NOTE
W.W. Toller ? R. Rowan ? N. Knowlton
Genetic evidence for a protozoan (phylum Apicomplexa) associated
with corals of the Montastraea annularis species complex
Received: 8 June 2001 /Accepted: 13 March 2002 / Published online: 8 June 2002

 Springer-Verlag 2002
Keywords Coccidia ? Apicomplexa ? Coral disease ?
srDNA ? Symbiodinium ? Montastraea
Introduction
Scleractinian corals associate with diverse microbial
communities (e.g., Rowan 1998; Rohwer et al. 2001).
Some eukaryotic microbes are common, well-known
associates of corals; notably dino?agellates of the genus
Symbiodinium (e.g., Trench 1987), endolithic algae (e.g.,
Lukas 1974), and fungi (e.g., Bentis et al. 2000). Other
eukaryotic associates have not received detailed study.
Descriptions of these normal associations are a neces-
sary ?rst step towards identifying abnormal microbial
associations (i.e., pathogens) which lead to coral disease
(Richardson 1998).
Methods for molecular genetic analysis have revolu-
tionized the study of marine microbes (e.g., Brinkmeyer
et al. 2000) and have proven useful in the study of coral
microbial communities (e.g., Rowan 1998; Rohwer et al.
2001). Here we give a brief account of a coral-associated
eukaryote which was identi?ed by molecular methods
during a bleaching experiment with the Caribbean coral
Montastraea annularis (Toller et al. 2001a). This sym-
biont, provisionally named genotype N, was identi?ed
during routine screening of Symbiodinium samples by
restriction fragment length polymorphism (RFLP)
analysis of small subunit ribosomal RNA genes
(srDNA). Our data indicate that genotype N represents
a coccidian protozoan (phylum Apicomplexa) which
commonly associates with corals of the Montastraea
annularis species complex.
Methods
Genotype N was identi?ed by an RFLP that was distinct from
those of Symbiodinium taxa known to associate with M. annularis
(Fig. 1a). Here, we report only those methods speci?c to
the identi?cation and characterization of genotype N; methods
for routine RFLP analysis of Symbiodinium srDNA, as well as
methods for experimental manipulation of corals, are given
elsewhere (Toller et al. 2001a, 2001b).
Initially, the srDNA of genotype N was detected following
polymerase chain reaction (PCR) with a ??host-excluding?? primer
combination (Rowan and Powers 1991): the primers were ss5 (a
universal primer) and ss3Z (a primer designed to exclude cnidarian
srDNA). We ?rst observed genotype N in samples where Symbio-
dinium cell numbers were extremely low (<0.4?105 cells/cm2)
following experimental bleaching of host corals (
100-fold reduc-
tion in Symbiodinium numbers relative to controls). These samples
yielded little srDNA when PCR-ampli?ed in the usual (Toller et al.
2001b) manner. To obtain su?cient srDNA, we used two rounds of
ampli?cation as follows: srDNA was ampli?ed from samples with
host-excluding primers for 34 PCR cycles. Aliquots (10 ll) of those
ampli?cations were electrophoresed on agarose gels (1.0% Nuseive
GTG; FMC BioProducts, Rockland, ME), and faint bands of
srDNA were excised and added to 100 ll water. The gel-puri?ed
srDNA was heated to 65 C for 2 min, and then 1 ll of each was
PCR-ampli?ed with host-excluding primers in the same manner.
The resulting re-ampli?ed srDNA was then analyzed by RFLP as
described elsewhere (Toller et al. 2001b).
To further characterize genotype N, srDNA was cloned and
sequenced, as described previously (Toller et al. 2001b). Samples
came from three di?erent colonies of M. annularis in experiment I
of Toller et al. (2001a). srDNA was ampli?ed with universal PCR
primers (clone N0?1) and with host-excluding primers (clones N0?2
and N0?3). For each sample, a minimum of 12 recombinants was
examined by RFLP with Dpn II and with Taq I. DNA sequences
were deposited in GenBank (http://www.ncbi.nlm.nih.gov/) under
accession numbers AF238264 (N0?1), AF238265 (N0?2), and
AF238266 (N0?3).
srDNAsequences fromgenotypeN,SymbiodiniumA,B,C, andE,
and the coral M. annularis were used to design ??N-speci?c?? PCR
primers that amplify only srDNA of genotype N from samples that
alsocontainedcoralandzooxanthellar srDNAs:The5?endsofprimer
18N-F2 (5?-TAGGAATCTAAACCTCTTCCA-3?; forward primer)
and of primer 18N-R1 (5?-CAGGAACAAGGGTTCCCGACC-3?;
Coral Reefs (2002) 21: 143?146
DOI 10.1007/s00338-002-0220-2
W.W. Toller (&) ? N. Knowlton
Marine Biology Research Division, 0202, Scripps Institution
of Oceanography, University of California San Diego,
La Jolla, California 92093-0202, USA
E-mail: westoller@hotmail.com
Tel.: +1-858-8222486
Fax: +1-858-5347313
R. Rowan
University of Guam Marine Laboratory,
Mangilao, Guam 96923, USA
reverse primer) correspond to nucleotides 515 and 1374 (respectively)
in clone N0?1. These primers were used in PCR ampli?cations as
before, except thatDNA synthesis steps were reduced to 1 min. PCR
products were analyzed by RFLP with Dpn II and with Taq I.
For phylogenetic analysis, we aligned srDNA sequences with
Clustal X software (Thompson et al. 1997) and used neighbor-
joining reconstruction (Saitou and Nei 1987) with 1,000 bootstrap
replications. We included srDNA sequences from representative
alveolate taxa (apicomplexans, dino?agellates, ciliates) which were
used previously in similar phylogenetic analyses (McNally et al.
1994; Escalante and Ayala 1995).
Results and discussion
Our initial observations of genotype N in experimentally
bleached corals (see Methods) suggested that N might
normally be present at undetectable levels, and
??revealed?? when Symbiodinium numbers fall. The design
of that experiment enabled us to return to archived
samples taken from the same corals, but prior to bleaching
(i.e., prior to reduction of Symbiodinium numbers).
Re-analysis of these samples withN-speci?c PCR primers
con?rmed that genotype Nwas present (in all cases) prior
to experimental treatment, when Symbiodinium were
abundant. Using the same method, we then screened
samples from other corals. Approximately 90% of
samples taken from apparently healthy, unmanipulated
colonies of M. annularis (45 colonies) and M. faveolata
(7 colonies) tested positive for genotypeN, yielding aPCR
product of ca. 860 bp which, by digestion withDpn II and
with Taq I, was indistinguishable from the PCR product
obtained from clone N0?1 (Fig. 1b). These data indicate
that genotype N commonly associates with M. annularis
and M. faveolata, and that its presence was previously
masked in our analyses by the much greater abundance of
Symbiodinium.
Sequence data con?rm that the srDNA of genotype N
is quite divergent (12?17%) from those of Symbiodinium
A, B, C, and E ? the zooxanthellae known to associate
with M. annularis (see Toller et al. 2001a, 2001b).
Sequence comparisons also explain why genotype N was
ampli?ed with host-excluding PCR primers ? the primer
ss3Z has an extensive match to the srDNA of genotype N
(clone N0?1; not shown; see also McNally et al. 1994). We
performed experiments with mixtures of srDNA clones
from genotype N (cloneN0?1) and Symbiodinium C (clone
C0; Toller et al. 2001b) and con?rmed that our host-
excluding protocol ampli?es both srDNA with approxi-
mately equal e?ciency (not shown). Thus, our method
apparently detects both genotype N and Symbiodinium in
proportion to the relative abundance of their srDNA.
Phylogenetic analysis (Fig. 2) indicates that genotype
N is a member of the protozoan phylum Apicomplexa, a
sister group to the dino?agellates (Gajadhar et al. 1991;
McNally et al. 1994). As far as is known, all apicom-
plexans are animal parasites (Levine 1982). Some, for
example the apicomplexans Plasmodium and Toxoplas-
ma, are notorious pathogens in humans and domestic
animals. In our analysis, genotype N shared greatest
similarity to Toxoplasma, Neospora, Isospora, and Sar-
cocystis, representative taxa from the apicomplexan
class Coccidia (Levine 1982). From these data, we con-
clude that genotype N represents a coccidian protozoan
that associates with corals.
Our three sequences from genotype N (from three
di?erent colonies of M. annularis) suggest that consid-
erable srDNA nucleotide variation (1.0?1.7%) exists
between samples obtained from di?erent colonies. These
srDNA sequences di?er as much as comparisons be-
tween di?erent species of coccidians (e.g., srDNA from
Toxoplasma gondii and Neospora caninum di?er by only
0.4%). However, srDNA heterogeneity is common in
Apicomplexa in general (e.g., Li et al. 1997), and the
accuracy and/or taxonomic signi?cance of individual
sequences may be suspect in such mixtures (see Wintz-
Fig. 1. a RFLP analysis of srDNA from genotype N and
Symbiodinium. srDNA was ampli?ed with host-excluding PCR
primers from a coral sample (N) and from di?erent srDNA clones
representing Symbiodinium A (A), Symbiodinium B (B), Symbiodi-
nium C (C), and Symbiodinium E (E). srDNA was digested with
Dpn II (left) or Taq I (right). The sample containing RFLP
genotype N was taken from Montastraea annularis following
experimental bleaching (see text). Lanes marked M contain size
standards of (from top to bottom) 1,500 bp, 1,200 bp, and then
1,000 bp to 100 bp in 100-bp increments. bRFLP analysis of srDNA
obtained by PCR with N-speci?c primers. srDNA was ampli?ed
from coral samples (lanes 1?3) and an srDNA clone (N0?1) and
digested with Dpn II (left) or Taq I (right). Samples came from three
di?erent colonies of M. annularis. Size markers (M) are as above
144
ingerode et al. 1997; Toller et al. 2001b). In our limited
examination of srDNA from genotype N, we found
evidence of within-sample heterogeneity ? approxi-
mately 25% of the srDNA clones obtained from each
sample had restriction site changes not detected by
RFLP of the original samples (not shown). srDNA
heterogeneity within samples may contribute to the nu-
cleotide variation that we observed between samples,
and our data will serve as a starting point for clarifying
this issue.
Does this coccidian associate with coral hosts outside
of the Montastraea annularis species complex? The an-
swer seems likely to be ??yes.?? Upton and Peters (1986)
used light microscopy to describe a coccidium, Gem-
mocystis cylindrus, from eight species of Caribbean
corals. We made a preliminary analysis of two temper-
ate, azooxanthellate coral species (Balanophyllia elegans
and Astrangia sp. collected from San Diego, CA) and
they also tested positive with N-speci?c PCR (not
shown). We suggest that the association of coccidian
apicomplexans, such as genotype N, with corals may be
common.
We do not yet know if genotype N is a coral par-
asite. Clearly, it is related to a group of totally parasitic
protozoa (Fig. 2), but its ecological relationship with
corals remains obscure. We detected it in corals with
no outward signs of pathology. Observations on
G. cylindrus represent an analogous case, where this
presumably parasitic coccidian was only infrequently
associated with host coral pathology (Upton and Peters
1986). The use of in situ hybridization methods (e.g.,
Distel et al. 1995) will enable localization of N within
host tissues and may further clarify the nature of their
association.
Ultimately, it is likely that understanding the ecology
of coral?apicomplexan associations will be more im-
portant than making a clear distinction between a be-
nign association and a parasitic one. Coral pathogens, in
some cases, may be otherwise benign microbes which
become pathogenic only when their coral host becomes
compromised, for example, under conditions of envi-
ronmental stress (Mitchell and Chet 1975). Assays like
ours can be used to rapidly assess the spatial and tem-
poral patterns of coccidian distribution within and
among host colonies. These methods can be used to test
for the potential involvement of apicomplexans in coral
diseases, which are of increasing concern worldwide
(Richardson 1998).
Acknowledgments We thank Nick Holland, David Kline, Forest
Rohwer, and three anonymous reviewers for their comments on
this manuscript.
References
Bentis CJ, Kaufman L, Golubic S (2000) Endolithic fungi in reef-
building corals (Order: Scleractinia) are common, cosmopoli-
tan, and potentially pathogenic. Biol Bull 198:254?260
Fig. 2. Phylogenetic analysis of
srDNA sequences from geno-
type N and representative pro-
tozoa. srDNA sequences were
aligned and neighbor-joining
reconstruction was used to ex-
plore the relationship of RFLP
genotype N to members of the
protozoan phyla Apicomplexa
(apicomplexans, light gray box),
Dinozoa (dino?agellates, medi-
um-gray box), and Ciliophora
(ciliates, dark gray box). Boot-
strap values (percent of 1,000
replicates) are shown where
support for grouping exceeded
70%
145
Brinkmeyer R, Rappe? M, Gallacher S, Medlin L (2000) Develop-
ment of clade- (Roseobacter and Alteromonas) and taxon-spe-
ci?c oligonucleotide probes to study interactions between toxic
dino?agellates and their associated bacteria. Eur J Phycol
35:315?329
Distel DL, Lee HK-W, Cavanaugh CM (1995) Intracellular coex-
istence of methano- and thiotrophic bacteria in a hydrothermal
vent mussel. Proc Natl Acad Sci USA 92:9598?9602
Escalante AA, Ayala FJ (1995) Evolutionary origins of Plasmodi-
um and other Apicomplexa based on rRNA genes. Proc Natl
Acad Sci USA 92:5793?5797
Gajadhar AA, Marquardt WC, Hall R, Gunderson J, Carmona
EVA, Sogin ML (1991) Ribosomal RNA sequences of Sarco-
cystis muris, Theileria annulata, and Crypthecodinium cohnii
reveal evolutionary relationships among apicomplexans, dino-
?agellates, and ciliates. Mol Biochem Parasit 45:147?154
Levine ND (1982) Apicomplexa. In: Parker SP (ed) Synopsis and
classi?cation of living organisms, vol 1. McGraw-Hill, New
York, pp 750?787
Li J, Gutell RR, Damberger SH, Wirtz RA, Kissinger JC, Rogers
MJ, Sattabongkot J, McCutchan TF (1997) Regulation and
tra?cking of three distinct 18 S ribosomal RNAs during de-
velopment of the malaria parasite. J Mol Biol 269:203?213
Lukas K (1974) Two species of the chlorophyte genus Ostreobium
from skeletons of Atlantic and Caribbean reef corals. J Phycol
10:331?335
McNally KL, Govind NS, Thome? PE, Trench RK (1994) Small-
subunit ribosomal DNA sequence analyses and a reconstruc-
tion of the inferred phylogeny among symbiotic dino?agellates
(Pyrrophyta). J Phycol 30:316?329
Mitchell R, Chet I (1975) Bacterial attack of corals in polluted
seawater. Microbial Ecol 2:227?233
Richardson LL (1998) Coral diseases: what is really known? Trends
Ecol Evol 13:438?443
Rohwer F, Breitbart M, Jara J, Azam F, Knowlton N (2001) Di-
versity of bacteria associated with the Caribbean coral Mon-
tastraea franksi. Coral Reefs 20:85?95
Rowan R (1998) Diversity and ecology of zooxanthellae on coral
reefs. J Phycol 34:407?417
Rowan R, Powers DA (1991) Molecular genetic identi?cation of
symbiotic dino?agellates (zooxanthellae). Mar Ecol Prog Ser
71:65?73
Saitou N, Nei M (1987) The neighbor-joining method: a new
method for reconstructing phylogenetic trees. Mol Biol Evol
4:406?425
Thompson JD, Gibson TJ, Plewniak F, Jeanmougin F, Higgins
DG (1997) The Clustal X windows interface: ?exible strategies
for multiple sequence alignment aided by quality analysis tools.
Nucleic Acids Res 24:4876?4882
Toller WW, Rowan R, Knowlton N (2001a) Repopulation of
zooxanthellae in the Caribbean corals Montastraea annularis
and M. faveolata following experimental and disease-associated
bleaching. Biol Bull 201:360?373
Toller WW, Rowan R, Knowlton N (2001b) Zooxanthellae of the
Montastraea annularis species complex: patterns of distribution
of four taxa of Symbiodinium on di?erent reefs and across
depths. Biol Bull 201:348?359
Trench R (1987) Dino?agellates in non-parasitic symbioses. In:
Taylor FJR (ed) The biology of dino?agellates. Botanical
Monographs vol 21. Blackwell, Oxford, pp 531?570
Upton SJ, Peters EC (1986) A new and unusual species of cocci-
dium (Apicomplexa: Agamococcidiorida) from Caribbean
scleractinian corals. J Invert Pathol 47:184?193
Wintzingerode Fv, Go?bel UB, Stackebrandt E (1997) Determina-
tion of microbial diversity in environmental samples: pitfalls of
PCR-based rRNA analysis. FEMS Microbiol Rev 21:213?229
146