 to 500 bp were
considered because the resolving capability of the sequencing gel functions best in this
size range. Bands that appeared predominantly in more than one strain but were absent
from or less frequent in the other strains were considered to be discriminatory bands. For
each primer pair, the number of polymorphic fragments was divided by the total number
of fragments to calculate a percentage of polymorphism.

Clustering and phylogenetic analysis. Only, distinct, reproducible, well-

 

resolved fragments were scored. The band pattern of each taxon was scored by visual

inspection in the GeneScan program and converted to binary codes based on the presence

176

or absence of the discriminatory bands (1 for presence and 0 for absence). To simplify
data treatment, quantitative polymorphisms (i.e. variable band intensities and/or
reproducible faint bands) were not taken into account. The 1/0 matrix was used to
calculate (dis)similarity coefficients following J accrad for each primer pair separately and
also for the data pooled over all the nine primer combinations (Rivera et al., 1995). The
resulting distance matrices were used to construct an unweighted pair group method by
using arithmetic averages (UPGMA) clustering method (Sneath and Sokal, 1973) as
performed with BioNumerics software package (version 3.0; Applied Maths, Kortrijk,
Belgium). A dendrogram was used to depict graphically the genetic relationship and
phylogeny of strains (Figure 1). The reliability and robustness of the phyletic trees were
tested by bootstrap analysis with 1,000 replications (Felsenstein, 1985).

Linkage Disequilibrium (LD) analysis. Patterns of linkage disequilibrium (LD)
were studied in 15 S. neurona populations and included a total of 434 alleles (2 each,
presence/absence). Lewontin’s D’ (pairwise linkage disequilibrium coefficient) was
computed using the ETLINK program modified for binary data by Thomas S. Whittam
(Microbial Evolution Laboratory, Michigan State University, East Lansing, MI 48824,

USA). All parsimonious uninformative sites are excluded from analysis.

177

 

RESULTS

The AF LP patterns generated were highly reproducible. When we tested the
repeatability of DNA extraction and AF LP technique using identical samples from either
S. neurona MIHl strain or S. falcatula, identical fingerprint patterns were obtained using
3 replicate DNA samples, isolated, and processed on separate occasions, from merozoites
of each isolate. These three different attempts produced no visible differences in
GeneScan for each of the parasites, which ensured the accuracy and reliability of the data.

We used the AF LP technique to develop a convenient and reliable method for
genetic differentiation of S. neurona at the inter- and intraspecific level. Sixty-four primer
combinations were initially tested to identify the most appropriate pairs of selective
primers that give the most polymorphic and phylogenetically informative patterns. Nine
primers were selected (Table 2) based on the number of fragments amplified and the
polymorphism rate observed. These primer pairs were applied to the complete set of
accessions listed in Table 1. Analysis of the 15 strains with nine AF LP primer pairs
identified a total of 548 fragments, of which 359 were polymorphic (62 %
polymorphism).

The correlation between the different genetic similarity matrices, obtained from
the nine separate primer pairs, was evaluated. Correlation findings showed that AF LP
markers obtained from different sets of amplification products provided complementary
information. Even though each primer pair individually could have given an
approximation of almost all the data, considerable differences existed between them.
Hence, data obtained from all nine-primer combinations was used for a more robust

analysis.

178

Table 2. Selected primer combinations and polymorphism rates for AF LP analysis of 15
Sarcocystis neurona strains.

 

 

Primer pairs Total number No. of Polymorphic Polymorphism
of bands bands (%)
E coRI-T/MseI-CAG 5 l 20 39
EcoRI-C/MseI-CTG 91 62 ' 68
EcoRI-C/MseI-CAG 64 34 53
EcoRI-O/MseI-CTG 62 33 53
EcoRI-A/ MseI-CTC 77 73 95
EcoRI-C/ MseI-CTC 53 36 68
EcoRI-C/ MseI-CAT 33 17 51
EcoRI-C/ MseI-CTT 44 18 41
EcoRI-A/MseI-CAT 73 66 90
Mean 60.9 39.9 62

 

179

The AF LP banding patterns were analyzed and genetic relationships among the
strains are indicated in an evolutionary tree constructed by the unweighted pair group
method using arithmetic averages (UPGMA) method from a similarity matrix of J accard
Coefficient using BioNumerics software. The results show that AF LP does not provide
characteristic markers to differentiate S. neurona strains recovered from different host
species in this sample set. At the intraspecific level, AF LP fingerprints of S. neurona
were grouped by numerical analysis in one main cluster, allowing a clear separation of S.
neurona from S. falcatula, N. caninum, T. gondii, and B. darlingi. Figure 1 shows the
UPGMA dendrogram of relatedness (J accard similarity coefficient) based on a total of
842 AF LP markers.

The dendrogram shows that there is a close relationship between S. neurona
strains. This clustering revealed a single monophyletic group of S. neurona species with a
high (100%) confidence value at 82 to 94 % level of genomic similarity. At this level of
similarity, S. neurona did not form separate monophyletic groups despite the fact that
they originated from different hosts. S. neurona strains were distantly separated from the
outgroup taxa S. falcatula at a similarity level of 56% and separated from N. caninum, T.
gondii, and B. darlingi at a similarity level of 41%. The similarity level between
individual fingerprints of each pair of (sub)species, calculated by the J accard coefficient,
ranged from 80 to 94%, indicating an overall low level of genetic heterogeneity among

the strains.

180

% Relative similarity

40 45 50 55 60 65 70 75 80 85 90 95 100

W

77 S. neurona (MIOP30)
S. neurona (MIOP17)
S. neurona (MIOP29)
S. neurona (MIOP7)
S. neurona (MIH6)
58

S. neurona (MIOP20)
S. neurona (MIOP4)
S. neurona (MIHS)
S. neurona (MIH9)
S. neurona (MIOP19)
S. neurona (MIH8)
S. neurona (MIH2)
S. neurona (MSUOPCBS)
S. neurona (MIH7)
S. neurona (MIOPl)
I S. falcatula (st)

100 I B. darlingi (MIBDl)
l 100 I T. gondii (RH)

I N. caninum (NCl)

 

 

 

 

 

 

 

 

100
100 I I

 

 

 

 

 

 

 

 

 

Figure 1. An UPGMA dendrogram based on AFLP markers obtained with 9 primer
pairs. Numbers shown at different nodes represent percentage confidence limits obtained
in the bootstrap analysis. Nodes without numbers had bootstrap values of less than 50.
The scale shown above is the measure of genetic similarity calculated according to the
Jaccard similarity coefficient prepared from AF LP banding patterns of 15 S. neurona
strains. Taxon names and designations are indicated on the right side of the panel.

181

 

In general, Figure 1 did not show any discrete clustering in the S. neurona clade
except for a few subgroups. For example, S. neurona strains MIOP30 and MIOP17 from
opossums clustered well together with 94% similarity, likewise the strains M1116 and
MIOP20 clustered together at 93% similarity. These and MIOP7 and MIOP14 from
opossums clustered together with an overall similarity of 92%. The UPGMA analysis
demonstrated that S. neurona strains from various hosts showed considerable genetic
variability and demonstrated no correlation between the strains and their hosts,
suggesting that the distinct host origins is not a constraint in the relatedness between the
S. neurona strains.

Among a total of 434 AF LP bands, 217 were parsimony informative for
determining phylogenetic relationships between S. neurona strains. A total number of
93744 comparisons were made. The distribution of Lewontin’s D’ (pairwise linkage
disequilibrium coefficient) had a peak around 0 which can be a result of recombination
and or parallel mutation to generate all 4 haplotypes. However, these analyses could not
document sufficient shuffling by recombination to get to linkage equilibrium. This can
be seen in the variance of the mismatch distribution, which is inflated over the

randomized data (observed stdev = 20.2, expected stdev, randomized data = 7.04).

182

 

DISCUSSION

The present study was conducted to evaluate the utility of whole-genome AF LP
technology as a new method for DNA fingerprinting of S. neurona in order to study the
phylogenetic relationships to the (sub)species level of S. neurona isolated from clinically
diseased horses and from opossums. Documentation of genetic relatedness of Sarcocystis
has been a significant impediment to advances in studying these pathogens. To our
knowledge, this is the first time that AF LP has been used to characterize genomic
variability of S. neurona species or any other cyst-forming coccidian taxa.

AF LP accessed multiple independent sites within the genome and allowed a better

 

definition of the relatedness of different S. neurona strains. The results demonstrated the
usefulness of the AF LP technique as a reliable taxonomic tool for the differentiation of S.
neurona, and other related species. It served as a powerful method for the
characterization of infraspecific polymorphism among populations of clinical and
environmental strains of S. neurona. To optimize the method, it was necessary to do
several repeat measures on two different isolates on three separate occasions to ensure
reproducibility. Additionally, criteria for comparisons were established as nine primer
pairs that yielded 30—60 DNA fragments after the PCR, within the size range of 50—500
bp. These were used to screen all of the isolates. The results of various primer pairs
conducted to check the robustness of the dendrogram/estimates of phylogeny clearly
establish that polymorphism revealed by AF LP is not only abundant but also statistically
reliable. Moreover, these results demonstrate that genetic resolution provided by AP LP is
amenable to phylogenetic analysis not only of closely related Sarcocystis species as

shown earlier but even of highly diverse genera such as T. gondii, N. caninum, and B.

183

darlingi.

Genotyping techniques based on a single genetic locus can lead to erroneous
conclusions if other methods or approaches are not employed. The AF LP method has a
higher multiplex ratio (i.e., number of loci simultaneously analyzed per experiment) than
single gene analysis. Also, as AF LP simultaneously accesses multiple independent sites
within the whole genome, it provides overall evaluation of the phylogenetic relatedness
of S. neurona strains. In general, the AF LP method has been a very useful and reliable
technique for detecting polymorphisms and its reproducibility is reported to be very high
(Vos et al., 1995). A noteworthy result of the present study is the finding that AF LP
displayed a higher rate of polymorphism among S. neurona strains compared to that
obtained with PCR-RFLP or sequencing of SAGl or SSUrRNA genes in previous studies
(Rosenthal et al., 2001; Elsheikha et al., 2004a).

At the specific level, the genetic relationships inferred from this study are in
agreement with SSUrRNA and SAGl gene sequence data in the present study in that S.
neurona strains were found to be genetically closer to each other and seemed to originate
from a common monophyletic origin. Indeed, the present study provides evidence
suggesting that S. neurona has originated from a common ancestral species. This is
apparent in the phyletic representation, wherein all the lineages in this species converge
to a single point before separating from other related species and genera included as
outgroup taxa. However, in the present study, the variability observed among S. neurona
populations based on AF LP analysis suggests that the 'species' S. neurona may be a
composite group including sub-lineages. Linkage disequilibrium (LD) analysis suggests

that S. neurona has an intermediate model of population structure. This intermediate

184

 

model exhibited LD values indicative of both clonal and panamictic population structure.
However, it is possible that the apparent linkage disequilibrium is a consequence of some
geographic structuring but the sample size used in this study was too limited
geographically to infer a firm delineation.

The cluster analysis did not reveal any correlation between genomic similarity
and host origin of the populations. Even though the number of populations tested here is
rather low and should be increased to strengthen the informative value of our data, this
study nevertheless provided arguments for the hypothesis that S. neurona may not be
indigenous throughout their current geographical distribution in the New World. The fact
that S. neurona populations from horses and opossums are similar at the genome level
suggests that they could share a common geographic center of origin. Additionally, the
presence of populations with very low genomic similarity in the same region could be the
result of random dissemination from a number of centers of origin and juxtaposition
through agronomical practices rather than arising from extreme genetic drift from a
common and local ancestor.

The ability to accurately detect polymorphisms between S. neurona is of extreme
importance for the design of effective control strategies for this parasite. This study has
established the AF LP conditions and primer combinations that permit the assessment of
genetic diversity between S. neurona subspecies. It further demonstrated that the
technique could be applied in Sarcocystis diversity studies, with potential use for the
identification of intra- or inter-specific genetic markers. Through the evaluation of large
numbers of clearly defined field strains, such AF LP fingerprinting may facilitate the

identification of polymorphisms linked to parasite virulence factors and, thus, contribute

185

to the understanding of host—parasite interactions at the molecular level. Therefore,
further research focusing on the identification of molecular markers able to accurately
discriminate at a number of levels (species, clones, populations, pathotypes) should be
encouraged.

In conclusion, the AF LP technique described in this paper was reproducible for
analyzing genomic DNA from S. neurona isolates and isolates of related taxa and was
effective for showing interspecific and intraspecific molecular variability between strains.
Although there was no obvious correlation between particular strains and their host
origin, this technique revealed close genetic similarity of S. neurona strains obtained
from horses, opossums, and a cowbird from the State of Michigan. Within the scope of
subsequent ecological and epidemiological studies, AF LP fingerprinting could be further
extended to a broad survey of populations present in different hosts and from different
geographic regions, in order to evaluate the impact of coevolution between host and S.
neurona on their biodiversity. Future studies may also include the recovery and
molecular cloning of AF LP marker bands for the identification of species-specific or

type-specific S. neurona markers.

186

REFERENCES

Dubey JP, Lindsay DS, Saville WJ, Reed SM, Granstrom DE, Speer CA (2001) A review
of Sarcocystis neurona and equine protozoal myeloencephalitis (EPM). Vet Parasitol 95:
89—131

Elsheikha HM, Murphy AJ, Fitzgerald SD, Mansfield LS, Massey JP, Saeed MA (2003)
Purification of Sarcocystis neurona sporocysts from opossum (Didelphis virginiana)
using potassium bromide discontinuous density gradient centrifugation. Parasitol Res
902104-109

Elsheikha HM, Mansfield LS (2004) Sarcocystis neurona Major Surface Antigen Gene 1
(SA GI) shows evidence of having evolved under positive selection pressure. Parasitol
Res 94:452-459

Elsheikha HM, Lacher DW, Mansfield LS (2004a) Phylogenetic relationships of
Sarcocystis neurona isolates of horse and opossum to other cyst-forming coccidia
deduced from SSUrRNA gene sequences. Parasitol Res (In Review)

Elsheikha HM, Rosenthal BM, Murphy AJ, Dunams DB, Neelis DA, Mansfield LM
(2004b) Generally applicable methods to purify intracellular coccidia fi'om cell cultures
and to quantify purification efficacy using quantitative PCR. J Microbiol Methods (In
Review)

F elsenstein J (1985) Confidence limits on phylogenies: An approach using the bootstrap.
Evolution 392783—791

J anssen P, Coopman R, Huys G, Swings J, Bleeker M, Vos P, Zabeau M, Kersters K
(1996) Evaluation of the DNA fingerprinting method AF LP as a new tool in bacterial
taxonomy. Microbiology 142:1881—1893

Mansfield LS, Schott HC, Murphy AJ, Rossano MG, Tanhauser SM, Patterson J S,
Nelson K, Ewart SL, Marteniuk JV, Bowman DD, Kaneene JB (2001) Comparison of
Sarcocystis neurona isolates derived from horse neural tissue. Vet Parasitol 95:167—178

Masiga DK, Tait A, Turner CMR (2000) Amplified restriction fragment length
polymorphism in parasite genetics. Parasitol Today 16:350—353

Murphy AJ, Mansfield LS (1999) Simplified technique for isolation, excystation, and
culture of Sarcocystis species from opossums. J Parasitol 852979—981

Rivera IG, Chowdhury MAR, Huq A, Jacobs D, Martins MT, Colwell RR (1995)
Enterobacterial repetitive intergenic consensus sequences and the PCR to generate
fingerprints of genomic DNAs from Vibrio cholera 01, 0139, and non-Ol strains. Appl
Environ Microbiol 61 :2898—2904

187

Rosenthal BM, Lindsay DS, Dubey JP (2001) Relationships among Sarcocystis species
transmitted by New World opossums (Didelphis spp.). Vet Parasitol 95:133—142

Savelkoul PH, Aarts HJ, de Haas J, Dijkshoorn L, Duim B, Otsen M, Rademaker J L,
Schouls L, Lenstra J A (1999) Amplified-fragment length polymorphism analysis: the
state of an art. J Clin Microbiol 37:3083—3091

Sneath PHA, Sokal RR (1973) Numerical Taxonomy: The Principles and Practice of
Numerical Classification. WH Freeman and Co., San Fransisco, CA

Spooner DM, Tivang J, Nienhuis J, Miller JT, Douches DS, Contreras MA (1996)
Comparison of four molecular markers in measuring relationships among the wild potato
relatives Solanum section Etuberosum (subgenus Potatoe). Theor Appl Genet 92:532-540

"'“W

Tanhauser SM, Yowell CA, Cutler TJ, Greiner EC, MacKay RJ, Dame JB (1999)
Multiple DNA markers differentiate Sarcocystis neurona and Sarcocystisfalcatula. J
Parasitol 85:221—228 1

.u It.

 

Vos P, Hogers R, Bleeker M, Reijans M, van de Lee T, Homes M, Frijters A, Pot J, F“
Peleman J, Kuiper M (1995) AF LP: a new technique for DNA fingerprinting. Nucleic
Acids Res 23:4407—4414

188

 

CHAPTER 7
DISCUSSION AND CONCLUDING REMARKS
DISCUSSION

Sarcocystis neurona is an ubiquitous pathogen of equids and can give rise to
severe neurological signs (Dubey et al., 2001; Mansfield et al., 2001). EPM is a widely
distributed disease in the US. The wide bio-diversity in the life cycles and hosts and the
extensive geographic range of S. neurona clearly demonstrate the plasticity of this
parasite and its ability to establish itself under diverse eco-geographic conditions.
However, the taxonomy and speciation of this species remains controversial. In the last
several years, the question has arisen as to whether each S. neurona—like strain isolated
from a different host animal is really an independent species. Also, to date, the genetic
variations between S. neurona strains threatening equine health remain ill defined.

As an initial and essential step towards an effective strategy for management of
EPM infections, the genetic diversity of S. neurona populations must be assessed.
Unfortunately, studies investigating the genetic diversity of S. neurona at the population
level are greatly hampered by the difficulty involved in parasite isolation, purification,
and cultivation. Furthermore, the molecular identification and comparative genotyping of
S. neurona associated with diseased equids and that isolated from normal opossums are
also hampered by a lack of resolution at the inter- or intraspecific taxonomic level.
Therefore, the present study was conducted mainly to provide new methodologies that (i)
enhance parasite isolation and purification and (ii) improve genotypic classification of S.
neurona at the intraspecies level. In the present study, original methods were developed

to facilitate the isolation of highly purified parasite materials. These methods

189

 

substantially improved identification and classification of all tested isolates when we
compared them to cloned and thoroughly defined Apicomplexan parasite taxa. Valuable
contributions provided here include new methods to ensure that each Sarcocystis isolate
was a pure isolate rather than a mixture of two or several isolates. Mixed isolates can be
a real problem because there is considerable confiision and concern regarding the identity
of Sarcocystis parasites of opossums since at least five species can be harbored in the gut
of a single opossum (Tanhauser et al., 1999; Elsheikha et al., 2004b). First, Sarcocystis
spp. sporocysts obtained from opossum small intestine were separated from the host cell

contaminants using Kbr discontinuous density gradient centrifugation (Elsheikha et al.,

 

2003) and were introduced into cell culture as described (Murphy and Mansfield, 1999). .
When this method failed, sporocysts were given orally to gamma interferon gene
knockout (y-IFN-KO) mice to secure the strains. Approximately, 35 days after infection,
mice were humanely euthanatized, the brain recovered at necropsy, ground, and placed
on cultured cells so that the parasites can be introduced into culture. y-IFN-KO mouse
brain acts as biological filters and only sporocysts of S. neurona and S. speeri are known
to induce encephalitis in KO mice. Also, if sporocysts given to the y-IFN-KO mice were
of the S. speeri type or a mixture of S. neurona and S. speeri, and S. speeri somwhow
managed to establish an infection in these mice and consequently were isolated in the
tissue cultures, we could still distinguish these species. S. speeri and S. neurona have
clearly distinct growth patterns and rates in tissue culture by which the two species could
be easily differentiated from one another (Dubey et al., 2000). In our study, the patterns
of development of all parasite isolates grown in tissue culture were consistent with those

of S. neurona (Dubey et al., 2001). Secondly, PCR-RFLP assays were used to confirm

190

the S. neurona species identity of culture-derived parasite materials. Finally, all culture-
derived parasite materials used in this study including S. neurona and other
Apicomplexan taxa were separated from any host cell contaminations by passing parasite
materials through PD-lO columns, which removed almost all non-parasite materials and
yielded not less than 95% pure parasite populations. Column- purified parasite materials
substantially improved the sensitivity of subsequent molecular analysis such as AF LP.
Parasite purity was not a problem with the T. gondii, N. caninum, B. darlingi, and S.
falcatula taxa. These are well-characterized species. For example, the Cornell isolate of
S. falcatula, was obtained from American type Culture collection (ATCC No. 5071,
Manassas, VA). This isolate has been cultivated in vitro and its development was studied
in detail.

Future studies will focus on obtaining and analyzing additional isolates of S.
neurona. Another technique that could be employed to obtain absolute clonal S. neurona
isolates is to use the limiting dilution method. Briefly, ~ 100 pl of the cell culture
maintenance Eagle Minimum Essential Medium containing merozoites at a concentration
of 0.1-merozoite ml'l will be inoculated into each well of the flat-bottomed 96-well
microtiter plates that contain confluent monolayers of bovine turbinate cells (BT cells,
ATTC CRL, 1390). Then, cultures would be incubated at 37 °C in a humidified
atmosphere containing 5% CO; and 95% air. The plate will be examined with an
inverted microscope for wells containing parasites for up to 1.5 months. Using this
method, it is anticipated that we could obtain a pure clonal parasite culture derived from a

single merozoite from each isolate within 8-10 weeks.

191

In this study, new methods were evaluated in addition to previously known
methods in order to improve the genotypic and systematic classification of S. neurona
strains from different animals, particularly from horses and opossums using various
molecular markers. This new AF LP method has been useful for answering these
questions in other systems (Blears et al., 1999; 2000). Molecular typing systems can be
characterized in terms of ability to discriminate, reproducibility, technical complexity, the
cost, the impact of information derived fiom the typing method, and the ease of
interpretation (Arbeit, 1995; Tenover et al., 1997; van Belkum et al., 2001). The choice
between various typing methods depends on a number of factors, such as the objectives of
the study, the level of discriminatory power required, the kinds of DNA preparations and
laboratory conditions available, and the technical skills of laboratory personnel. The
application of AF LP methodology in the presenty study allowed us to detect more genetic
variations between S. neurona isolates and better elucidated their population structure.

In the present study, genotypic analysis based on SSUrRNA sequence data
demonstrated high homology in S. neurona strains from opossums and horses from
different localities. Chapter 3, Figure 2 showed the parsimony tree for 4 SSUrRNA
sequences of 4 S. neurona strains (three from Michigan and one from Kentucky) in
relation to each other and other coccidian sequences. This demonstrated that all S.
neurona strains recovered from Michigan show overall sequence divergence of 2.7%
from the Kentucky strain. These findings indicate that S. neurona is composed of a single
monophyletic group. It also suggests that these S. neurona strains from opossums and
horses have a common evolutionary origin, and that, wildlife reservoirs (opossums) are

very important in the epidemiology of S. neurona infections in horses. Although

192

 

phylogenetic analysis based on SSUrRNA gene sequence may appear as a promising
method for inferring the taxa of S. neurona (Elsheikha et al., 2004a), S. neurona strains
from opossums and horse from the State of Michigan could not be distinguished based on
SSUrRNA sequences because all of them showed absolute sequence homology. These
results suggest that the SSUrRNA region examined has limited value as a marker to
distinguish strains within S. neurona populations. There is not enough polymorphic
variation at the sub-species level in this gene target. The low variation observed among
these S. neurona SSUrRNA sequences agrees with other literature data (Escalante and
Ayala, 1995; Slapeta et al., 2003) and supports the observation that other genetic variants
(e. g. polymorphic genes) are generally needed for designation of a new strain/species.

Outer membrane proteins can be useful for analysis of selection pressure and
genetic diversity. The SAGl gene, the major surface antigen gene 1 is highly conserved
among all Sarcocystidean protozoa. In this study, SAGl gene sequence analysis
indicated that in the absence of recombination, all S. neurona isolates analyzed to date
have a common evolutionary history. The same conclusion has also been drawn from
sequence and phylogenetic analysis of the SSUrRNA gene (Elsheikha et al., 2004a).
Also, SAGl sequence and evolutionary analyses revealed the presence of two alleles of
the SAGl gene, a finding that is first described in the present study. However, both the
SSUrRNA and SAGl gene sequences were not able to clarify the population structure of
S. neurona.

We uncovered a group of S. neurona isolates that could be distinguished from
other S. neurona based on the presence of a unique “TA” insertion within the 334 bp

sequences of the 25/396 diagnostic marker Elsheikha et al., (2004c). This genetically

193

 

distinct group was located in the lower peninsula of Michigan. Also, eighteen other S.
neurona strains in the US were separated into two groups, those in the Northern US (N-
US) and those in the Southern US (S-US), in phylogenetic trees constructed from
sequences of the 25/396 DNA marker. These findings employed PCR-RFLPs of the
25/396 and 33/54 DNA markers and were corroborated by using sequence comparisons
of the 25/396 DNA marker. However, these discoveries should‘be interpreted with
caution because the amount of sequence diversity that separated the N-US and S-US
groups was minimal and documentation of more genetic differences at other loci between
the N-US and S-US groups of S. neurona are required before a firm conclusion about
their segregation according to their geographic location could be made.

The high sensitivity of the AF LP technique and its rich banding pattern (30—60
bands for each pair of selective primers) provided more information about polymorphism
between S. neurona strains, allowing them to be clustered based on percent similarity.
This better reflects their intraspecific genetic relatedness. Linkage disequilibrium, (LD)
analysis of AF LP data provided evidence that S. neurona has an intermediate model of
genetic structure and exhibited (LD) statistical values suggestive of both clonal and
panamictic population stuructre. Even though considerable genetic variation was
revealed by AF LP between S. neurona populations, no correlation between fingerprinting
patterns of S. neurona strains and host origin could be detected. Some authors have
suggested that AF LP may have little power to resolve complex and recent diversification
(Kim et al., 2001). The fact that the evolutionary rate in the Apicomplexan parasite
genome is low (Escalante and Ayala, 1995) and that there was a lack of genetic

structuring observed in S. neurona in this study, both support the notion that AF LP

194

markers seem to be unsuitable to investigate diversification of recent organisms such as
S. neurona.

Thirty S. neurona isolates is the minimum number of isolates required to precisely
study the population structure of this organism. Additionally, these isolates should be
diverse and collected from various animal hosts and from wide geographic regions.
While more intra-population diversity was detected in S. neurona and its population
structure was partly clarified with AF LP compared to other anlysis based on a single
gene/marker, one could argue that small sample sizes (= 15 S. neurona isolates) such as
ours could be responsible for these results. However, in our study we proceeded to
analysis with a smaller number of S. neurona isolates due to the difficulties associated
with isolating and culturing this parasite. The samples analyzed here represent 10 years
of active work to obtain isolates from critically ill horses suffering from EPM. One major
outcome of small sample size studies is that low-frequency polymorphisms may have
gone undetected. However, it is important to note that sampling error associated with use
of small sample size for estimating population differentiation is greatly reduced through
the use of large numbers of loci (Nei, 1987). Furthermore, some studies on other systems
were based on an even more limited isolate sample size and these clearly showed a
significant genetic structuring (Blears et al., 2000; for Cryptosporidium parvum; Giannasi
et al., 2001; for snakes).

In general, a molecular typing system for S. neurona should be simple to perform,
simple to interpret and should be amenable to standardization and quality control.
Molecular typing methods also should be adaptable, to accommodate emerging and

evolving S. neurona strains, and should be cost-effective. Molecular typing methods, as

195

applied to S. neurona, are still in their infancy and the procedures vary from laboratory to
laboratory and lack standardization, making it difficult to compare them. The PCR-RFLP
analysis of DNA markers is currently the reference method for S. neurona species
delineation (Tanhauser et al., 1999). In the present study, PCR— RF LP analysis was
cheap, simple and was applied to uncultured, unpure specimens. It is therefore ideally
suited for molecular epidemiological studies of S. neurona. Although PCR-RFLP can be
easily performed and the reproducibility of the method is acceptable, this technique
shows a relatively poor typing ability particularly for S. neurona sporocysts obtained
from opossums due to the high prevalence of co-existance of more than one Sarcocystis
spp. in opossum intestines. Therefore, this procedure is useful mainly as a secondary
method in combination with 25/3 96 marker sequence analyses (Rosenthal et al., 2001;
Elsheikha et al., 2004c). DNA sequence analysis of 25/396 marker or of the SAGl gene
were more discriminatory than PCR-RFLP and were able to genotype S. neurona isolates
to the species and sub-species level. The main advantage of DNA sequencing over PCR-
RFLP was that sequence data obtained by an individual could easily be added to by
another group, which should propel the field at a much higher rate. Use of well-defined
genetic loci enhanced our ability to genotype S. neurona strains. At this time, AF LP
analysis is the most robust approach by which S. neurona can be classified at the sub-
species level because the results were ultimately based on the entire genome sequence of
the organism. The AF LP method has been useful for the study of microbial genetics,
evolution, taxonomy, and epidemiology in a variety of eukaryotic and prokaryotic

organisms (Blears et al., 1999). However, the outcome and robustness of AF LP analysis

196

are highly contingent on the analysis of a large number of isolates, which can be
problematic when dealing with cyst-forming coccidia.

111 general, a rapid and definitive typing result is most likely to be achieved by
PCR, and post-PCR processing, including nucleotide sequencing or AF LP fingerprinting.
These methods have been made more feasible by the introduction of rapid, cheaper
automated methods and enhanced softwares like GeneScan for analyzing band patterns.
Sequence analysis of specific informative gene amplicons may be required for
epidemiological studies and quality assurance, whereas AF LP analysis may be sufficient
and more useful for other purposes such as studying population genetic structure. A
robust method like AF LP is probably necessary to identify mixed infections with multiple
sub-species. Based on the results of the present study, we recommend multilocus
sequencing and/or AF LP as the primary method(s) for S. neurona intra-species
identification in molecular epidemiological studies. The discriminatory power of these
approaches is high, and both methods can be used to evaluate the genetic heterogeneity
within S. neurona strains and to screen large numbers of strains. However, obtaining S.
neurona isolates from horses was a difficult task and presented a major challenge because
these isolates could not be obtained fi‘om live horses and had to be isolated from the
neural tissues of horses right after their death or euthanasia.

The purpose of conducting population studies on S. neurona is to develop
realistic, effective means of control. From an epidemiological standpoint, it is necessary
to determine the correlation between the reproductive patterns in S. neurona and the
degree of genetic isolation that might lead to identification of discrete populations or

cryptic species in order to improve the management of this infection. We need to answer

197

the basic question of whether all horses with severe EPM have genetically similar
organisms or whether host immunity or other mechanisms control the overall range of
responses observed. Before this study, the clonality in S. neurona and the impact of
sexual recombination on the genetic diversity of S. neurona have not been investigated.
The applicability of different typing methods to investigate these epidemiological
hypotheses depends mainly on the pattern of changes within the targeted polymorphic
loci. For instance, for tracking a transmission cycle between different hosts, it is essential

to use a typing method that remains stable during a few cycles of transmission, but which

 

can change over a longer evolutionary time. S. neurona is well-known to produce
oocysts via meiotic division in the intestinal cells of its definitive hosts; therefore, it is
almost certain that sexual reproduction is occurring in this organism, even if it has not
been documented. Genetic recombination between Sarcocystis spp. or between S.
neurona strains, whether or not it is natural or experimentally induced, has not been
reported. However, there are some examples of more than one Sarcocystis spp. infecting
a given host; for instance, S. neurona and S. falcatula often co-exist in the same
opossums (Elsheikha et al., 2004c). Mixed Sarcocystis infections in opossums have been
reported in different geographic regions (T anhauser et al., 1999; Dubey, 2000; Elsheikha
et al., 2004b). Such mixed infections may result directly from feeding on a host infected
with multiple Sarcocystis species or on multiple hosts with single infections. However, it
is still unclear if recombination events have taken place and, if so, how frequently they
have occurred.

In the present study, all S. neurona strains used for genotypic analysis were

derived from the culture-derived merozoite stage. The rationale of using this particular

198

 

stage of the organism is that S. neurona reproduces clonally in this asexual phase of the
life cycle. Since there is no recombination in this case, the entire genome will be
transmitted intact from one generation to the next and every region of the genome will
have the same evolutionary information. Hence, just one variable region would suffice to
identify different genotypes. Indeed, due to clonality of the merozoite stage, we found
that the relatedness of individual S. neurona strains could be assessed by using any of the
following identified molecular molecules (e. g. SSUrRNA gene, SAGl gene, 25/396
DNA marker, AF LP markers), and the relationship was straightforward. In contrast,
during the sexual phase of the S. neurona life cycle, every region of the genome might
have different evolutionary information due to the potential for genetic recombination.
Therefore, it would be impossible for a single variable locus to identify genetically
different individuals; a collection of loci would be necessary. In this situation, genetic
distance among isolates could be assayed, but the accuracy of the estimate would be
proportional to the number of loci assayed.

In this study, nucleotide and amino acid sequence analyses of SAGl of S.
neurona strains from various geographic regions revealed that two short segments located
in the 3’ domain and the intron regions of this gene have evolved under positive selection
pressure. Although, the mechanisms underlying the generation of positive selection in
SAGl remain unknown, the identification of areas of the SA GI gene under diversifying
selection has important implications. SAGl gene encodes one of the major surface
antigens in S. neurona and other cyst-forming coccidia. It is known that surface antigens

are continuously subjected to selection pressure forces exerted from the host immune

199

 

 

defense. This evolutionary information could be used for the selection of particular S.
neurona genotypes for development and testing of chemotherapeutics or vaccines.

EPM cases have been reported in France, Japan, and South America (Dubey et al.,
1999; Pitel et al., 2002; Katayama et al., 2003). However, these cases were diagnosed
based on serology alone. S. neurona has not been isolated from horses from these
countries. Recently, S. neurona was detected in a horse with neurological signs in India;
however, the infection may have been acquired in the USA (Mansfield et al., 2001). S.
neurona strains in the US have been isolated from a large diversity of mammalian hosts
and birds (Dubey, 2000; Cheadle et al., 2001; Dubey et al., 2001; Tanhauser et al., 2001;
Mansfield et al., unpublished data). The principal vectors of S. neurona are opossums of
the genus Didelphis (Rosenthal et al., 2001). However, not all of the described
Sarcocystis species of opossums are pathogenic for horses. Currently, only S. neurona
has been cultured from horses with EPM. Although it is uncertain whether the other
Sarcocystis species harbored by opossums can cause equine disease, their pathogenic role
is most likely far less important than that of S. neurona. As reservoir hosts, opossums
infected by various Sarcocystis species may not transmit them to horses and if
transmission occurs, horses may act as filters. In fact, fiom early experimental infection
studies (Dubey et al., 2001) horses seem to be predisposed mainly to infection with only
S. neurona. The mechanisms underlying the apparent differential transmission of S.
neurona to various mammalian, marine, and avian hosts remain to be determined.

The reasons for the recent widespread distribution of S. neurona species could be
attributed to prevalence and pattern of opossum distribution. Opossums have been

radiating throughout the USA and even into Canada aggressively in recent years

200

(Rosenthal et al., 2001). However, SSUrRNA, SAGl, and 258% sequence data in this
study demonstrated low nucleotide divergence within S. neurona strains from horses and
opossums (Elsheikha et al., 2004a, 2004b). These findings suggest that establishment of
S. neurona as a distinct species was a relatively recent event. The limited genetic
diversity observed in this study among the characterized S. neurona strains based on
SSUrRNA, SAGl, and 25/396 sequence analysis suggest also that other events might
have been superimposed on the process of genetic diversity. This may be attributed to a
bottleneck effect that has allowed the survival of only one or few lineages of this
organism (Elsheikha et al., 2004c). Alternatively, considering the enormous host range
and the large geographical distribution of S. neurona, another explanation could involve
both recent evolution followed by spread through mismanagement and poor agricultural
practices (e. g. contaminated food materials).

As indicated earlier, S. neurona is composed of two lineages based on SAGl gene
sequence analysis (Elsheikha and Mansfield, 2004). For example, if one SAGl lineage
of S. neurona colonized the USA, then later diverged into two lineages, then, it is
conceivable to expect that the two lineages would be each other’s closest relatives.
Indeed, there are significant pairwise genetic differences observed between these two
lineages and each SAGl lineage of S. neurona forms an independent clade. This
suggests that these two lineages might have colonized the USA independently. However,
current knowledge of evolutionary relationships between S. neurona strains and their
hosts is really too fragmentary at this time to attempt to infer the origin of an ancestral S.

neurona strain. Further studies should explore this question.

201

CONCLUSIONS

Equine infections due to S. neurona always involve the brain and spinal cord,
resulting in neurological and/or musculoskeletal disorders. In the present study, an
extensive comparative genetic analysis of intraspecies diversity among several S.
neurona strains from different animals, particularly from opossums and horses from the
State of Michigan, USA was conducted using a whole genome fingerprinting AF LP and
DNA sequence analyses. This work was based on the analysis using statistical and
evolutionary genetic means in order to identify parasite clones suitable for the
identification of Discrete Typing Units (DTUs) for epidemiological tracking purposes.
The study is significant in that it applies new molecular tools such as the high-resolution
analytical fingerprinting technique AF LP that has not been used before for studying the
phylogenetic relationships and population structure of S. neurona or any other cyst-
forrning coccidian. These tools can be adapted to initiate a standardized methodology for
laboratories across the US in order to investigate the molecular epidemiology and
evolution of S. neurona. Furthermore, these tools will also help in the examination of the
evolutionary origin and phylogenetic relatedness of other cyst-forming coccidian
parasites.

In this study, several genetic methods and DNA targets were also used to test
previously published hypotheses [e.g., S. neurona and S. mucosa have a common
ancestor (Jenkins et al., 1999) and the surface antigen gene 1 (SAGl) sequence analysis
from several isolates is useful for molecular diagnostic identification of S. neurona and
other related Apicomplexia (Hyun et al., 2003)] in order to determine the appropriateness

of past conclusions proposed in the literature.

202

The following hypotheses were tested in the present work:

H1: S. neurona strains from opossums and horses form a monophyletic group.
H2: Phylogenies of S. neurona strains are consistent with a common history.

H3: The SAGl gene of S. neurona has evolved under positive selection pressure.
H4: Sarcocystis neurona has a clonal population structure.

H5: A genome-wide fingerprinting analysis using AF LP can resolve intra-species

relationships among S. neurona strains.

The following findings resulted from the study:

1.

Potassium bromide discontinuous density gradient centrifugation was a rapid and
efficient procedure for isolation and purification of S. neurona sporocysts from

mucosal scrapings of opossums’ small intestine. (Chapter 2)

PD-10 columns were used successfully for rapid and efficient purification of S.
neurona merozoites and other coccidian species from tissue-cultured cells. (Chapter

2)

There were high SSUrRNA sequence homologies in S. neurona strains from
opossums and horses from different localities. S. neurona strains from opossums and
horses shared a common origin. Phylogenies of S. neurona strains were consistent

with a common history. (Chapter 3)

The SAGl gene of S. neurona is composed of two alleles and has two segments that

have evolved under positive selection pressure. (Chapter 4)

203

5. Sarcocystis neurona has distinct clonal lineages that are consistent with geographic

segregation based on 25/3 69 DNA marker sequence. (Chapter 5).

6. Genome-wide fingerprinting AF LP analysis was more sensitive and detected more
polymorphisms among S. neurona strains than sequence analysis based on a single

gene. (Chapter 6).

The ability to differentiate between S. neurona strains/clones and related species
in this work provided clear insights into population structure and phylogenetic
relationships among S. neurona and other Sarcocystis spp. Linkage disequilibrium
analysis and analysis of the genetic differences/similarities between S. neurona strains
supports the conclusion that S. neurona has an intermediate model of population structure
(i.e., a pOpulation genetic structure that shares the characteristics of both clonal and
panamictic population structure) and that this small population of S. neurona strains in a
limited geographic area cannot be considered to comprise either a single super-species or
multiple sub-species.

These investigations into the phylogenetic relationships of S. neurona clearly
indicate the potential limitations of creating groupings based on morphological properties
of the parasite and host species infected. We recommend that molecular data in the form
of DNA gene sequences, combined with standard taxonomic morphologic criteria should
be given greater consideration in taxonomic placements of S. neurona or any Sarcocystis
species.

The ability to type S. neurona is crucial for epidemiological surveillance and for a
sound evaluation of the impact of this parasite’s genetic diversity on its relevant medical

properties such as virulence or pathogenesis. However, for S. neurona genotyping to

204

have a practical application in these areas, specific DNA bands related to host origins,
particular mutations and/or virulence genes should be cloned, sequenced, and compared
to relational databases as tools for future investigations. These well-identified markers
could be used in future studies aimed to prevent or reduce the severity of equine EPM
and to sustain equine health and well-being.

The results presented in this work should contribute to the understanding of S.
neurona evolution and provide additional grounds for subsequent research. It is essential
that future research be aimed at prevention of EPM through: 1) identification of risk
factors associated with interactions with wildlife, 2) definitive diagnostic tests, and 3)
vaccines that provide adequate protection. The molecular tools developed, and the
findings generated from this project should provide the basis for addressing these three
goals and help to decrease the use of vaccines and anti-microbial agents that do not
improve the health status of horses.

Methods developed and findings obtained in this work emphasized the importance
of using many phylogenetic tree-building methods and comparing alternative models of
DNA sequence evolution in order to make the most reasonable inferences from a given
dataset. Indeed, information obtained in this study also provided us with a broader range

of alternative hypotheses for firrther study and testing.

205

FUTURE DIRECTIONS
EPM is a serious, growing problem for the horse industry in the US and continues
to be a heavy economic burden on the industry and on the life expectancy and
performance of horses in the US. Current infection prevention and control programs
have led to some improvements in the management and control of these infections in the
past several years. However, molecular epidemiology research on EPM faces a series of
challenges, and further progress in prevention regimens against S. neurona infections
demands new approaches. Given the genetic diversity of S. neurona and the disease
manifestations of EPM, the genotyping of S. neurona can contribute valuable data for
possible elucidation of the correlation between a particular S. neurona genotype and the
occurrence and severity of distinct clinical manifestations of EPM. Such data have
important implications for the clinical and laboratory diagnosis of EPM as well as for
epidemiology, vaccine development, and studies on evolution of cyst-forming coccidians.
1. Clinical implications. The association between S. neurona genotypes and the ability
to induce disease has been reported (Butcher et al., 2002). Strain variation has been
observed for some biological and molecular characters in S. neurona (Mansfield et
al., 2001). However, the mechanisms by which infection with different S. neurona
strains may lead to distinct clinical syndromes are largely unknown. Not all horses
exposed to the parasite develop the disease; but when disease exists, clinical signs can
vary from mild ataxia to paralysis, recumbency, and death (Dubey et al., 2001).
Variable responses to treatment in horses infected with S. neurona also suggests that
strain variation may be a factor in the clinical management of EPM. Currently, we

know from sampling soil and opossum feces that naturally infected horses are

206

exposed to multiple Sarcocystis spp. in the environment that may complicate the
understanding of the life cycle of S. neurona and the pathogenesis and diagnosis of
EPM. The extreme difficulty encountered in culturing S. neurona from clinical cases
has hampered assessments of the association between S. neurona isolates and distinct
clinical presentations. Since the recovery of S. neurona from clinical specimens of S.
neurona-infected horses by culture is difficult, assessment of the correlation between
infecting strain and clinical signs will require other approaches. Further study is
needed to enhance recovery of the organism from infected horses antemortem so that
genetic analyses may be conducted. Also, further study is needed to test the
hypothesis that serological tests based on genotype-specific epitopes are capable of
differentiating the antibody response of horses infected with different S. neurona
genotypes. Additionally, if this hypothesis is accepted then these tests would be
helpful in assessments of the association between S. neurona genotypes and distinct
clinical presentations. Since different S. neurona genotypes may be associated with
distinct pathogenic properties, application of sensitive molecular typing methods
which do not require large amounts of parasite material or cultivation of the parasite
to the characterization of S. neurona will provide the framework for a systematic
approach to identify differences in infectivity and pathogenicity between S. neurona
genotypes. Among the typing methods available for S. neurona, SAGl gene
sequencing and AF LP fingerprinting analyses employed in this study can be used to
analyze the genotypic characteristics and to elucidate the genetic relationship and the

pathogenic potential of different S. neurona genotypes. However, a typing method

207

that gives a complete correlation between S. neurona genotype and pathogenicity or
virulence will require more knowledge of the parasite determinants of virulence.
Diagnostic implications. Although specific treatment for EPM is not yet available,
diagnosis is nonetheless important to distinguish between EPM and other related
neurological conditions in order to better assess the prognosis. Disease
manifestations of EPM are diverse and few of these are unique to S. neurona
infection, thus, diagnosis cannot be made on the basis of clinical signs alone. The
current diagnosis of EPM is primarily clinical, but serological tests for parasite
specific antibody in serum or CSF provide usefiil supporting evidence. The
serological tests used to assess S. neurona antibodies include enzyme-linked
immunosorbent assay (ELISA), immunofluorescence (IF A) assays, and Western
blotting. Western blotting and IFA assays are the most widely used but have low
specificity (Daft et al., 2002). Cross-reactive antibodies can produce a false-positive
test in animals with other Sarcocystis infections (Daft et al., 2002). Rossano et al.,
(2000) developed criteria that minimized false-positive results in the western blot test
for detecting equine serum antibodies to Sarcocystis neurona. However, the antigenic
heterogeneity of S. neurona still influences the sensitivity and specificity of
serological tests for EPM. Therefore, this issue should be readdressed, since
molecular typing results based on AF LP data have indicated that the genetic diversity
S. neurona is greater than was previously thought. Therefore, further studies should
be conducted to test the hypothesis that serological tests involving a combination of
antigens derived from two or more different strains are more informative in

determining infection with S. neurona and identifying classical EPM. Such tests

208

 

would also determine the impact of strain heterogeneity on EPM disease outcome.
This approach is expected to overcome some of the problems associated with
phenotypic analyses and would provide additional information regarding the
epidemiology of S. neurona.

. Vaccine development. One of the possible protective measures against EPM is
successfill vaccination of horses at risk. Recently, a whole-cell vaccine was
developed, intrduced and partially approved for horse use by the Food and Drug
Administration in the USA. The low success of this vaccine in preventing cases of

EPM may be due to the low antibody levels generated in vaccinated animals in

 

response to the vaccine or possibly may be due to other intrinsic factors. The efficacy .
of this vaccine is still under study. In addition to the controversy on the efficacy of
vaccination of horses for EPM, the affect of this vaccination procedure on the
efficacy of current serological tests for EPM also decreases the justification of using a
whole cell killed vaccine. Also, it is not known whether antibodies generated in
horses that are immunized with this vaccine derived from one strain can provide
coverage against a wide range of S. neurona strains.

Fascinating future prospects are raised by new findings on passive immunization
using monoclonal antibodies (MAbs). Passive immunization with a high dose of anti-
S. neurona MAbs directed against the major surface antigen gene, SAGl , of the
parasite appeared to be partially protective against S. neurona challenge infection in
y-IFN—KO mice (Elsheikha et al., unpublished data). MAbs improved the survival
time of y-IFN-KO mice challenged with S. neurona sporocysts and inhibited the

attachment to and invasion of the S. neurona merozoites in cultured cells. However,

209

it is possible that the protective immunity seen here could be strain specific and that
the applicability of this approach may be limited by the documented heterogeneity of
SAGl among S. neurona strains (Elsheikha and Mansfield, 2004). One strategy to
circumvent SAGl heterogeneity is to test the hypothesis that a polyvalent-subunit
vaccine made of a cocktail of various recombinant SAG gene products derived from
different strains yield protective immunity against infection by different genotypes of
S. neurona. Even though the SAGl gene is known to be the immunodominant
surface antigenic protein in S. neurona, other potential candidates for vaccines

include the SnMIClO microneme protein. Other surface proteins like Sn14 and Sn16

 

have also been reported to interact with antibodies against S. neurona (Liang et al.,
1998; Hoane et al., 2003) and should be included in a future polyvalent vaccine.

. Evolutionary implications. Phylogenetic methods used for construction of
evolutionary trees provide an alternative tool for the elucidation of the population
structure of Sarcocystis and for discrimination between recombination vs. clonality as
an evolutionary trend in this parasite. Clonal organisms evolve in tree-like patterns;
each individual has one ancestor and all genetic loci of parents and offsprings will
reflect the same phylogeny. In the recombining organisms, evolution is not tree-like
but is more like a net, with each individual having two parents and horizontal gene
transfer occurring with each mating (Maddison, 1995; Tibayrenc, 1998). An
important challenge in carrying out such a population-based approach for S. neurona
or any other cyst-forming coccidia will be the need to obtain a sufficient number of
isolates from different hosts collected over a long time frame in order to provide a

comprehensive representation of the parasite population over a wide geographic range

210

and wide time span. Therefore, further study is needed to test the hypothesis that
analysis of a large number of S. neurona isolates from different hosts and from
diverse geographic region would elucidate precisely the genetic species structure of S.
neurona. The AF LP technique or a similar procedure would be the appropriate
method to be used in this analysis. Also, it is very important to include in this
analysis S. speeri and S. lindsayi species obtained originally from opossums because

these species serve as intermediate missing links between S. neurona and other

 

Sarcocystis spp. and should provide important information about the evolutionary

origin of S. neurona.

 

 

211

REFERENCES

Arbeit R D (1995) Laboratory procedures for the epidemiologic analysis of
microorganisms, p. 190—208. In: Murray PR, Baron EJ, Pfaller MA, Tenover FC, Yolken
RH (ed.), Manual of clinical microbiology, 6th ed. ASM Press, Washington, DC

Blears MJ, De Grandis SA, Lee H, Trevors JT (1999) Amplified fragment length
polymorphism (AF LP): review of the procedure and its applications. J Ind Microbiol
Biotechnol 21 :99—1 14

Blears MJ, Pokomy NJ, Carreno RA, Chen S, De Grandis SA, Lee H, Trevors J T (2000)
DNA Fingerprinting of Cryptosporidium parvum Strains Using Amplified Fragment
Length Polymorphism (AF LP). J Parasitol 862838—841

Butcher M, Lakritz J, Halaney A, Branson K, Gupta GD, Kreeger J, Marsh AE (2002)
Experimental inoculation of domestic cats (F elis domesticus) with Sarcocystis neurona or
S. neurona-like merozoites. Vet Parasitol 107:1—14

Daft BM, Barr BC, Gardner IA, Read D, Bell W, Peyser KG, Ardans A, Kinde H,
Morrow J K (2002) Sensitivity and specificity of western blot testing of cerebrospinal
fluid and serum for diagnosis of equine protozoal myeloencephalitis in horses with and
without neurologic abnormalities. Am Vet Med Assoc 221 :1007—1013

Dubey JP (2000) Prevalence of Sarcocystis species sporocysts in wild-caught opossums
(Didelphis virginiana). J Parasitol 86:705—7 l 0

Dubey JP, Kerber CE, Granstrom DE (1999) Serologic prevalence of Sarcocystis
neurona, T oxoplasma gondii, and Neospora caninum in horses in Brazil. J Am Vet Med
Assoc 215:970—972

Dubey JP, Speer CA, Lindsay DS (2000) In vitro cultivation of schizonts of Sarcocystis
speeri Dubey and Lindsay, 1999. J Parasitol 86:671—678

Dubey JP, Lindsay DS, Saville WJ A, Reed SM, Granstrom DE, Speer CA (2001) A
review of Sarcocystis neurona and equine protozoal myeloencephalitis (EPM). Vet
Parasitol 95 :89—1 31

Elsheikha HM, Murphy AJ, Fitzgerald SD, Mansfield LS, Massey JP, Saeed MA (2003)
Purification of Sarcocystis neurona sporocysts fi'om opossum (Didelphis virginiana)
using potassium bromide discontinuous density gradient centrifugation. Parasitol Res
90:104—109

Elsheikha HM, Lacher DW, Mansfield LS (2004a) Phylogenetic relationships of S.

neurona isolates of horse and opossum to other cyst-forming coccidia deduced from
SSUrRNA gene sequences. Parasitol Res (In Press)

212

 

 

Elsheikha HM, Mansfield LS (2004) Sarcocystis neurona Major Surface Antigen Gene 1
(SAGI) shows evidence of having evolved under positive selection pressure. Parasitol
Res 94:452—459

Elsheikha HM, Murphy AJ, Mansfield LS (2004b) Prevalence of Sarcocystis species
sporocysts in Northern opossums (Didelphis virginiana). Parasitol Res 93:427—431

Elsheikha HM, Murphy AJ, Mansfield LS (2004c) Phylogenetic congruence among
Sarcocystis neurona isolates from the United States based on sequence and restriction
fragment Length polymorphism analyses. Sys Parasitol (In Press)

Escalante AA, Ayala FJ (1995) Evolutionary origin of Plasmodium and other
Apicomplexa based on rRNA genes. Proc Natl Acad Sci U S A 92: 5793—5797

Giannasi N, Thorpe RS, Malhotra A (2001) The use of amplified fragment length
polymorphism in determining species trees at fine taxonomic levels: analysis of a
medically important snake, T rimeresurus albolabris. Mol Ecol 10: 19—426

Hoane J S, Carruthers VB, Striepen B, Morrison DP, Entzeroth R, Howe DK (2003)
Analysis of the Sarcocystis neurona microneme protein SnMIClO: protein characteristics
and expression during intracellular development. Int J Parasitol 33:671—679

Hyun C, Gupta GD, Marsh (2003) Sequence comparison of Sarcocystis neurona surface
antigen from multiple isolates. Vet Parasitol 112211—20

Jenkins MC, Ellis JT, Liddell S, Ryce C, Munday BL, Morrison DA, Dubey JP (1999)
The relationship of Hammondia hammondi and Sarcocystis mucosa to other heteroxenous
cyst-forming coccidia as inferred by phylogenetic analysis of the 188 SSU ribosomal
DNA sequence. Parasitology 119:135—142

Katayama Y, Wada R, Kanemaru T, Sasagawa T, Uchiyama T, Matsumura T, Anzai T
(2003) First case report of Sarcocystis neurona-induced equine protozoal
myeloencephalitis in Japan. J Vet Med Sci 65:757—759

Kim KS, J eong HW, Park CK, Ha J H (2001) Suitability of AF LP markers for the study
of genetic relationships among Korean native dogs. Genes Genet Syst 76:243—250

Liang FT, Granstrom DE, Zhao XM, Timoney JF (1998) Evidence that surface proteins
Sn14 and Sn16 of Sarcocystis neurona merozoites are involved in infection and
immunity. Infect Immun 66:1834—1838

Maddison W (1995) Phylogenetic histories within and among species, p. 273—287. In:

Hoch PC, Stephenson AG (ed.), Experimental and molecular approaches to plant
biosystematics. Missouri Botanical Garden, St. Louis

213

 

 

Mansfield LS, SchottII HC, Murphy AJ, Rossano MG, Tanhauser SM, Patterson J S,
Nelson K, Ewart SL, Marteniuk JV, Bowman DD, Kaneene JB (2001) Comparison of
Sarcocystis neurona strains derived from horse neural tissue. Vet Parasitol 95:67—178

Murphy AJ, Mansfield LS (1999) Simplified technique for isolation, excystation, and
culture of Sarcocystis spp. from opossums. J Parasitol 85:979—981

Nei M (1987) Molecular Evolutionary Genetics. Columbia University Press, New York

Pitel PH, Pronost S, Gargala G, Anrioud D, Toquet MP, F oucher N, Collobert-Laugier C,
Fortier G, Ballet J J (2002) Detection of Sarcocystis neurona antibodies in French horses
with neurological signs. Int J Parasitol 322481—485 '

Rossano MG, Mansfield LS, Kaneene JB, Murphy AJ, Brown CM, Schott HC 2nd, Fox
J C (2000) Improvement of western blot test specificity for detecting equine serum
antibodies to Sarcocystis neurona. J Vet Diagn Invest.12:28—32

 

Rosenthal BM, Lindsay DS, Dubey JP (2001) Relationships among Sarcocystis species b
transmitted by New World opossums (Didelphis spp.). Vet Parasitol 951133—142

Slapeta JR, Modry D, Votypka J, J irku M, Lukes J, Koudela B (2003) Evolutionary
relationships among cyst-forming coccidia Sarcocystis spp. (Alveolata: Apicomplexa:
Coccidea) in endemic Afiican tree vipers and perspective for evolution of heteroxenous
life cycle. Mol Phylogenet Evo 27:464—475

Tanhauser SM, Yowell CA, Cutler TJ, Greiner EC, MacKay RJ, Dame JB (1999)
Multiple DNA markers differentiate Sarcocystis neurona and Sarcocystisfalcatula. J
Parasitol 85 2221—228

Tanhauser SM, Cheadle MA, Massey ET, Mayer BA, Schroedter DE, Dame J B, Greiner
EC, MacKay RJ (2001) The nine-banded armadillo (Dasypus novemcinctus) is naturally
infected with Sarcocystis neurona. Int J Parasitol 31:325—329

Tenover FC, Arbeit RD, Goering RV (1997) How to select and interpret molecular strain
typing methods for epidemiological studies of bacterial infections: a review for
healthcare epidemiologists. Infect Control Hosp Epidemiol 181426—439

Tibayrenc M (1998) Genetic epidemiology of parasitic protozoa and other infectious
agents: the need for an integrated approach. Int J Parasitol 28:85—104

van Belkum A, Struelens M, de Visser A, Verbrugh H, Tibayrenc M (2001) Role of

genomic typing in taxonomy, evolutionary genetics, and microbial epidemiology. Clin
Microbiol Rev 14: 547—560

214

   

111111111111111119111