Available online at www.sciencedirect.

com

International Journal for Parasitology 38 (2008) 861–868

www.elsevier.com/locate/ijpara

Genetic characterization and phylogenetic position of Echinococcus

felidis Ortlepp, 1937 (Cestoda: Taeniidae) from the African lion q
Marion Hüttner a,b,*,1, Minoru Nakao b,1, Torsten Wassermann a,
Ludwig Siefert c, Joop D.F. Boomker d, Anke Dinkel a, Yasuhito Sako b,
Ute Mackenstedt a, Thomas Romig a, Akira Ito b
a
Department of Parasitology, University of Hohenheim, Emil-Wolff-Str. 34, 70599 Stuttgart, Germany
b
Department of Parasitology, Asahikawa Medical College, Midorigaoka Higashi 2-1, 078-8510 Asahikawa, Hokkaido, Japan
c
Department of Wildlife and Animal Resources Management, Makerere University, P.O. Box 7062, Kampala, Uganda
d
Department of Veterinary Tropical Diseases, University of Pretoria, Private Bag X04, 0110 Onderstepoort, South Africa

Received 6 September 2007; received in revised form 22 October 2007; accepted 23 October 2007

Abstract

Echinococcus felidis had been described in 1937 from African lions, but was later included in Echinococcus granulosus as a subspecies
or a strain. In the absence of any genetic characterization, most previous records of this taxon from a variety of large African mammals
remained unconfirmed due to the lack of diagnostic criteria and the possible confusion with the sympatric E. granulosus sensu stricto,
Echinococcus ortleppi and Echinococcus canadensis. In this study, we obtained taeniid eggs from lion feces in Uganda and amplified DNA
from individual eggs. Mitochondrial and nuclear DNA sequences showed similarities with those of other Echinococcus spp., but high
values of percentage divergence of mitochondrial genes indicated the presence of a distinct species. In a second step, we compared this
material with the preserved specimens of adult E. granulosus felidis, which had been identified morphologically approximately 40 years
ago in South Africa. All DNA fragments (<200 bp) that could be amplified from the adults showed 100% similarity with the Ugandan
material. In the phylogenetic tree of Echinococcus which was constructed from the mitochondrial genes, E. felidis is positioned as a sister
taxon of E. granulosus sensu stricto. The data obtained will facilitate the development of diagnostic tools necessary to study the epide-
miology of this enigmatic parasite.
Ó 2007 Australian Society for Parasitology Inc. Published by Elsevier Ltd. All rights reserved.

Keywords: Echinococcus felidis; Panthera leo; Phylogeny

1. Introduction intestine of carnivorous definitive hosts belonging to the

families Canidae, Felidae and Hyaenidae (Nelson et al.,
Cestodes of the genus Echinococcus are causative agents 1965; Rausch and D’Alessandro, 2002; Eckert and Depla-
of various forms of zoonotic echinococcosis. The adult zes, 2004). The gravid segments are released into the envi-
tapeworms are only 2–7 mm in length and inhabit the small ronment with the feces of these carnivores. Herbivorous
intermediate hosts including humans become infected with
q the larvae of Echinococcus spp. following ingestion of the
Nucleotide sequence data reported in this paper are available in
DDBJ/EMBL/GenBank databases under Accession Nos. EF558355– eggs. The development of massive hydatid cysts, mainly
EF558360. in the liver and lungs, cause severe pathological effects in
*
Corresponding author. Address: Department of Parasitology, Univer- the intermediate hosts. In spite of the medical and
sity of Hohenheim, Emil-Wolff-Str. 34, 70599 Stuttgart, Germany. Tel.: veterinary importance of these cestodes (Budke et al.,
+49 711 459 23072; fax: +49 711 459 22274.
E-mail address: [email protected] (M. Hüttner).
2006), their taxonomic classification is still controversial,
1
Theses authors contributed equally to this work. particularly concerning the cryptic species complex of

0020-7519/$34.00 Ó 2007 Australian Society for Parasitology Inc. Published by Elsevier Ltd. All rights reserved.
doi:10.1016/j.ijpara.2007.10.013
862 M. Hüttner et al. / International Journal for Parasitology 38 (2008) 861–868

Echinococcus granulosus sensu lato (Bowles et al., 1995). in 2005. Since lions are strictly protected there, the samples
Recently, a robust molecular phylogeny of Echinococcus were collected from the ground without any manipulation
spp. was reconstructed from their complete mitochondrial of the animals. In total, 52 fecal samples were collected,
genomes (Ito et al., 2007; Nakao et al., 2007), and E. gran- originating from an unknown number of animals. The
ulosus sensu lato was split into E. granulosus sensu stricto identification of the feces was made by experienced field
(genotypes G1–G3), Echinococcus equinus (genotype G4), workers, based on shape, colour, odour, ingested hair
Echinococcus ortleppi (genotype G5) and Echinococcus (from grooming) and the presence of field signs such as
canadensis (genotypes G6–G10). Only the enigmatic ‘lion footprints (Breuer, 2005). The feces were kept refrigerated
strain’ from Africa (Thompson and McManus, 2002) has for 3 weeks, were placed at 80 °C for at least 5 days for
eluded any study of its taxonomic position and DNA pro- safety reasons and subsequently stored at 20 °C until use.
file due to the unavailability of suitable isolates. The preserved adult specimens of E. felidis are in the
The lion strain was first described as Echinococcus felidis Department of Veterinary Tropical Diseases, University
in 1937 by Ortlepp (1937) from the lion, Panthera leo, in of Pretoria in South Africa. The material consisted of a
South Africa. The basis for its description as a new species piece of small intestine taken from a lion in South Africa
was the marked rugosity of the rostellar hooks and its about 40 years ago. Many adult worms were still attached
occurrence in a felid as definitive host, which is unique to the mucosal surface. Worms from this collection had
for a member of the E. granulosus complex. In 1963, been morphologically identified by Anna Verster, who
Rausch and Nelson considered this taxon to be conspecific had described the morphology of this taxon. When we
with E. granulosus. However, Verster (1965) re-examined received the material, the tissue was stored in alcohol.
the morphology of Echinococcus spp., and noted the distri- However, the difficulty of DNA purification by a commer-
bution of testes anterior to genital pore in the mature seg- cial kit using proteinase K suggested that the tissue had
ment of the lion parasite. She proposed a subspecific status, originally been fixed in formalin.
as E. granulosus felidis, but this was considered invalid due
to the sympatric occurrence of various such ‘subspecies’ of 2.2. DNA amplification and sequencing
E. granulosus in southern Africa (Rausch, 1967). To the
present, the ‘lion strain’ has been viewed as a form of Taeniid eggs were recovered from lion feces by zinc chlo-
E. granulosus of uncertain taxonomic status, and which is ride flotation (Mathis et al., 1996), and then suspended in
transmitted between lions and large wild herbivores in 10 mM Tris–HCl, pH 7.5 containing 1 mM EDTA (TE
Africa (Macpherson and Wachira, 1997). buffer). To reduce the risk of mixed parasitic infections
Hydatid cysts allocated to the lion strain have been (Müller-Graf, 1995), individual eggs were isolated by using
recorded from a large number of wild mammals such as wil- a capillary pipette. As reported previously (Nakao et al.,
debeest, warthog, bushpig, zebra and various antelopes 2003a), a single egg was lysed in 10 ll of 0.02 N NaOH at
(reviewed by Macpherson and Wachira, 1997). Most of these 95 °C for 10 min, and used directly as a template for PCR.
records are of doubtful validity due to the lack of specific The intestinal tissue including E. felidis adults was washed
diagnoses. Likewise, a number of transmission experiments several times with distilled water and immersed in excess TE
was performed using hydatid cysts from wild mammals in buffer overnight to remove preservative agents. The adults
Africa; some were successful in infecting dogs, while others were recovered from the intestinal mucosa with a fine needle.
were not (Macpherson et al., 1983; Macpherson and Wach- Whole lysates with alkaline solution were made without fur-
ira, 1997). These ambiguous results are possibly due to con- ther DNA purification. Each of the adults was placed in a
fusion with the other taxa of E. granulosus sensu lato, which microtube containing 25 ll of 0.02 N NaOH, and crushed
may also be present in African wildlife. The issue can only be with a pestle. After heating at 95 °C for 10 min, the adult
clarified by molecular characterization of adult Echinococcus homogenate was used as a template for PCR.
worms from lions which correspond morphologically with Template preparations and PCR for the eggs and the
E. felidis. However, obtaining intestinal worms from lions adults were performed using different rooms, divided
in the wild has proven difficult. Therefore, we collected tae- reagents and individual pipettes to avoid DNA contamina-
niid eggs from fecal samples of lions in Uganda, and tested tion. DNA amplifications were mostly performed using
the conspecificity of these isolates with E. felidis by compar- nested PCR because of the minute amount of DNA in sin-
ison with short DNA fragments amplified from the original gle taeniid eggs (Rishi and McManus, 1987) and the severe
adult worm material from South Africa, which had been pre- damage caused to DNA in long-preserved samples by for-
served for approximately 40 years. malin or other acid preservatives. For the first PCR, a 50 ll
reaction mixture containing 0.2 lM of each external pri-
2. Materials and methods mer, 200 lM of each dNTP and 1 U of Ex-Taq polymerase
(Takara, Japan) was prepared as recommended by the
2.1. Parasite specimens manufacturer. Whole parasite lysates were individually
added to the reaction mixture. The added volumes were
Freshly deposited fecal samples from lions were col- 1 ll in the egg lysate and 3 ll in the adult lysate. After
lected in the Queen Elizabeth National Park in Uganda amplification, 1 ll of the first PCR was transferred to a
 M. Hüttner et al. / International Journal for Parasitology 38 (2008) 861–868 863

second PCR consisting of the same reaction mixture, but et al., 2002, 2007) and from Taenia solium (Nakao et al.,
instead of external primers, each internal primer was added 2003b), served as comparative sequences (Accession Nos.
at 0.2 lM. All thermal reactions were performed for 35 AB018440, AB086256, AB208063-4, AB208545-6,
cycles of denaturation (94 °C for 30 s), annealing (55 °C AB235846-8, AF297617 and AF346403). The nuclear
for 30 s) and extension (72 °C for 30–90 s). DNA sequences of elp (Xiao et al., 2005; Hüttner et al.,
Mitochondrial genes for cytochrome c oxidase subunit 1 unpublished data) and ef1a (Nakao et al., unpublished
(cox1), NADH dehydrogenase subunit 1 (nad1), cyto- data) in Echinococcus spp. were also used for comparisons
chrome b (cob) and rRNA (rrn), and nuclear genes for (Accession Nos. AB159141-3, EF660053-4 and AB306933-
elongation factor 1 alpha (ef1a) and ezrin-radixin-moesin 40). The mtDNA sequences of cox1, nad1 and cob were
(ERM)-like protein (elp) were amplified from taeniid eggs translated to amino acid sequences by the echinoderm
by using primers listed in Table 1. The rrn includes the mitochondrial genetic code (Nakao et al., 2000; Telford
large (rrnL) and small (rrnS) subunit rRNA genes. The et al., 2000).
external and internal primers for these mitochondrial genes Multiple alignments of nucleotide and amino acid
were designed from the conserved regions of E. granulosus sequences were made gene by gene, using the Clustal W
mtDNA (database Accession No. AF297617) (Le et al., program via the internet (http://clustalw.ddbj.nig.ac.jp).
2002). In the case of the preserved E. felidis adults, short The alignment of rrn was corrected by deleting ambiguous
mtDNA fragments (<200 bp) of cox1, nad1 and cob were positions such as loops and indels. The conversion of the
amplified by using additional primers (Table 1). aligment format to PHYLIP or NEXUS was done by the
Amplification products were purified by using the QIA- program Seqret of EMBOSS (Rice et al., 2000). The per-
quick PCR purification kit (QIAGEN, Germany) and used centage divergence values of nucleotide sequences were
as a template for direct sequencing. The sequencing reaction computed by PAUP4.0b10 (Swofford, D.L., 2002. PAUP:
was carried out using BigDye terminator ver. 1.1 (Applied phylogenetic analysis using parsimony (and other methods)
Biosystem, USA) and the resultant ladders were read by 4.0 beta. Sinauer Associates, Sunderland, Massachusetts,
ABI PRISM 377 genetic analyser (Applied Biosystem, USA.) using Kimura’s two parameter model (Kimura,
USA). Long DNA templates were sequenced by primer 1980) with a c-shape parameter (a = 1).
walking. The nucleotide and amino acid sequences derived from
mitochondrial genes (cox1, nad1, cob and rrn) were used
2.3. Data processing for phylogenetic analyses. The alignments of nucleotide
sequences were concatenated into a DNA data set (total
Complete mitochondrial genomes, which were available 5170 sites). Likewise, the alignments of amino acid
from 10 taxa of Echinococcus spp. (Le et al., 2002; Nakao sequences were concatenated into a protein data set (total

Table 1
Primer pairs used for PCR
Target genes Primers for the first PCR (5 0 –3 0 )a Primers for the second PCR (5 0 –3 0 )
cox1 (mtDNA) F: TTACTGCTAATAATTTTGTGTCAT –
R: GCATGATGCAAAAGGCAAATAAAC –
nad1 (mtDNA) F: TGGAACTCAGTTTGAGCTTTACTA F: TATTAAAAATATTGAGTTTGCGTC
R: ATATCAAAGTAACCTGCTATGCAG R: TCTTGAAGTTAACAGCATCACGAT
cob (mtDNA) F: GTGGTATTGACATTTTGTAGATTA F: GTTACTAATAGGTTAGTTTAAACT
R: AATTTCCAGAGTAGAAATCACCAT R: CGCCCAACAGTTAAAAAATAACTA
rrn (mtDNA) F: TCCTGTAGCTTGTCATAATGATTA F: ACTTATGGTGTGTATTATATGCGT
R: CTGATTCTCACTACCAAATTTAAC R: ATAAATACACAAACCGCCACAATA
ef1a (nuclear DNA) F: TGGGCAAGTCCTCTTTCAAGTACG F: CTTGATAAGCTCAAGGCTGAGCGT
R: GCGACAGTCGATCTCATGTCACGA R: ATTGGTTTGCTGGGCACCATTCTA
elp (nuclear DNA) F: ATGCGCGTGAGAGTCTTCAGAAGAb F: GAGATGAACAGAAAGCTGAAGGAG
R: ATTCTGCGAAGCTCAGCTTCAb R: CTTGGCCATGGCGACCTTCTGAGC
cox1 (mtDNA)c F: ACTGTTGGGTTGGATGTTAAGACG F: TAGTTCTGTTACTATGATTATAGG
R: CATAACATAATGAAAATGAGCCA R: GTATCATGTAAAACATTATCCAAC
nad1 (mtDNA)c F: TGTTTTTGAGATCAGTTCGGTGTG F: CAGTTCGGTGTGCTTTTGGGTCTG
R: CATAATCAAACGGAGTACGATTAG R: GAGTACGATTAGTCTCACACAGCA
cob (mtDNA)c F: TTATGCTATACTTCGGTGTATTA F: TCGGTGTATTAATTCGAAGATTG
R: ATAAGGATACTCCGGATGACAAC R: GATGACAACCACCCAAATAAGTC
a
F, forward primer; R, reverse primer.
b
Primers previously designed by Xiao et al. (2005).
c
Primers used for historical collections of Echinococcus felidis adults.
864 M. Hüttner et al. / International Journal for Parasitology 38 (2008) 861–868

1188 sites). PAUP4.0b10 was employed for maximum like- feces, in accordance with a previous survey in the region
lihood (ML) analysis using the DNA data set. The substi- (Müller-Graf, 1995). Out of 52 fecal samples, 36 contained
tution model GTR + G + I and its parameters were taeniid eggs. Based on preliminary result of cox1-sequenc-
determined by Akaike Information Criterion (AIC) imple- ing, one fecal sample was selected and the eggs purified
mented in MODELTEST 3.7 (Posada and Crandall, 1998). from this sample were used for the present genetic study.
A full heuristic search algorithm was run to estimate the Using Echinococcus-derived primers (Table 1), the DNA
ML tree. The ML analysis using the protein data set was fragments of four complete mitochondrial genes (cox1,
implemented using Tree-Puzzle 5.2 (Schmidt et al., 2002) nad1, cob and rrn) and two partial nuclear genes (ef1a
with the substitution model JTT + G + I. In both analyses, and elp) were amplified from each of three taeniid eggs.
the robustness of inferred trees was tested by bootstrapping The length of nucleotide sequences determined were
with 1000 replicates. 1608 bp in cox1, 891 bp in nad1, 1068 bp in cob, 1603 bp
Partitioned Bayesian analysis was carried out by MrBa- in rrn, 973 bp in ef1a and 907 bp in elp. The sequences of
yes 3.1 (Ronquist and Huelsenbeck, 2003). The DNA data each gene were completely identical among the three eggs.
sets were divided into four partitions. The nucleotide sub- As shown in Table 2, all the sequences of the taeniid eggs
stitution models GTR + G + I for cox1, nad1 and rrn par- were compared with the known sequences of Echinococcus
titions and HKY + G + I for cob partition were selected by spp. and T. solium. The mitochondrial gene sequences
AIC implemented in MrModeltest 2.2 (modified from obtained from the eggs were similar to those of Echinococ-
MODELTEST by Johan Nylander, Uppsala University, cus spp., but the pairwise divergence values showed that the
Sweden). The protein data sets were divided into three par- taeniid egg species is distinct from all known species of the
titions, and their substitution models were determined by genus. In the nuclear genes, the pairwise divergence values
AIC in PROTTEST 1.26 (Abascal et al., 2005) as follows: were very low when compared with Echinococcus spp., but
Mtmam + G + I for cox1 protein, JTT + G for nad1 pro- the sequences of the eggs were clearly distinguishable from
tein and Mtmam + G for cob protein. Using each of the those of E. granulosus sensu stricto, E. ortleppi and E.
data sets, a Metropolis-coupled Markov chain Monte Car- canadensis (genotype G6), which are known to be present
lo analysis was run for 1 million generations to estimate the in eastern Africa. In conclusion, the eggs were found to
posterior probabilities of trees. The tree sampling was belong to a taxon of Echinococcus, which had not yet been
undertaken every 1000 generations and all tree samples genetically characterized.
before the chain reached a stationarity were discarded as
burn-in.
In all the phylogenetic analyses, the trees were rooted 3.2. Species identification
with T. solium as the outgroup taxon.
To test the hypothesis that the eggs belong to E. felidis,
we examined the historical specimen of E. felidis. Standard
3. Results PCR protocols did not succeed in amplifying mitochon-
drial genes from the long-preserved adults of E. felidis.
3.1. DNA profile of taeniid eggs from lion feces However, the use of alkaline lysate as a template for nested
PCR and the size reduction of target DNA to <200 bp suc-
Parasite eggs of Diphyllobothriidae, Taeniidae, Ancy- ceeded in amplifying short regions of cox1, nad1 and cob.
lostomatidae and Trichuridae were detected in the lion The nucleotide sequences of cox1 from position 913 to

Table 2
Pairwise divergence values (%) of mitochondrial and nuclear DNA sequences between lion-derived taeniid egg and confirmed species
Lion-derived taeniid egg compared with Mitochondrial genes Nuclear genes
cox1 nad1 cob rrn ef1a elp
Echinococcus granulosus sensu stricto 8.4 17.1 11.6 6.6 1.4 1.7
Echinococcus canadensis G6 10.6 17.9 14.8 9.2 1.0 1.6
Echinococcus canadensis G7 10.5 17.9 15.0 9.2 1.0 –
Echinococcus canadensis G8 11.1 19.3 15.0 8.6 0.9 –
Echinococcus equinus 8.1 16.7 12.4 7.6 – –
Echinococcus ortleppi 10.6 18.4 14.8 8.9 0.9 1.5
Echinococcus vogeli 10.0 17.8 14.6 8.4 1.8 –
Echinococcus oligarthrus 11.1 20.0 18.1 10.1 2.4 –
Echinococcus multilocularis 10.4 19.3 15.2 10.4 2.9 5.1
Echinococcus shiquicus 10.4 22.9 15.7 12.1 2.2 4.2
Taenia solium 21.6 37.4 30.2 28.3 – –
(11.2)a (20.8) (15.1) (11.0) (3.3) (5.4)
The complete sequences of mitochondrial genes and the partial sequences of nuclear genes were used for comparison.
a
Percentage divergences between E. multilocularis and E. granulosus sensu stricto are shown in parentheses.
 M. Hüttner et al. / International Journal for Parasitology 38 (2008) 861–868 865

1110, cob from position 818 to 988 and nad1 from position previously (Nakao et al., 2007), the neotropical endemic
405 to 589 were used for comparison. The positions were species Echinococcus oligarthrus and Echinococcus vogeli
numbered from each start codon, based on the complete occupied basal positions. Sister species relationships
mtDNA sequences (Accession No. AF297617) of E. granu- between E. canadensis and E. ortleppi and between E. mul-
losus sensu stricto. These short mtDNA sequences of E. tilocularis and E. shiquicus were also reconfirmed in these
felidis adults were clearly distinguishable from those of analyses. In addition, a new pair of sister species was recog-
other Echinococcus spp., and showed a 100% identity with nized, E. granulosus sensu stricto and E. felidis. The high
those of the taeniid eggs from Uganda (Supplementary values of bootstrap proportions and Bayesian posterior
Fig. S1). Therefore, we conclude that both the taeniid eggs probabilities support these relationships. The species pair
from Ugandan lion feces and the preserved worms from E. felidis and E. granulosus sensu stricto forms the most
South Africa belong to the same taxon, E. felidis. basal lineage among the examined taxa except for E. oli-
Unfortunately, we were unable to reconfirm the mor- garthrus and E. vogeli. A phylogram obtained by ML anal-
phological characteristics of the adult E. felidis, because ysis shows the degree of genetic relatedness (Fig. 1).
the long-preserved worms were too hard, brittle and The phylogeny of Echinococcus spp. was also recon-
incomplete for observation of the diagnostic morphological structed from mitochondrial proteins deduced from cox1,
features. nad1 and cob. Maximum likelihood and partitioned Bayes-
ian analyses using the protein data set resulted in a clado-
3.3. Phylogenetic analyses gram with some differences relating to the position of E.
equinus and the positions of E. vogeli and E. oligarthrus rel-
Using the long mtDNA sequences (cox1, nad1, cob and ative to each other. However, the sister species relationship
rrn) from the taeniid eggs, the taxonomic status of E. felidis between E. felidis and E. granulosus sensu stricto was still
and its phylogenetic relationship with other Echinococcus highly supported by both the analyses (Supplementary
spp. were determined by ML and partitioned Bayesian Fig. S2B).
analyses. The bootstrapped phylogenetic trees were identi-
cal in both analyses (Supplementary Fig. S2A). As reported 4. Discussion

In recent years, the phylogenetic tree of Echinococcus,

mainly based on mtDNA, has been gradually completed
by inclusion of most of the described taxa (Thompson
et al., 2006; Nakao et al., 2007; Ito et al., 2007). The present
study is a further step towards an understanding of the
phylogeny of the genus by including one of its most elusive
members, E. felidis. Our analysis shows that E. felidis is an
independent taxon, as its genetic distance (pairwise diver-
gence values of mt genes) to its closest relative, E. granulo-
sus sensu stricto, is in the range of the distance between E.
multilocularis and E. shiquicus, and far exceeds the distance
between E. ortleppi and genotypes of E. canadensis (Nakao
et al., 2007). Divergence values are also far higher than
between the well accepted species pair Taenia saginata
and Taenia asiatica (Jeon and Eom, 2006; Nakao et al.,
2007). The recent proposition to split E. granulosus sensu
lato, of which E. felidis is doubtlessly a member, into var-
ious species is presently under debate. It is acknowledged
that there are a number of details which still need further
clarification, e.g. the status of genotypes in the cluster E.
ortleppi–E. canadensis and the extent of intergradation
between various species in sympatric situations. In the case
of E. felidis we have no doubt about the species status due
to the maintenance of genetic identity over a wide geo-
Fig. 1. Phylogram of Echinococcus species obtained using the maximum graphical range (isolates from South Africa collected in
likelihood method. The scale bar represents the estimated number of the 1960s versus isolates from Uganda collected in 2007),
nucleotide substitutions per nucleotide site. The sister species relationship the distinct morphology (Verster, 1965), and the unique
between Echinococcus felidis and Echinococcus granulosus sensu stricto is
use of a felid as definitive host.
shown in boldface. The bootstrap proportions (%) of maximum likelihood
(ML) and the posterior probabilities (%) of Bayesian analyses (PB) for Recently, the phylogenetic utility of mitochondrial pro-
nodes Nos. 1–9 are: (1) 44/50, (2) 85/100, (3) 82/100, (4) 97/100, (5) 49/82, tein-coding genes was assessed by using six taxa of cestodes
(6) 94/100, (7) 100/100, (8) 100/100, (9) 100/100, (ML/PB). and five taxa of trematodes (Hardman and Hardman,
866 M. Hüttner et al. / International Journal for Parasitology 38 (2008) 861–868

2006). They demonstrated that reliable nodes were recov- hosts for E. felidis. Since the metacestodes of these para-
ered when 4.0 kb had been sampled and recommended sites develop into similar unilocular cysts, the morphologi-
nad2 or cox1 as a region to include in a sampling program. cal identification of cystic larvae is impossible in domestic
In this study, three protein-coding genes including cox1 and wildlife intermediate hosts. Likewise, taeniid eggs in
(total 3567 nucleotide sites) and ribosomal RNA genes the feces of African lions are microscopically indistinguish-
(1603 nucleotide site) were sampled to reconstruct the phy- able, and at least six species of taeniid tapeworms have
logeny of Echinococcus spp. Maximum likelihood analysis been recorded from Panthera (Dinnik and Sachs, 1972;
using the mtDNA data set resulted in two unreliable nodes Loos-Frank, 2000). Due to this diagnostic uncertainty,
with low bootstrap values (Fig. 1 and Supplementary there exist very few confirmed data on host range and geog-
Fig. S2A), suggesting that our sampling was insufficient. raphy of the parasite. We do not know if lions are exclusive
However, three nodes involving sister species pairs were definitive hosts, or if other felids, canids or hyaenids are
robust in the resultant tree. In particular, the sister species also involved in the lifecycle. Likewise, we know very little
relationship between E. felidis and E. granulosus sensu about the intermediate host range. There is, of course, no
stricto was supported further by partitioned Bayesian anal- doubt that transmission occurs via the important prey spe-
yses using nucleotide and amino acid sequences. cies of lions. However, lions have a very catholic diet that
In cladistic taxonomy, the discovery of sister species can differ substantially between regions. Hayward and Ker-
relationships is important in studying the evolutionary pro- ley (2005) compiled data from 32 studies on the diet of
cesses of closely related organisms. As reported in the lions from different locations and found that the diet
genus Taenia (Hoberg, 2006), sister species and their geo- included 42 prey species, mainly ungulates of the orders
graphical distribution play a crucial role in understanding Artiodactyla and Perissodactyla. Almost 20 species of wild
the evolutionary origin of human parasites. In this study, ungulates are on record as harbouring hydatid cysts (Mac-
molecular phylogenetic analyses demonstrated a sister spe- pherson and Wachira, 1997). Not all of these cysts neces-
cies relationship between E. felidis and E. granulosus sensu sarily belong to E. felidis, as was recently demonstrated
stricto. Human cystic echinococcosis caused by E. granulo- by the recovery of E. granulosus sensu stricto (G1) from
sus sensu stricto has a global impact (Budke et al., 2006), a wild warthog in East Africa (Hüttner et al., unpublished
and its evolutionary origin remains to be determined. A data). Zebra (Equus quagga) can be considered as an inter-
recent molecular phylogenetic study revealed that modern mediate host, as experimental infection of lions with cysts
Felidae arose in Asia (Johnson et al., 2006). According to from zebra has resulted in successful infections (Young,
their estimation, Asian-derived Panthera species spread 1975).
into America (jaguar and lion) and into Africa (lion and At present, there are no data available on the pathoge-
leopard) in late Pliocene (3–2 million years ago). Because nicity of E. felidis to humans. Its close relationship with
of the lack of a molecular clock in tapeworms, we cannot E. granulosus sensu stricto suggests a zoonotic potential,
estimate the molecular date of bifurcation between E. feli- as the latter species is apparently responsible for the major-
dis and E. granulosus sensu stricto exactly. However, the ity of human cases worldwide (Jenkins et al., 2005). How-
high values of pairwise divergence in their mitochondrial ever, the public health impact of this species is expected to
genes (Table 2) indicate that the bifurcation had occurred be minimal, as lions are largely restricted to national parks
already in Asia, when the general estimation of base substi- and game reserves with limited human activity. It may have
tution in mtDNA (2% divergence per million years) was some impact on pastoralists in East Africa who still coexist
applied (Avise et al., 1992). In this scenario, E. felidis with wildlife such as the Maasai in Kenya and Tanzania.
invaded Africa together with lions in the late Pliocene. Ultrasound scanning surveys demonstrated the frequent
The dog tapeworm E. granulosus sensu stricto utilizes sheep occurrence of cystic echinococcosis among the Maasai of
as the main intermediate host, and the populations of northern Tanzania (prevalence 1.1%) (Macpherson et al.,
moufflon (Ovis gmelini) in Asia and Asia Minor are widely 1989). These pastoralists settle on the eastern border of
accepted as the ancestors of domestic sheep (Pedrosa et al., the Serengeti where wildlife is common even outside the
2005). This information also supports the hypothesis that protected area. An exposure to lion-derived Echinococcus
E. granulosus sensu stricto originated in Asia. Further stud- eggs is therefore feasible, but further epidemiological stud-
ies on host range and geographical distribution are needed ies are required to evaluate the pathogenicity of E. felidis to
to test these hypotheses. In contrast to E. granulosus sensu humans.
stricto which was globally transferred with livestock, E.
felidis as been less subject to anthropogenic re-distribution, Acknowledgements
and any record of its presence should reflect the natural
range. This work was financially supported by the Eiselen
In Africa, the dog tapeworms E. granulosus sensu Foundation Ulm, German Academic Exchange Service
stricto, E. ortleppi and E. canadensis (genotype G6) occur (DAAD) and by the Japan Society for the Promotion of
in sheep, cattle and camels, respectively (McManus and Science (JSPS) for international projects (14256001,
Thompson, 2003). The distribution of these domesticated 17256002) and JSPS-Asia/Africa Science Platform Fund
mammals overlaps with the natural habitats of wildlife and Infection Matrix Fund from the Ministry of Educa-
 M. Hüttner et al. / International Journal for Parasitology 38 (2008) 861–868 867

tion, Japan, to A.I. We thank Mzee James Kalyewa for his Macpherson, C.N.L., Wachira, T.W.M., 1997. Cystic echinococcosis in
help in the field in Uganda. Africa south of the Sahara. In: Andersen, F.L., Ouhelli, H., Kachani,
M. (Eds.), Compendium of Cystic Echinococcosis in Africa and in
Middle Eastern Countries with special Reference to Morocco. Brig-
Appendix A. Supplementary data ham Young University, Provo, pp. 245–277.
Mathis, A., Deplazes, P., Eckert, J., 1996. An improved test system for
Supplementary data associated with this article can be PCR-based specific detection of Echinococcus multilocularis eggs. J.
found, in the online version, at doi:10.1016/j.ijpara.2007. Helminthol. 70, 219–222.
Müller-Graf, C.D.M., 1995. A coprological survey of intestinal parasites
10.013. of wild lions (Panthera leo) in the Serengeti and Ngorongoro Crater,
Tanzania, East Africa. J. Parasitol. 81, 812–814.
References Nakao, M., Sako, Y., Yokohama, N., Fukunaga, M., Ito, A., 2000.
Mitochondrial genetic code in cestodes. Mol. Biochem. Parasitol. 111,
Abascal, F., Zardoya, R., Posada, D., 2005. ProtTest: selection of best-fit 415–424.
models of protein evolution. Bioinformatics. 21, 2104–2105. Nakao, M., Yokoyama, N., Sako, Y., Fukunaga, M., Ito, A., 2002. The
Avise, J.C., Bowen, B.W., Lamb, T., Meylan, A.B., Bermingham, E., complete mitochondrial DNA sequence of the cestode Echinococcus
1992. Mitochondrial DNA evolution at a turtle’s pace: evidence for multilocularis (Cyclophyllidea: Taeniidae). Mitochondrion 1, 497–509.
low genetic variability and reduced microevolutionary rate in the Nakao, M., Sako, Y., Ito, A., 2003a. Isolation of polymorphic microsat-
Testudines. Mol. Biol. Evol. 9, 457–473. ellite loci from the tapeworm Echinococcus multilocularis multilocu-
Budke, C.M., Deplazes, P., Torgerson, P.R., 2006. Global socioeconomic laris. Inf. Genet. Evol. 3, 159–163.
impact of cystic echinococcosis. Emerg. Infect. Dis. 12, 296–303. Nakao, M., Sako, Y., Ito, A., 2003b. The mitochondrial genome of the
Bowles, J., Blair, D., McManus, D.P., 1995. A molecular phylogeny of the tapeworm Taenia solium: a finding of the abbreviated stop codon U. J.
genus Echinococcus. Parasitology 110, 317–328. Parasitol. 89, 633–635.
Breuer, T., 2005. Diet choice of large canivores in northern Cameroon. Nakao, M., McManus, D.P., Schantz, P.M., Craig, P.S., Ito, A., 2007. A
Afr. J. Ecol. 43, 97–106. molecular phylogeny of the genus Echinococcus inferred from complete
Dinnik, J.A., Sachs, R., 1972. Taeniidae of lions in East Africa. mitochondrial genomes. Parasitology 134, 713–722.
Z. Tropenmed. Parasitol. 23, 197–210. Nelson, G.S., Pester, F.R., Rickman, R., 1965. The significance of wild
Eckert, J., Deplazes, P., 2004. Biological, epidemiological, and clinical animals in the transmission of Cestodes of medical importance in
aspects of echinococcosis, a zoonosis of increasing concern. Clin. Kenya. Trans. R. Soc. Trop. Med. Hyg. 59, 507–524.
Microbiol. Rev. 17, 107–135. Ortlepp, R.J., 1937. South African Helminths, Part I. Onderstepoort J.
Hardman, M., Hardman, L.M., 2006. Comparison of the phylogenetic Vet. Sci. Anim. Ind. 9, 311–336.
performance of neodermatan mitochondrial protein-coding genes. Pedrosa, S., Uzun, M., Arranz, J.J., Gutiérrez-Gil, B., San Primitivo, F.,
Zool. Scr. 35, 655–665. Bayón, Y., 2005. Evidence of three maternal lineages in Near Eastern
Hayward, M.W., Kerley, G.I.H., 2005. Prey preferences of the lion sheep supporting multiple domestication events. Proc. R. Soc. B. 272,
(Panthera leo). J. Zool. Lond. 267, 309–322. 2211–2217.
Hoberg, E., 2006. Phylogeny of Taenia: species definitions and origins of Posada, D., Crandall, K.A., 1998. MODELTEST: testing the model of
human parasites. Parasitol. Int. 50, S23–S30. DNA substitution. Bioinformatics 14, 817–818.
Ito, A., Nakao, M., Sako, Y., 2007. Echinococcosis: serological detection Rausch, R.L., Nelson, G.S., 1963. A review of the genus Echinococcus
of patients and molecular identification of parasites. Future Microbiol. Rudolphi, 1801. Ann. Trop. Med. Parasitol. 57, 127–135.
2, 439–449. Rausch, R.L., 1967. A consideration of infraspecific categories in the
Jenkins, D.J., Romig, T., Thompson, R.C.A., 2005. Emergence/re- genus Echinococcus Rudolphi, 1801 (Cestoda: Taeniidae). J. Parasitol.
emergence of Echinococcus spp. – a global update. Int. J. Parasitol. 53, 484–491.
35, 1205–1219. Rausch, R.L., D’Alessandro, A., 2002. The epidemiology of echinococ-
Jeon, H.K., Eom, K.S., 2006. Taenia asiatica and Taenia saginata: genetic cosis caused by Echinococcus oligarthrus and E. vogeli in the
divergence estimated from their mitochondrial genomes. Exp. Paras- Neotropics. In: Craig, P., Pawlowski, Z. (Eds.), Cestode Zoonoses:
itol. 113, 58–61. Echinococcosis an Emergent and Global Problem. IOS Press, Amster-
Johnson, W.E., Eizirik, E., Pecon-Slattery, J., Murphy, W.J., Antunes, A., dam, the Netherlands, pp. 107–113.
Teeling, E., O’Brien, S.J., 2006. The late miocene radioation of modern Rice, P., Longden, I., Bleasby, A., 2000. EMBOSS: the European
felidae: a genetic assessment. Science 311, 73–77. molecular biology open software suite. Trends Genet. 16, 276–277.
Kimura, M., 1980. A simple method for estimating evolutionary rates of Rishi, A.K., McManus, D.P., 1987. Genomic cloning of human Echino-
base substitutions through comparative studies of nucleotide coccus granulosus DNA: isolation of recombinant plasmids and their
sequences. J. Mol. Evol. 16, 111–120. use of genetic markers in strain characterization. Parasitology 94, 369–
Le, T.H., Pearson, M.S., Blair, D., Dai, N., Zhang, L.H., McManus, D.P., 383.
2002. Complete mitochondrial genomes confirm the distinctiveness of Ronquist, F., Huelsenbeck, J.P., 2003. MrBayes 3: Bayesian phylogenetic
the horse–dog and sheep–dog strains of Echinococcus granulosus. inference under mixed models. Bioinformatics 19, 1572–1574.
Parasitology 124, 97–112. Schmidt, H.A., Strimmer, K., Vingron, M., von Haeseler, A., 2002.
Loos-Frank, B., 2000. An up-date of Verster’s (1969) ‘Taxonomic revision TREE-PUZZLE: maximum likelihood phylogenetic analysis using
of the genus Taenia Linnaeus’ (Cestoda) in table format. Syst. quartets and parallel computing. Bioinformatics 18, 502–504.
Parasitol. 45, 155–183. Telford, M.J., Herniou, E.A., Russel, R.B., Littlewood, D.T.J., 2000.
McManus, D.P., Thompson, R.C., 2003. Molecular epidemiology of Changes in mitochondrial genetic codes as phylogenetic characters:
cystic echinococcosis. Parasitology 127, S37–S57. two examples from the flatworms. Proc. Natl. Acad. Sci. USA 97,
Macpherson, C.N.L., Karstad, L., Stevenson, P., Arundel, J.H., 1983. 11359–11364.
Hydatid disease in the Turkana District of Kenya III. Ann. Trop. Thompson, R.C.A., McManus, D.P., 2002. Towards a taxonomic revision
Med. Parasitol. 77, 61–73. of the genus Echinococcus. Trends Parasitol. 18, 452–457.
Macpherson, C.N.L., Spoerry, A., Zeyhle, E., Romig, T., Gorfe, M., Thompson, R.C.A., Boxell, A.C., Ralston, B.J., Constantine, C.C.,
1989. Pastoralists and hydatid disease: an ultrasound scanning Hobbs, R.P., Shury, T., Olson, M.E., 2006. Molecular and morpho-
prevalence survey in East Africa. Trans. R. Soc. Trop. Med. Hyg. logical characterization of Echinococcus in cervids from North
83, 243–247. America. Parasitology 132, 439–447.
868 M. Hüttner et al. / International Journal for Parasitology 38 (2008) 861–868

Verster, A.J.M., 1965. Review of Echinococcus species in South Africa. cestode from Tibetan fox and plateau pika in China. Int. J. Parasitol.
Onderstepoort J. Vet. Res. 32, 7–118. 35, 693–701.
Xiao, N., Qui, J., Nakao, M., Li, T., Yang, W., Chen, X., Schantz, P.M., Young, E., 1975. Echinococcosis (Hydatidosis) in wild animals of the
Craig, P.S., Ito, A., 2005. Echinococcus shiquicus n. sp., a taeniid Kruger National Park. J. S. Afr. Vet. Assoc. 46, 285–286.