Cestode genomics - progress and prospects for advancing basic and applied aspects of flatworm biology.

Abstract

Characterization of the first tapeworm genome, Echinococcus multilocularis, is now nearly complete, and genome assemblies of E. granulosus, Taenia solium and Hymenolepis microstoma are in advanced draft versions. These initiatives herald the beginning of a genomic era in cestodology and underpin a diverse set of research agendas targeting both basic and applied aspects of tapeworm biology. We discuss the progress in the genomics of these species, provide insights into the presence and composition of immunologically relevant gene families, including the antigen B- and EG95/45W families, and discuss chemogenomic approaches toward the development of novel chemotherapeutics against cestode diseases. In addition, we discuss the evolution of tapeworm parasites and introduce the research programmes linked to genome initiatives that are aimed at understanding signalling systems involved in basic host-parasite interactions and morphogenesis.

Full text

Page 1

Review Article
Cestode genomics – progress and prospects for advancing basic
and applied aspects of flatworm biology
P. D. OLSON,1M. ZAROWIECKI,1,2F. KISS3& K. BREHM3
1Department of Zoology, The Natural History Museum, London, UK,2Wellcome Trust Sanger Institute, Wellcome Trust Genome
Campus, Hinxton, Cambridge, UK,3Institute of Hygiene and Microbiology, University of W?rzburg, W?rzburg, Germany
SUMMARY
Characterization of the first tapeworm genome, Echinococ-
cus multilocularis, is now nearly complete, and genome
assemblies of E. granulosus, Taenia solium and Hymenol-
epis microstoma are in advanced draft versions. These initia-
tives herald the beginning of a genomic era in cestodology
and underpin a diverse set of research agendas targeting both
basic and applied aspects of tapeworm biology. We discuss
the progress in the genomics of these species, provide insights
into the presence and composition of immunologically
relevantgenefamilies,including
EG95⁄45W families, and discuss chemogenomic approaches
toward the development of novel chemotherapeutics against
cestode diseases. In addition, we discuss the evolution of
tapeworm parasites and introduce the research programmes
linked to the genome initiatives that are aimed at under-
standing signalling systems involved in basic host–parasite
interactions and morphogenesis.
theantigenB-and
Keywords antigen B, cestode, Echinococcus, EG95, genome,
Hymenolepis, S3Pvac
INTRODUCTION
Whole-genome sequencing of cestodes began in 2004 and
currently includes the aetiological agents of alveolar
echinococcosis (AE; Echinococcus multilocularis), cystic
echinococcosis (CE; E. granulosus) and neurocysticercosis
(NCC; Taenia solium) in addition to the rodent-hosted
laboratory model, Hymenolepis microstoma. With the
genomes of Echinococcus spp. near completion, and those
of Taenia and Hymenolepis in advanced drafts, we have
only begun to explore their full content, structure and
general characteristics. Nevertheless, genomic and tran-
scriptomic data are already advancing research in both
basic and applied aspects of tapeworm biology and herald
the beginning of a new era in cestodology. Here, we review
the progress made in the genomics of tapeworms and pro-
vide initial insights into the presence of immunologically
relevant molecules and chemogenomic approaches to the
development of new vaccines. We begin by discussing their
evolution and diversification into homeothermic hosts
including humans and finish by introducing the research
programmes on signalling systems involved in host–para-
site interactions and development that underpin two of
the genome initiatives.
EVOLUTION OF TAPEWORM PARASITES
Tapeworms represent an extreme example in the evolution
of parasitism in flatworms (phylum Platyhelminthes),
being distinguished from the other parasitic groups by the
complete loss of a gut and a highly modified, segmented,
body plan. They are almost exclusively enteric parasites of
vertebrates as adults, with complex life cycles involving
ontogenetically distinct larval stages that first develop in
arthropod hosts, although variation in everything from
their basic body architecture to their host associations is
found among an estimated 6000 species. Like free-living
flatworms, tapeworms maintain
(called neoblasts) throughout their lives (1–5), providing
them with an extraordinary degree of developmental
plasticity and a theoretical potential for indeterminate
growth (6). Although tapeworm infection of humans is
less prevalent than that of trematodes such as Schistosoma
and Fasciola, their enormous reproductive output and
totipotent stem cells
PIM
1319B
-
Dispatch: 8.8.11 Journal: PIM
CE: Deepika K
Journal NameManuscript No.
Author Received:No. of pages: 21 PE: Padmalekshmi
Correspondence: Klaus Brehm, Institute of Hygiene and Microbi-
ology, University of W?rzburg, Josef-Schneider-Strasse 2,
D-97080 W?rzburg, Germany
(e-mail: kbrehm@hygiene.uni-wuerzburg.de).
Disclosures: None.
Received: 16 May 2011
Accepted for publication: 07 July 2011
Parasite Immunology, 2011, 33, 1–21DOI: 10.1111/j.1365-3024.2011.01319.x
? 2011 Blackwell Publishing Ltd
1
2
3
4
5
6
7
8
9
10
11
12
13
14
15
16
17
18
19
20
21
22
23
24
25
26
27
28
29
30
31
32
33
34
35
36
37
38
39
40
41
42
43
44
45
46
47
48
49
50
51
52
1
2
3
4
5
6
7
8
9
10
11
12
13
14
15
16
17
18
19
20
21
22
23
24
25
26
27
28
29
30
31
32
33
34
35
36
37
38
39
40
41
42
43
44
45
46
47
48
49
50
51
52

Page 2

potential for metastatic growth can produce severe patho-
logical consequences (7), and cestode diseases remain a
significant threat to our health and agriculture.
The notion of flatworms as representing the proto-bila-
terian condition promoted throughout most of the 20th
century has been difficult to dispel, and they continue to
be cited as such today. Wide adoption of the 18S-based
‘new animal phylogeny’ (Figure 1; 8,9) that showed them
to be members of the Lophotrochozoa (a diverse group
including annelid worms and molluscs that together with
the Ecdysozoa encompasses the spiralian animals) refuted
this notion, and their lophotrochozoan affinities have
been supported consistently by studies based on increas-
ingly large numbers of genes. Less support has been
found for their exact position within the Lophotrocho-
zoa, but they appear to have closer affinities to ‘platyzo-
an’ groups including rotifers and bryozoans than to
either annelids or molluscs (10). Based on their position,
there is no longer any a priori reason to assume them to
be representative of an early, or ‘primitive’, bilaterian
condition. Moreover, not only are flatworms a more
recently evolved animal lineage than previous ideas
suggested, but the parasitic flatworms form also a derived
clade (i.e. Neodermata; ‘new skin’) within the phylum,
having appeared after the major diversification of their
free-living cousins (11). We should expect then that
flatworm biology, including their genomes, will reflect
both their shared affinities to other lophotrochozoan
phyla and their unique, lineage-specific adaptations, such
as the maintenance of totipotent stem cells and adoption
of parasitism.
Phylogeneticstudies(11,12)
parasitism first arose through association (e.g. predation,
symbiosis) of free-living or symbiotic flatworms and fishes,
most likely early in the host’s evolution (13), and still
today, bony fish and elasmobranch (sharks, rays and
chimaeras) are host to sexual stages of the majority of
flatworm parasite families. Although free-living species
display a high propensity for symbioses spanning the spec-
trum from commensalism to parasitism, there is strong
evidence that the major parasitic lineages form a mono-
phyletic group, demonstrating that obligate parasitism
arose only once during the course of flatworm evolution
(11). This was associated with a major developmental shift
involving the separation of ontogenetically distinct larval
and adult stages, with replacement of the larval epidermis
by a syncytial tegument. Within this clade, we now recog-
nize four independent lineages: the cestodes (tapeworms),
digeneans (flukes) and monopisthocotylean and polyopis-
thocotylean ‘monogeneans’. Interrelationships of these
lineages remain controversial, but have begun to point
towarda sister relationship
digeneans, and paraphyly of the ‘Monogenea’ (11,14,15),
in contrast to previous hypotheses (and classifications)
that considered ‘monogeneans’ to be both monophyletic
and the sister group to tapeworms. The main implications
of the molecular-based hypotheses are a common origin
of both enteric parasitism and complex life cycles in
indicatethat obligate
betweencestodesand
Figure 1
Lophotrochozoa, and the derived positions of both the true tapeworms (Eucestoda) and the Cyclophyllidea, the order to which all cur-
rently sequenced tapeworm genomes belong (Topologies from 10,11,21,166).
14
Phylogenetic position of the tapeworm genera Echinococcus, Hymenolepis and Taenia. Note position of the flatworms within the
LOW RESOLUTION FIG
P. D. Olson et al.
Parasite Immunology
2
? 2011 Blackwell Publishing Ltd, Parasite Immunology, 33, 1–21
1
2
3
4
5
6
7
8
9
10
11
12
13
14
15
16
17
18
19
20
21
22
23
24
25
26
27
28
29
30
31
32
33
34
35
36
37
38
39
40
41
42
43
44
45
46
47
48
49
50
51
52
1
2
3
4
5
6
7
8
9
10
11
12
13
14
15
16
17
18
19
20
21
22
23
24
25
26
27
28
29
30
31
32
33
34
35
36
37
38
39
40
41
42
43
44
45
46
47
48
49
50
51
52

Page 3

tapeworms and flukes despite major differences in their
life histories, and that the first neodermatan flatworms
were nonenteric and direct-developing, as seen in contem-
porarymonopisthocotylean
parasites.
Only in the last two decades has our understanding of
tapeworm interrelationships begun to stabilize, thanks to a
more concerted effort on the part of cestodologists (16)
and the wide application of molecular phylogenetic
techniques (14). Circumscription of even the primary tape-
worm lineages has required major revisions to reflect new
insights into their affinities, resulting in the proposal of
three new tapeworm orders since 2008 (17,18). Interrela-
tionships of the 15 or more natural (i.e. monophyletic)
groups of tapeworms have yet to be resolved satisfactorily,
but it is clear that early branching lineages colonized a
wide spectrum of cartilaginous and bony fishes before
subsequent diversification led to the colonization of
homeothermic hosts(e.g.
Among the early branching groups, only the Diphyllobo-
thriidea [n.b. formally classified as a family of Pseudo-
phyllidea (18)] radiated into homeotherms, but retained its
association with fishes (which became 2nd intermediate
hosts) and transmission via aquatic life cycles (22). There
was thus a single primary colonization of homeothermic
hosts coincident with the adoption of fully terrestrial life
cycles that gave rise to the most speciose contemporary
group, the Cyclophyllidea.
The extent to which tapeworm–host associations were
shaped by the unique adaptive immunity of the mamma-
lian host is not clear from an evolutionary perspective.
However, even enteric infections are subject to immuno-
logical defences and nonpermissive hosts readily expel
foreign species (23), resulting in a high specificity of tape-
worms to their definitive hosts. In general, mammals act
as apex predators in tapeworm life cycles, playing host to
andpolyopisthocotylean
birds,mammals) (19–21).
adult, enteric stages. In the unique case of taeniid cyclo-
phyllideans, in which mammals also act as intermediate
hosts (24), they are the primary prey items of larger mam-
mals, such as in the rodent⁄fox cycles of Echinococcus,
Mesocestoides and some Taenia species (25). With regard
to human infection with tapeworms, there is at least some
evidence that the Taenia species infecting humans evolved
before the development of agriculture, animal husbandry
and the domestication of cattle and swine (24,26), indicat-
ing that humans were responsible for introducing Taenia
solium and T. saginata to contemporary agricultural cycles.
Moreover, phylogenetic analysis showed that these species
evolved in humans independently (26): T. solium associ-
ated with the tapeworms of hyenas and T. saginata with
those of lions. This unsettling scenario suggests that in
prehistoric times, food webs selected a role for ourselves
not only as definitive hosts, but also as intermediate hosts,
in transmission cycles including larger carnivores as the
apex predators.
PROGRESS IN CESTODE GENOMICS
Table 1
tapeworm genomes as represented by three taeniid and
one hymenolepidid cyclophyllidean species. At present, the
only published flatworm genomes are those of the human
bloodflukes Schistosoma mansoni (27) and S. japonicum
(28), but available draft data for the planarian model
Schmidtea mediterranea (29) and the ‘turbellarian’ Macro-
stomum lignano (30) provide important reference genomes
of free-living flatworms. By comparing parasitic and free-
living species, identification of both loss and expansion of
gene families will provide the most comprehensive picture
to date of the effects of evolving obligate parasitism,
allowing its signature to be compared with that in other
animal groups, such as the nematodes (31). Much of this
summarizesthe generalcharacteristicsof
Table 1 General characteristics of cestode genomes
Species (common name) Disease
Genome
size (Mb) No. of chromosomesaG⁄C (%)
N50 (Mb)
(no. of scaffolds) Institutions
Echinococcus multilocularis
(fox tapeworm)
Echinococcus granulosus
(dog⁄fox tapeworm)
Taenia solium (pork tapeworm) Neuro-cysticercosis
Alveolar echinococcosis 1132N = 18 (158–161)424Æ9 (1816) WTSI and UW
Cystic echinococcosis⁄
hydatid disease
1062N = 18 (159,162)421Æ9 (2874) WTSI and UM
117Unknown 2N = 18?
(43,163)
12 (164,165)
430Æ07 (6284)UNAM
Hymenolepis microstoma
(mouse bile duct tapeworm)
Rodent-hosted
laboratory model
184350Æ8 (5387)WTSI and NHM
NHM, Natural History Museum, London, UK; UM, University of Montevideo, Uruguay; UW, University of W?rzburg, Germany; WTSI,
Wellcome Trust Sanger Institute, Hinxton, UK.
aNumbers in brackets indicate citation from which chromosome estimates were obtained.
Volume 33, Number **, **, 2011Cestode genomics
? 2011 Blackwell Publishing Ltd, Parasite Immunology, 33, 1–21
3
1
2
3
4
5
6
7
8
9
10
11
12
13
14
15
16
17
18
19
20
21
22
23
24
25
26
27
28
29
30
31
32
33
34
35
36
37
38
39
40
41
42
43
44
45
46
47
48
49
50
51
52
1
2
3
4
5
6
7
8
9
10
11
12
13
14
15
16
17
18
19
20
21
22
23
24
25
26
27
28
29
30
31
32
33
34
35
36
37
38
39
40
41
42
43
44
45
46
47
48
49
50
51
52

Page 4

signature will surely relate factors evolved to counter host
immune defences, and comparative genomics thus hold
great promise for advancing the immunology of parasitic
flatworms.
Tapeworm genomes are small in size at ?110 Mb, com-
pared with 363 Mb in Schistosoma (27), 700 Mb in Sch-
midtea
and
?330–1100 Mb
www.genomesize.com/index.php
to the fact that tapeworm genomes contain fewer mobile
genetic elements and retroposons than trematodes or pla-
narians, in which they are common (32,33). However, it is
clear that there has also been significant gene loss. For
example, the components for de novo synthesis of choles-
terol are missing, as is ornithine decarboxylase (a key
enzyme in spermidine⁄putrescine biosynthesis), and these
essential components must therefore be acquired from the
host. Indeed, the complete loss of a gut has presumably
resulted in the loss of many enzymes. Similarly, highly
conserved developmental genes families such as Hox and
Wnt also show highly reduced numbers of gene classes
and orthologs, as discussed later. From a practical stand-
point, the small size of tapeworm genomes and minimal
amount of repetitive elements make their characterization
less problematic than other flatworms and aids in deter-
mining the structures and synteny of genes and other
genetic elements.
Later in the following sessions, we discuss the history
and state of play in ongoing initiatives. Full details of
these genomes will be discussed in an article being led by
Matt Berriman of the Parasite Genome Group at the
Wellcome Trust Sanger Institute (WTSI).
in
). Differences may be due
Macrostomum
(http://
1
Echinococcus multilocularis Leukart, 1863
An initial meeting to set priorities in pathogen genome
sequencing led by Rick Maizels (University of Edinburgh)
was held at the WTSI Genome Campus in March 2004.
E. multilocularis, the causative agent of AE, was chosen as
the reference system for all further cestode genome
projects (Table 1). Although infections caused by E. granu-
losus or T. solium are more prevalent worldwide, E. multi-
locularis was selected primarily because of the availability
of better laboratory cultivation techniques. During recent
years, several systems for efficient in vitro cultivation of
the E. multilocularis metacestode stage (34,35) as well as a
system for complete regeneration of metacestode vesicles
from totipotent parasite stem cells (36) have been estab-
lished, so that the life cycle of this cestode within the
intermediate host, from the initial infecting oncosphere to
the stage that is passed on to the definitive host, can now
be mimicked under controlled laboratory conditions. As a
source of genomic DNA, the natural parasite isolate JAVA
(37) was used, which is derived from a cynomolgus mon-
key (Macaca fascicularis) that was kept in a breeding
enclosure in the German Primate Center (Gçttingen) and
which was intraperitoneally passaged in laboratory mice
for a few months prior to DNA isolation. This step
appeared important because of the fact that long-term lab-
oratory ‘strains’ of larval cestodes (i.e. material that has
been passaged for years or decades within the peritoneum
of mice) usually undergo morphological and physiological
(and most probably also genomic) alterations that no
longer reflect the in vivo situation (1). To minimize
contamination with host DNA, it was further necessary to
isolate DNA from protoscoleces that had previously been
treated with pepsin at pH 2, leading to almost complete
digestion of host material but leaving parasite material
intact.
After extensive generation of bacterial artificial chromo-
somes libraries and determination of the parasite’s genome
size (36), a first round of conventional Sanger capillary
sequencing to ?4-fold coverage was carried out which was
complemented by several runs of paired and unpaired
454- and Solexa-sequencing. At the time of most analyses
presented here, sequence information representing 140-fold
coverage of the genome had been generated which, in
version 1 (13 August 2010), had been assembled into 600
supercontigs with an N50 contig size of more than 1Æ6 Mb
(n.b. in the latest assembly, half the genome is contained
in only 18 supercontigs; see Table 1). Thus, by combining
classicalcapillary sequencing
sequencing methodology, a data set has been produced for
the E. multilocularis genome that is more comprehen-
sive than those of the already published genomes of
S. mansoni, S. japonicum and B. malayi, which had not
been assembled into versions of <5000 contigs (38,39).
Interestingly, although the initial determination of the
E. multilocularis genome size by flow cytometry on iso-
lated parasite cells yielded values around 300 Mb (36), the
assembled sequence data strongly suggest a haploid gen-
ome size of ?110 Mb. The reason for this discrepancy is
currently unknown, but may represent a case of poly-
ploidy. However, in BLAST analyses of a set of several
thousand ESTs that are available for E. multilocularis
(40,41) and E. granulosus (41) against the genome assem-
bly, none could be identified that was not represented on
one of the 600 supercontigs. This indicates that at least
the protein-encoding portion of the genome is very well
covered by the latest assembly version, which is publicly
available via http://www.sanger.ac.uk/resources/downloads/
helminths/echinococcus-multilocularis.html.
In parallel to genome sequencing and assembly, tran-
scriptomes of different life cycle stages of E. multilocularis
are currently being characterized using next-generation
withnext-generation
P. D. Olson et al.
Parasite Immunology
4
? 2011 Blackwell Publishing Ltd, Parasite Immunology, 33, 1–21
1
2
3
4
5
6
7
8
9
10
11
12
13
14
15
16
17
18
19
20
21
22
23
24
25
26
27
28
29
30
31
32
33
34
35
36
37
38
39
40
41
42
43
44
45
46
47
48
49
50
51
52
1
2
3
4
5
6
7
8
9
10
11
12
13
14
15
16
17
18
19
20
21
22
23
24
25
26
27
28
29
30
31
32
33
34
35
36
37
38
39
40
41
42
43
44
45
46
47
48
49
50
51
52

Page 5

sequencing (NGS). Initial data sets are available at the
WTSI webpage of the E. multilocularis sequencing project
for isolated primary cells after one week of regeneration
(representing the early oncosphere–metacestode transition;
36), for in vitro cultivated metacestode vesicles and for
protoscoleces prior to or after activation by low-pH⁄pep-
sin treatment, which mimics the transition into the
definitive host. Further RNA sequencing is carried out for
regenerating primary cells after three weeks of culture (late
phase of oncosphere–metacestode transition), for metaces-
tode vesicles with brood capsules (early formation of
protoscoleces) and for the adult stage. Thus, transcriptome
data that almost completely cover the E. multilocularis life
cycle will soon be available, although it will still be
difficult to obtain material of activated E. multilocularis
oncospheres in amounts that are sufficient for RNA
sequencing.
Using the available transcriptome data as well as a
large set of E. multilocularis and E. granulsous EST infor-
mation(availableunder http://www.nematodes.org/Ne-
glectedGenomes/Lopho/LophDB.php, http://fullmal.hgc.jp
/em/docs/echinococcus.html and http://www.sanger.ac.uk/
resources/downloads/helminths/echinococcus-multilocular-
is.html), gene prediction and annotation is currently
under way. In a first, AUGUSTUS-based analysis of the
assembled genome, we identified ?11 000 protein-encod-
ing genes, which is slightly less than the gene number
(11 800) that has been predicted for the trematode Schis-
tosoma mansoni (27). For 70% of these genes, we could
identify clear orthologs in other organisms, whereas the
remaining 30% are most probably Echinococcus- or ces-
tode-specific genes or gene families.
Echinococcus granulosus (Batsch, 1786)
Mostly for comparative studies with the Echinococcus mul-
tilocularis reference genome, NGS has very recently also
been used for a first characterization of the genome of
E. granulosus. This project is being carried out by the par-
asite genomics group of the WTSI led by Matt Berriman
in collaboration with Cecilia Fernandez (University of
Montevideo). Becauseof
infections, the G1 (sheep) strain was chosen for sequenc-
ing and, like in the case of E. multilocularis, protoscoleces
after treatment with low pH⁄pepsin were used as a source
for genomic DNA to minimize host contamination
(C. Fernandez, pers. comm.). After a first round of Illu-
mina sequencing, the genome has been assembled into
5200 contigs that, using the E. multilocularis genome as a
reference framework, have been further assembled into
?2000 scaffolds that are available via http://www.san-
ger.ac.uk/resources/downloads/helminths/echinococcus-gra
its importancein human
nulosus.html. As expected, the genomes of E. granulosus
and E. multilocularis are highly homologous with overall
96% identity at the nucleotide sequence level within the
coding regions of predicted genes, and still around 91%
identity in promoter regions. Because the E. granulosus
contigs have been assembled into supercontigs using
E. multilocularis as a reference, no valid conclusions
concerning genomic rearrangements between the species
can been made at present. Direct comparisons of longer
contigs of the E. granulosus genome assembly with the
E. multilocularis sequence, however, indicate that there is
also a high level of synteny between both species. Differ-
ences in gene structure and sequence can mostly be
observed in the case of expanded gene families, such as
the recently described hsp70 family (42) that contains a
significant number of pseudogenes. The E. granulosus gen-
ome assembly is currently awaiting additional Illumina
data, and thus, substantial improvement is expected soon.
Taenia solium L., 1758
A third important project on a taeniid cestode addresses
the whole genome of T. solium (43) and is being carried
out by a Mexican consortium directed by Juan-Pedro Lac-
lette (http://bioinformatica.biomedicas.unam.mx/taenia/)
located at the Universidad National Autonoma de Mexico.
As in the case of the E. multilocularis genome, this project
has followed a hybrid strategy in which classical capillary
sequencing of cloned genome fragments has been com-
bined with NGS. In a first phase of the project, ?20 000
ESTs from adult worms and cysticerci were generated, fol-
lowed by estimation of the parasite’s genome size. Using
genomic DNA from cysticerci as a source for analysis, 2·
coverage by capillary sequencing and 5· coverage by 454
sequencing have been reached, and the hybrid assembly
process has so far yielded ?50 000 contigs (N50 >
5000 bp) that cover ?90% of the EST-based transcriptome
profile. Additional 454- and Solid-reads are planned in
this project so that a much more comprehensive assembly
will soon be available. Furthermore, because EST informa-
tion and next-generation transcriptome data from Echino-
coccus spp. are informative for identifying genes in Taenia
spp. (and vice versa), close collaboration of the bioinfor-
matic teams that work on all three ongoing taeniid cestode
genome projects has been established that should greatly
facilitate the annotation process. Interestingly, as in the
case of E. multilocularis, the haploid genome size of T. so-
lium was first determined to be ?260 Mb using flow
cytometry, whereas the NGS assembly so far indicates a
genome size of 130 Mb (43). Whether this is, in both
cases, associated with genome duplications or polyploidy
remains to be elucidated.
Volume 33, Number **, **, 2011Cestode genomics
? 2011 Blackwell Publishing Ltd, Parasite Immunology, 33, 1–21
5
1
2
3
4
5
6
7
8
9
10
11
12
13
14
15
16
17
18
19
20
21
22
23
24
25
26
27
28
29
30
31
32
33
34
35
36
37
38
39
40
41
42
43
44
45
46
47
48
49
50
51
52
1
2
3
4
5
6
7
8
9
10
11
12
13
14
15
16
17
18
19
20
21
22
23
24
25
26
27
28
29
30
31
32
33
34
35
36
37
38
39
40
41
42
43
44
45
46
47
48
49
50
51
52

Page 6

Hymenolepis microstoma (Dujardin, 1845)
Hymenolepis microstoma, the mouse bile duct tapeworm,
is one of three beetle⁄rodent-hosted hymenolepidid labora-
tory models commonly used in research and teaching since
they were first domesticated in the 1950s. Although less
studied than either the rat tapeworm H. diminuta (44) or
the dwarf tapeworm H. nana, H. microstoma has advanta-
ges of being small and mouse-hosted unlike H. diminuta
and is refractory to both human infection and cross-con-
tamination of rodents via a direct life cycle, unlike
H. nana. Use of H. microstoma has thus both practical
and regulatory advantages that make a good model for
research requiring easy access to both larval and strobilate
stages of the tapeworm life cycle. Although we expect the
genome of H. microstoma to be representative of all three
model species, it is worth noting that they are not each
other’s closest relatives (45) and that there has long been
disagreement as to whether or not Hymenolepis spp. bear-
ing hooks should be classified in their own genus (i.e.
Rodentolepis) (see 46). If so, then we expect H. microstoma
to be better representative of H. nana than to H. diminuta.
Genome characterization of H. microstoma began in
2009 as a pilot project in collaboration with the Sanger
Institute after their implementation of NGS allowed for
expansion of existing genome sequencing programmes.
Although Hymenolepis tapeworms are not significant
human pathogens, they represent an important laboratory
model in cestodology and access to a highly inbred culture
made them well suited for de novo genome assembly. Geno-
mic and transcriptomic data are based on specimens of a
‘Nottingham’ strain maintained by the author (PDO) in vivo
using natural hosts (flour beetles, Tribolium confusum, and
BALB⁄c mice). The origin of the strain can be traced back
to the original domestication of the species by the C. P.
Read laboratory at Texas Rice University in the 1950s (47),
making the genome data directly relevant to a large body of
previous research based on isolates of this strain. A com-
plete description of the strain and it origins, including a
review of its general biology and use as a laboratory model,
has been recently published in open access format (46).
The H. microstoma genome assembly consists entirely of
data generated via NGS technologies and has been assem-
bled and analysed using bioinformatic pipelines developed
by the Parasite Genomics Group at the WTSI (48–53) and
others(54–57). Thecurrent
comprises data from six full Roche 454 Titanium runs
(three unpaired runs, two paired runs with 3–4-kb inserts,
and one with 9-kb inserts) and three Illumina Solexa lanes
(76-bp reads, two lanes with 250-bp inserts, and one lane
with 3-kb inserts). The combined data resulted in more
than 40· coverage of the estimated 147-Mb genome
assembly (April2011)
(Table 1). Separate de novo assemblies of the two technolo-
gies were made using the software NEWBLER 2.5 (58) (for
Roche⁄454) and ABYSS 1.2.1 (55) (for Illumina), and con-
tigs then merged using the pipe-line GARM (A. Sanchez,
unpubl. data), based on the genome assembler Minimus
(59). Remaining gaps were closed with IMAGE (dev. vs.
(48) for 20 iterations with gradually more permissive
parametersettings (kmer = 61–30, overlap = 100–200).
The final sequences were corrected using five iterations of
iCORN (dev. ver.) (49). Genome data are made available
fromhttp://www.sanger.ac.uk/resources/downloads/helm-
inths/hymenolepis-microstoma.html.
Transcriptomic data are also being profiled using Illu-
mina technologies for the purposes of RNA-seq analysis
and annotation, as well as to address specific questions in
adult development. Presently, this includes whole adult
cDNA from the mouse gut, and thus profiles all grades of
development represented by the strobilate adult worm, as
well as cDNA from a combined developmental series of
metamorphosing larvae (i.e. 3–7 days PI) from the haemo-
coel of beetles. Additional cDNA samples representing
progressively mature regions of the adult tapeworm stro-
bila are being sequenced by the WTSI, and each sample
will be replicated multiple times for statistical support.
This will allow us to determine differential expression
associated with the process of segmentation in the neck
region, the maturing of the reproductive organs in the
strobila and the process of embryogenesis occurring in
gravid segments.
Unlike E. multilocularis and E. granulosus, the H. micros-
toma genome assembly has not undergone manual curation
or refinement and is thus a good example of the kind of
assembly that can be achieved using medium-coverage NGS
and bioinformatics alone. For comparative purposes, com-
pleteness was assessed using CEGMA 2.0 (60), which looks for
a set of 458 ‘core’ genes that are highly conserved in eukary-
otes. This method estimated the H. microstoma genome
assembly to be 90% complete, compared to 87–93% in Echi-
nococcus species, and demonstrates that genome projects on
a medium scale, with restricted coverage and without man-
ual curation, are feasible and can give excellent estimates of
gene content. Moreover, as some percentage of these ‘core’
genes will have been lost from the reduced genomes of para-
sitic flatworms, estimating completeness on this basis
almost certainly provides an underestimation. Indeed, the
very high sequence coverage of the current cestode genome
assemblies suggests that tapeworms have simply lost ?7 to
10% of these ‘core’ genes. The biggest difference between
the H. microstoma and E. multilocularis assemblies is seen
in the scaffold-statistics: more than 50% of the E. multilocu-
laris genome is contained in 13 scaffolds in the latest
assembly(N50;Table 1),
2
)
whereas
H. microstoma
is
P. D. Olson et al.
Parasite Immunology
6
? 2011 Blackwell Publishing Ltd, Parasite Immunology, 33, 1–21
1
2
3
4
5
6
7
8
9
10
11
12
13
14
15
16
17
18
19
20
21
22
23
24
25
26
27
28
29
30
31
32
33
34
35
36
37
38
39
40
41
42
43
44
45
46
47
48
49
50
51
52
1
2
3
4
5
6
7
8
9
10
11
12
13
14
15
16
17
18
19
20
21
22
23
24
25
26
27
28
29
30
31
32
33
34
35
36
37
38
39
40
41
42
43
44
45
46
47
48
49
50
51
52

Page 7

contained in 747 scaffolds. Besides
E. multilocularis genome has more long-range mapping
information and has undergone several rounds of dedicated
manual curation to join scaffolds and resolve miss-assem-
blies because of repeat elements or heterozygosity. The dif-
ference in genome coverage is negligible for most research
questions, such as those that primarily make use of gene
sequence information and expression data, but could be
problematic for research requiring long-range mapping
information.
3
more read depth, the
TOWARD IMPROVED CHEMOTHERAPY
The drugs most frequently employed in the treatment for
cestode infections are praziquantel (PZQ) and benzimidaz-
oles (BZs; e.g. albendazole, mebandazole). PZQ, which is
well known for its activity against adult schistosomes, is
also a highly potent drug against cestode adult stages and
is frequently used to treat taeniasis, or is employed in
deworming campaigns against foxes or dogs in endemic
areas (61). Although the precise cellular target(s) for PZQ
in schistosomes are not yet known, voltage-gated calcium
channels are considered very good candidates and have
thus already been experimentally addressed using the
Xenopus oocyte expression system (62). Interestingly,
unlike other organisms, schistosomes express two different
b subunits of calcium channels, one of which confers PZQ
sensitivity in the Xenopus system, the other not (63). A
major difference between these subunits is the presence or
absence of two canonical serine residues in the so-called
beta interactiondomain
phosphorylated through protein kinase C (PKC). In the
case of the b subunit that conferred PZQ sensitivity, these
residues were replaced by amino acids that can no longer
be phosphorylated by PKC, and this difference might be
the structural reason for the general PZQ sensitivity of
schistosomes (63). Recently, Jeziorski and Greenberg (64)
also identified calcium channel b subunits in T. solium and
demonstratedthatthiscestode,
expresses an unusual subunit in which the PKC target resi-
dues were replaced by Asp and Ala, alongside a canonical
subunit with Thr⁄Ser residues at these positions. In the
ongoing sequencing projects, this could be verified for all
four cestode species under study. Both Echinococcus
species and H. microstoma, like T. solium, express two b
subunits of calcium channels of which one represents the
canonical form and the second a modified version with
amino acid replacements at the PKC responsive sites (data
not shown). This could, at least in part, explain the PZQ
sensitivity of adult cestodes. Although PZQ resistance will
most probably never be an issue in the treatment of
taeniasis patients, it could become a problem in large scale
(BID)thatare typically
likeschistosomes,
deworming campaigns against E. multilocularis, E. granu-
losus and Mesocestoides spp. that have been suggested
already for parts of Central Europe and China (25,65,66).
Particularly for such projects, genetic information on the
cellular targets of PZQ, as available through the genome
projects, will be highly valuable in assessing treatment
efficacy and the emergence of drug resistance.
In sharp contrast to its activity on adult cestodes, PZQ
has very limited effects on metacestode stages (67). The
underlying reason could be that the calcium channel b
subunits (or other potential PZQ targets) are expressed in
an adult-specific manner, and in the currently available
transcriptome profiles for E. multilocularis metacestode
vesicles, the respective genes are indeed expressed at a
marginal level (data not shown). Because of the low
efficacy of PZQ treatment, the current drugs of choice in
chemotherapy against AE, CE and NCC are BZs that
havea high affinity for
isoforms, thus inhibiting microtubule polymerization that
eventually leads to parasite death. Although prolonged BZ
treatment of the intermediate host can be effective in elim-
inating E. granulosus cysts or T. solium cysticerci (68,69),
its activity against E. multilocularis is very limited. In AE,
BZ treatment is mostly parasitostatic rather than parasito-
cidal and, as a consequence, has to be given lifelong (68).
Furthermore, in all three types of infection, BZ treatment
can be associated with severe side effects that are because
of limited bioavailability of the drug at the site of infection
and high structural homology of b-tubulin of parasite and
host. Three major b-tubulin isoforms that are expressed
by
E. multilocularis
have
several years ago and were shown to be highly homolo-
gous (>90% amino acid identity) to b-tubulin of humans
(40; Table 2). In the E. multilocularis genome assembly, we
have identified at least nine b-tubulin encoding loci,
although transcriptome profiling clearly shows that the
three previously identified isoforms (40) are abundantly
expressed in all larval stages, whereas the other six loci are
mostly silent or may even represent pseudogenes. Studies
on mechanisms of BZ resistance and sensitivity in nema-
todespreviously identified
(Phe200 and Phe167 in BZ-sensitive isoforms) that are
particularly important for drug binding to b-tubulin. In
BZ-resistant strains of
Haemonchus
residues were frequently exchanged by Tyr or His, leading
to diminished BZ binding (70). In this context, it is inter-
esting to note that the b-tubulin isoform which, according
to transcriptome profiling, shows the highest expression
level in the E. multilocularis metacestode (i.e. the target of
BZ treatment) displays Tyr residues at positions 200 and
167 and might thus represent a potentially BZ-resistant
isoform (Table 2). Highly homologous isoforms with Tyr
helminth-specific
b-tubulin
alreadybeencharacterized
twoaminoacid residues
contortus,these
Volume 33, Number **, **, 2011Cestode genomics
? 2011 Blackwell Publishing Ltd, Parasite Immunology, 33, 1–21
7
1
2
3
4
5
6
7
8
9
10
11
12
13
14
15
16
17
18
19
20
21
22
23
24
25
26
27
28
29
30
31
32
33
34
35
36
37
38
39
40
41
42
43
44
45
46
47
48
49
50
51
52
1
2
3
4
5
6
7
8
9
10
11
12
13
14
15
16
17
18
19
20
21
22
23
24
25
26
27
28
29
30
31
32
33
34
35
36
37
38
39
40
41
42
43
44
45
46
47
48
49
50
51
52

Page 8

at these two positions are also encoded by the genomes of
E. granulosus and T. solium (Table 2), and in the respective
EST databases, transcripts for this isoform are particularly
abundant (data not shown), indicating high expression in
the metacestodes of these species as well. Hence, limited
bioavailability of the drug at the site of infection, which is
particularly an issue for the infiltratively growing E. multi-
locularis metacestode, combined with a potentially reduced
affinity of BZs to the major b-tubulin isoform of the met-
acestode, could be the main reasons for limited efficacy of
BZ treatment in AE.
Employing in vitro cultivation systems for the E. multiloc-
ularis metacestode stage and classical approaches of testing
selected compounds for anti-parasitic activities, Andrew
Hemphill’s laboratory and others (71) have recently identi-
fied several compounds such as nitazoxanide, isoflavones or
amphotericin B that could be used as drugs in AE
treatment, mostly in combination with BZs (reviewed in
68). However, compounds that act not only parasitostatic
but truly parasitocidal against E. multilocularis in vivo have
not been discovered to date, indicating that new chemother-
apeutic strategies against AE are urgently needed. With the
availability of the E. multilocularis whole genome together
with those of E. granulosus and T. solium, targeted drug
design should be one of the most promising approaches for
the development of anti-cestode drugs in the next years. On
the one hand, comparative genomics can be employed to
identify factors that are unique to cestodes or flatworms
and could serve as targets for compound screening. The
drawback of this approach is that the function and
biochemical properties of parasite-specific factors are
usually unknown, which severely hampers the design of
efficient inhibitors. Furthermore, many of these parasite-
specific proteins have redundant functions and are often not
essential.Analternativeand muchmorepromising
approach should rather concentrate on drug targets that
are, to a certain degree, homologous between parasite and
host, thus providing information on function and biochem-
istry, but that display sufficient functional modification
between both species to allow the development of parasite-
specific inhibitors. A highly promising group of factors in
this regard are protein kinases (Table 3) that are crucially
involved in the regulation of metazoan development and
that mediate cell–cell communication by participating in
cellular signalling systems (72). Because of their important
role in cancer, the general biochemistry of these proteins is
extremely well studied and a plethora of compounds to
modify their activities, mostly directed against the well-con-
served ATP-binding pocket, is available (73). Protein kinas-
es have thus already been suggested as promising targets in
drug design against schistosomiasis (74), and their suitabil-
ity as targets in cestodes has recently been demonstrated by
Gelmedin et al. (75) who identified pyridinyl imidazoles,
directed against the p38 subfamily of mitogen-activated
protein kinases (MAPK), as a novel family of anti-Echino-
coccus compounds. A number of E. multilocularis protein
kinases such as the Erk- and p38-like MAPKs EmMPK1
(76) and EmMPK2 (75), respectively, the MAPK kinases
EmMKK1 and EmMKK2 (77), or the Raf-like MAPK
kinase kinase EmRaf (78) have already been characterized
on the molecular and biochemical level, and particularly in
the case of the two MAPKs, functional biochemical assays
have been established that can be used for compound
screening (75,76). Of further interest are already character-
ized receptor kinases of the insulin- (EmIR; 79), the epider-
mal growth factor- (EmER; 80) and the transforming
growth factor-b- (EmTR1; 81) receptor families that are
expressed by the E. multilocularis metacestode stage and
that are involved in host–parasite cross-communication by
interacting with the evolutionary conserved cytokine- and
hormone-ligands that are abundantly present in the
Table 2 Structural features and expression of b-tubulin isoforms
of Echinococcus multilocularis, Echinococcus granulosus and Taenia
solium
Gene Em.167 Em.200 Eg.167 Eg.200 Ts.167 Ts.200
Expr.MC
(%)
tub-1
tub-2
tub-3
Phe
Tyr
Phe
Phe
Tyr
Phe
Phe
Tyr
Phe
Tyr
Tyr
Tyr
Tyr
Tyr
Tyr
Tyr
Tyr
Phe
21
100
26
Nomenclature of the main b-tubulin isoforms expressed by ces-
todes was made according to Brehm et al. (40). Amino acid resi-
dues encoded at positions 167 and 200 are given for the
orthologous genes in E. multilocularis (Em), E. granulosus (Eg)
and T. solium (Ts). ‘Expr.MC’ refers to the relative expression
level in the E. multilocularis metacestode stage according to
RNA-seq data.
Table 3 Numbers
multilocularis genome that encode members of
druggable enzyme families
of genespresent inthe
Echinococcus
particularly
Enzyme familyNo. of genes
Protein kinases
Phosphatases
Peptidases
Ligand-gated ion channels
GPCR (Rhodopsin family)
GPCR (Secretin family)
Nuclear hormone receptors
Glycosyl hydrolases
Glycosyl transferases
250
47
113
9
44
3
17
24
37
GPCR, G-protein-coupled receptors.
P. D. Olson et al.
Parasite Immunology
8
? 2011 Blackwell Publishing Ltd, Parasite Immunology, 33, 1–21
1
2
3
4
5
6
7
8
9
10
11
12
13
14
15
16
17
18
19
20
21
22
23
24
25
26
27
28
29
30
31
32
33
34
35
36
37
38
39
40
41
42
43
44
45
46
47
48
49
50
51
52
1
2
3
4
5
6
7
8
9
10
11
12
13
14
15
16
17
18
19
20
21
22
23
24
25
26
27
28
29
30
31
32
33
34
35
36
37
38
39
40
41
42
43
44
45
46
47
48
49
50
51
52

Page 9

intermediate host’s liver (1,72). In total, we could thus far
identify ?250 protein kinase-encoding genes on the genome
assembly versions of E. multilocularis (Table 3) and E. gran-
ulosus, the majority of which displays considerable homolo-
gies to orthologous genes in schistosomes, which could be
particularly important for the design of compounds that
have a broad spectrum of activity not only against cestodes
but also against other parasitic flatworms.
An important issue in rational drug design is not only
the identification of targets that display structural and
functional differences between the respective parasite and
host components, thus ensuring that compounds with suf-
ficient parasite specificity can be found, but also the gen-
eral ‘druggability’ of the target, i.e., whether it contains
structural features that favour interactions with small mol-
ecule compounds (82). Apart from protein kinases, several
other protein families such as G-protein-coupled receptors
(GPCR) or ligand-gated ion channels proved to be partic-
ularly druggable in previous compound screens and che-
mogenomic approaches (83). For a selection of protein
families that are particularly suitable as drug targets,
Table 3 lists the number of coding genes that we have
identifiedusingthecurrent
assembly. In addition to a large number of protein kinases,
several of which are already under study in the E. multi-
locularis system (72), and nuclear hormone receptors,
which have been characterized in cestodes very recently
(84), the list also contains multiple peptidases, phosphata-
ses, GPCRs and ligand-gated ion channels which have, so
far, been characterized to a certain degree in schistosomes
(85), but never in tapeworms. Based on the criteria like
expression strength, essentiality, involvement in multiple
metabolic pathways, assayability and druggability, Crow-
ther et al. (86) recently established a highly interesting
in silico approach to prioritise parasite proteins for tar-
geted drug design and, in the case of S. mansoni, pre-
sented a list of particularly promising candidates such as
Na+⁄K+-ATPase, transketolase, vacuolar proton ATPases
and a number of additional protein and enzyme compo-
nents. Once gene annotation for E. multilocularis is fin-
ished and more extensive data on the larval transcriptome
are available, similar approaches are also possible for this
species and can, by comparative genomics, also be applied
to E. granulosus and T. solium.
Taken together, all technical and methodological prereq-
uisites for targeted drug design against larval cestodes
should soon be (or are already) available. Once suitable
targets are identified by in silico approaches, respective
small molecule lead compounds can be tested for anti-par-
asitic activity using the established in vitro cultivation sys-
tems for the E. multilocularis metacestode (87) and stem
cell systems(1). Asan
E. multilocularis
genome
importantcomplementary
approach, the essentiality of the target components can be
tested using RNA interference (RNAi) assays that have
been established very recently for regenerating E. multiloc-
ularis primary cells (88) and protoscoleces (89). On the
basis of the identified lead compounds and libraries of
related molecules, parasite-specific drugs can subsequently
be identified in comparative host- and parasite cell cultiva-
tion systems and eventually be tested in vivo in well-estab-
lished animal models for secondary AE. Based on the
considerable homologies between all taeniid cestodes, it is
highly likely that all identified anti E. multilocularis drugs
will be also active against E. granulosus and T. solium.
PARASITE ANTIGENS AND
IMMUNOMODULATORY MOLECULES
Larval
T. solium induce chronic, long-lasting infections during
which the host immune system is modified in various
ways through surface components of the metacestode
stage (e.g. the acellular ‘laminated layer’ of Echinococcus
species) or by excretory⁄secretory (E⁄S) products (90,91).
In all three cases, the induction of Th2-dominated
immune responses is observed in intermediate hosts that
are highly susceptible to an infection, and a picture is
beginning to emerge that, as in helminth infections
because of nematodes and trematodes, regulatory T cells
and alternatively activated macrophages might play a
crucial role in suppressing antiparasitic immune responses
(91,92). Although little is known on the molecular nature
of taeniid cestode E⁄S products with immunomodulatory
activities, previous investigations at least identified a
number of parasite antigens or laminated layer compo-
nents that might be involved in deviating or dampening
the immune response (reviewed by Gottstein & Hemphill;
93). With the availability of cestode genome sequences,
several important questions concerning immunomodula-
tory factors can now be addressed from a genomic
perspective, and in the following, we will present some
initial analyses on important antigen families and mole-
cules that arelikelyto
properties.
stages of
E. multilocularis,
E. granulosus
and
haveimmunomodulatory
The Echinococcus antigen B gene family
Undoubtedly, the most studied factor in Echinococcus is
the so-called antigen B (AgB), a highly immunogenic lipo-
protein and major component of hydatid cyst fluid (94).
Although there are several reports on immunomodulatory
properties of AgB in vitro (94), and biochemical investiga-
tions that demonstrate binding of different hydrophobic
ligands to AgB (95), the precise function of this protein in
Volume 33, Number **, **, 2011 Cestode genomics
? 2011 Blackwell Publishing Ltd, Parasite Immunology, 33, 1–21
9
1
2
3
4
5
6
7
8
9
10
11
12
13
14
15
16
17
18
19
20
21
22
23
24
25
26
27
28
29
30
31
32
33
34
35
36
37
38
39
40
41
42
43
44
45
46
47
48
49
50
51
52
1
2
3
4
5
6
7
8
9
10
11
12
13
14
15
16
17
18
19
20
21
22
23
24
25
26
27
28
29
30
31
32
33
34
35
36
37
38
39
40
41
42
43
44
45
46
47
48
49
50
51
52

Page 10

the biology of Echinococcus or in the immune response
duringechinococcosis is
described as a 160 kDa lipoprotein, AgB was later shown
to be built up of several 8 kDa monomers that are
encoded by a gene family (96), and since the first full
description of an AgB-encoding gene by Frosch et al. (97),
there is a constant debate on how many of these genes are
actually expressed in these parasites. By studies of Fernan-
dez et al. (98), Chemale et al. (99), Arend et al. (100) and
Mamuti et al. (101), the number of AgB subunit genes
had grown to five in 2007 (named EmAgB1-EmAgB5 in
E. multilocularis and EgAgB1-EgAgB5 in E. granulosus),
whereas genomic Southern blot analyses indicated that
there are at least seven loci (102). Studies by Haag et al.
(103) and Arend et al. (100) even suggested the presence
of further AgB genes (up to 10 in E. granulosus and up to
110 copies in the related E. ortleppi) as well as a high
degree of genetic polymorphism among those genes (even
within protoscoleces that derived from one single cyst).
These authors proposed that numerous AgB copies might
be involved in gene conversion mechanisms through
recombination processes and DNA rearrangements similar
to the situation in protozoans such as Plasmodium sp. or
trypanosomes (103). This theory was recently contradicted
by Zhang et al. (104) who characterized AgB genes in
E. granulosus isolates from different geographic origin and
proposed the presence of 10 unique genes (or alleles) that
are, however, highly homologous between these isolates
and did not show gross polymorphisms. To shed more
light on the situation, we have analysed the presence and
location of AgB genes in the current assemblies of the
E. multilocularis and E. granulosus genomes. As described
by Brehm (72), using the first assembly version of the
E. multilocularis genome (19 000 contigs), a total of seven
AgB loci appears to form a cluster on a distinct region of
the genome. In the latest genome version (600 supercon-
tigs), all these copies are now assembled into one continu-
ous sequence fragment of 57 kbp that is present on
still unknown.Originally
scaffold_29 (Figures 2 and 3). The antigen B cluster is
flanked by two genes, EmLDLR and EmMTA, which are
highly conserved among cestodes. The gene product of
one of these, EmLDLR, displays significant homologies to
low-density lipoprotein (LDL) receptors from other spe-
cies and contains one single class A LDL receptor
domain. The second encodes a factor with considerable
homologies (50% identical, 66% similar residues) to the
human ‘metastasis-associated-protein’ MTA3 which is a
component of the nucleosome-remodelling and histone-de-
acetylase complex (105) and, like the human protein,
contains one BAH (bromo-adjacent homology) domain,
one GATA-type zinc finger domain and one classical zinc
finger domain (data not shown). As previously suggested
(72), the antigen B cluster is formed of one copy each of
AgB1, AgB2, AgB4 and AgB5, two identical genes encod-
ing AgB3 and one slightly altered AgB3 gene (AgB3’).
The only difference to the previously suggested cluster
organization (72) is that in the newest assembly version
the AgB5 locus and one AgB3 locus have changed
position (Figure 2). All genes of the cluster display the
typical organization (103) of two exons, with a signal pep-
tide encoded by exon 1, separated by a small intron. Tran-
scriptome analyses on in vitro cultivated metacestode
vesicles further indicate that AgB1 is, by far, the most
abundantly expressed isoform, followed by AgB3’ (20% of
the expression level of AgB1) and AgB3 (10%). Only mar-
ginal expression could be detected for AgB2, AgB4 and
AgB5 in the metacestode, and likewise, almost no expres-
sion was measured for any AgB isoform in the protoscolex
(data not shown).
In E. granulosus, the situation appears to be highly
similar to E. multilocularis (Figure 2). Within a region of
approximately the same size as in E. multilocularis, close
orthologs of EmLDLR (EgLDLR) and EmMTA (EgMTA)
are present and are flanking a cluster of seven loci with
one copy each of AgB1, AgB2, AgB4 and AgB5, as well as
three slightly differing copies of AgB3 (AgB3-1, AgB3-2,
Figure 2
fold_29 (positions 624.079–681.301) of the E. multilocularis genome (above) and supercontig_30740 (680.925–741.961) of the current
E. granulosus assembly (below). Arrows indicate the location and direction of transcription of AgB isoforms and Echinococcus orthologs of
the conserved genes LDLR and MTA. Regions of the E. granulosus assembly that are not yet covered by sequencing data are marked by
‘N’.
15
The Echinococcus multilocularis and Echinococcus granulosus antigen B clusters. Displayed is the genomic organization on scaf-
LOW RESOLUTION FIG
P. D. Olson et al.
Parasite Immunology
10
? 2011 Blackwell Publishing Ltd, Parasite Immunology, 33, 1–21
1
2
3
4
5
6
7
8
9
10
11
12
13
14
15
16
17
18
19
20
21
22
23
24
25
26
27
28
29
30
31
32
33
34
35
36
37
38
39
40
41
42
43
44
45
46
47
48
49
50
51
52
1
2
3
4
5
6
7
8
9
10
11
12
13
14
15
16
17
18
19
20
21
22
23
24
25
26
27
28
29
30
31
32
33
34
35
36
37
38
39
40
41
42
43
44
45
46
47
48
49
50
51
52

Page 11

AgB3-3). Although care has to be taken in suggesting
complete synteny between both species in this region,
because of the fact that the single E. granulosus contigs
(flanked by ‘N’ in Figure 2) have been assembled into
supercontigs using the E. multilocularis sequence as a refer-
ence, at least the E. granulosus copies of AgB1, AgB4 and
AgB3-2 are clearly assembled into one contig and display
the same gene order and transcriptional orientation as in
E. multilocularis (Figure 2). This makes it highly likely that
the genome arrangement as suggested for E. granulosus in
Figure 2 reflects the true situation. Apart from the AgB
cluster, we could not detect any AgB-related sequences
elsewhere in the genomes of E. multilocularis and E. granu-
losus, with one notable exception of an AgB-like gene on
E. multilocularis
scaffold_7,
represented in EST databases, does not show a detectable
transcription profile in RNA-seq data, contains inactivat-
ing mutations within the reading frame (data not shown),
and thus most likely represents a pseudogene.
Taken together, the apparently high level of homology
and synteny within the AgB clusters of E. multilocularis
and E. granulosus, and the absence of functional AgB
copies outside these clusters, does not support the theory
that this region is a hot spot for genomic rearrangements.
Furthermore, the structure as depicted in Figure 2 clearly
supports previous data on the occurrence of just five
distinct subfamilies of AgB genes (101) and the presence
of seven distinct bands in Southern blot analyses under
low-stringency conditions (102). The gross discrepancies
between the genomic situation around the AgB clusters of
E. granulosus and E. multilocularis and previous reports
on very high copy numbers of the AgB genes in Echino-
coccus protoscoleces (100,103) are difficult to explain at
present. On the one hand, Arend et al. (100) and Haag
that is,however, not
et al. (103) exclusively relied on PCR-based methodology
to estimate the numbers of AgB genes in isolated parasite
material which, because of the amplification process,
might be prone to significant errors. On the other hand,
because of an as yet unknown mechanism, these genes
could be amplified as extra-chromosomal DNA aggregates
that might have slipped the genome assembly process.
Finally, because the highest number of AgB copies was
detected in laboratory material of E. ortleppi (103), this
species might significantly differ from E. multilocularis and
E. granulosus concerning the AgB cluster. In future stud-
ies, it might thus be worthwhile to also characterize the
E.ortleppi AgB cluster and the surrounding genomic
regions.
Interestingly, when analysing the current Hymenolepis
genome assembly, we also identified four AgB-related
genes (on contigs 10534, 20275, 23242 and 25502) with a
typical exon–intron structure (Figure 3), suggesting that
the AgB family is not taeniid cestode specific but occurs
in a wide variety (if not all) cestodes. Unfortunately, the
H. microstoma assembly used at the time of analysis was
too fragmented to determine whether the AgB genes are
also clustered in this species. However, the most recent
version of its genome, and targeted analyses of additional
cestode genomes using sequence information of the con-
served LDLR and MTA genes, should provide valuable
information to further dissect the evolution of the Echino-
coccus AgB cluster.
The host protective oncosphere antigens
The prototype of another highly interesting taeniid cestode
gene family is encoding the oncospheral antigen EG95
which has been successfully used in vaccination trials
Figure 3
Displayed is a CLUSTALW alignment of amino acid sequences encoded by exon 1 (signal peptide) and exon 2 of AgB isoforms of E. mul-
tilocularis (EmAgB1–EmAgB5), E. granulosus (EgAgB1–EgAgB5), and four AgB-like proteins of H. microstoma encoded on contigs 10534,
20275, 23242 and 25502 of the current genome assembly version. Highly conserved residues are printed in white on black background, resi-
dues with similar biochemical function and printed in black on grey background. The signal peptide region is indicated by asterisks below
the alignment. Note that the signal peptides of EgAgB3-2, Hm20275 and Hm25502 could not yet be identified on the genome sequence or
are not contained in the current assembly versions of the genomes.
16
Antigen B isoforms encoded by the genomes of Echinococcus multilocularis, Echinococcus granulosus and Hymenolepis microstoma.
LOW RESOLUTION FIG
Volume 33, Number **, **, 2011 Cestode genomics
? 2011 Blackwell Publishing Ltd, Parasite Immunology, 33, 1–21
11
1
2
3
4
5
6
7
8
9
10
11
12
13
14
15
16
17
18
19
20
21
22
23
24
25
26
27
28
29
30
31
32
33
34
35
36
37
38
39
40
41
42
43
44
45
46
47
48
49
50
51
52
1
2
3
4
5
6
7
8
9
10
11
12
13
14
15
16
17
18
19
20
21
22
23
24
25
26
27
28
29
30
31
32
33
34
35
36
37
38
39
40
41
42
43
44
45
46
47
48
49
50
51
52

Page 12

against CE in sheep (reviewed by Lightowlers; 106). The
EG95 gene has been demonstrated to belong to a gene
family that consists of six functional genes in E. granulosus
of which four encode a protein identical to the original
isolate (now named EG95-1; 107). The EG95 gene family
is structurally homologous to the 45W gene family and the
16K and 18K groups of antigens that are expressed in
various Taenia species (108). Like in the case of E. granulo-
sus, recombinant antigens of this family were already suc-
cessfully employed for the development of vaccines against
larval Taenia infections (106). The biological function of
the EG95⁄45W proteins is largely unknown. However, they
all share a common domain structure of a signal peptide,
followed by one single fibronectin III (Fn3) domain and a
hydrophobic transmembrane region close to the C-termi-
nus (107). Very interesting recent work on different Taenia
species (109,110) and E. granulosus (111) also demon-
strated that these proteins are primarily located in the
penetration glands of the nonactivated oncosphere and are
distributed over the oncospheral parenchyma upon activa-
tion with low-pH⁄pepsin treatment (mimicking the transi-
tion to the intermediate host). Because Fn3 domains are
typically found in extracellular matrix-associated proteins,
it is conceivable that the EG95⁄45W proteins play a role in
providing or organizing a primary matrix framework to
which totipotent parasite stem cells (delivered by the
oncosphere) can attach to undergo the early oncosphere–
metacestode transition, although experimental evidence
supporting this theory is still lacking. A close ortholog to
EG95 has also already been identified in E. multilocularis
(named EM95), and the respective recombinant protein
was effective in protecting mice against challenge infection
with E. multilocularis oncospheres (112). Because this was,
so far, the only report on these genes in E. multilocularis
and because the overall genomic organization of the
EG95⁄45W encoding genes had not been determined in the
other cestode species, we carried out respective analyses on
the assembled E. multilocularis genome. When the EM95,
EG95 and 45W sequences were used in tBLAST analyses,
we could indeed identify a relatively large number (up to
15) of related genes dispersed over the genome, most of
which were, however, transcriptionally silent according to
RNA-seq data and many contained inactivating mutations
in their reading frames. Only five of the genes showed
significant levels of transcription and only two of those,
located on scaffold_159 (Em95; position 5963–4694) and
scaffold_125 (Em95-2; 15880–14568) were closely related to
the previously identified EM95 (112) and displayed the
same conserved exon–intron structure (Figure 4). Intrigu-
ingly,in theRNA-seq transcription
oncosphere-specific genes displayed considerable levels of
expression in regenerating primary cells but not in metaces-
tode or protoscolex (Figure 5) which underscores the
suitability of the E. multilocularis stem cell cultivation sys-
tem to mimic the oncosphere–metacestode transition not
only morphologically (36), but also concerning gene
expression profiling. Two additional EM95-like genes that
we identified, located on scaffold_104 (Emy162a; position
44001–45896) and scaffold_7 (Emy162b; 35094–33349)
showed considerable homologies to the recently identified
EMY162 antigen (113). Unlike EM95, this antigen lacks
the C-terminal hydrophobic transmembrane domain (but
contains the signal peptide and the Fn3 domain) and is not
strictly expressed in the oncosphere but also in other devel-
opmental stages (113), which is clearly supported by our
RNA-seq transcriptome data (Figure 5). The fifth gene,
located on scaffold_45 (Emoal for oncosphere-antigen-like;
position 4212–3089) represents a novel, distantly related
member of the EG95⁄45W family that has not yet been
described in studies on vaccine development (Figure 4).
profiles,these
Figure 4
ment of the amino acid sequences of two Em95 isoforms (Em95, Em95-2; 107), two isoforms of EmY162 (EmY162a⁄b; 113) and an addi-
tional isoform, newly identified in this study (Emoal). Highly conserved residues are printed in white on black background, residues with
similar function in black on grey background. The location of signal peptides (SP), Fn3 domains (FN3) and transmembrane domains
(TMD) is indicated above the alignment. Note that no TMDs are predicted for the two EmY162 isoforms. Sequence- and primary cell-spe-
cific expression of the newly identified Emoal cDNA was verified by PCR amplification, cloning and sequencing. The Emoal cDNA
sequence is available in the EMBL, GenBank and DDJB databases under the accession no. FR848832.
17
Members of the EG95⁄45W protein family encoded by the Echinococcus multilocularis genome. Displayed is a CLUSTALW align-
LOW RESOLUTION FIG
P. D. Olson et al.
Parasite Immunology
12
? 2011 Blackwell Publishing Ltd, Parasite Immunology, 33, 1–21
1
2
3
4
5
6
7
8
9
10
11
12
13
14
15
16
17
18
19
20
21
22
23
24
25
26
27
28
29
30
31
32
33
34
35
36
37
38
39
40
41
42
43
44
45
46
47
48
49
50
51
52
1
2
3
4
5
6
7
8
9
10
11
12
13
14
15
16
17
18
19
20
21
22
23
24
25
26
27
28
29
30
31
32
33
34
35
36
37
38
39
40
41
42
43
44
45
46
47
48
49
50
51
52

Page 13

Very much like EM95, Emoal is specifically expressed in
regenerating primary cells; it displays an exon–intron struc-
ture that is typical for the EG95 gene family, and its gene
product comprises a signal peptide, one Fn3 domain and a
C-terminal transmembrane domain, suggesting that it has
a similar function as the EG95⁄45W proteins described so
far. A close ortholog to Emoal, Egoal, is also present on
the genome of E. granulosus (contig_32513; position 4699–
3576), which could prove important for the further
development and improvement of vaccine formulations
against CE. Interestingly, and in contrast to the AgB fam-
ily, the genome of H. microstoma is absolutely free of
EG95⁄45W-like sequences, which supports the idea that
this gene family is indeed highly specific to taeniid tape-
worms.
Additional antigens and immunomodulators
In addition to the TSOL18 and TSOL45 antigens of
T. solium, extensive vaccination trials against porcine cys-
ticercosis have already been undertaken using the so-
called S3Pvac vaccine (114,115). S3Pvac consists of three
synthetic peptides (named KETc12, KETc1, GK1) that
hadbeen identified by
T. crassiceps cDNA libraries and when tested under field
conditions, SP3vac could reduce the number of T. solium
infected pigs by 50% and lowered parasite load by >90%
(90). Interestingly, in spite of the fact that a considerable
amount of information has already been published on
S3Pvac (90), including a recent report on the presence of
similar sequences in other cestodes (116), the proteins
and genes which correspond to the synthetic peptides
have never been characterized so far. We therefore analy-
sed the situation for E. multilocularis using the published
KETc1 and GK1 sequences as well as E. multilocularis
genome and transcriptome data. The GK1 peptide clearly
maps to the amino acid sequence encoded by a predicted
gene on scaffold_13 (position 1.570.711–1.568.292). The
encoded protein (264 amino acids; 29 kDa; Figure 6) con-
tains one Glucosyltransferase⁄Rab-like GTPase activa-
tors⁄Myotubularin domain (GRAM domain), which is
thought to be an intracellular protein-binding or lipid-
binding signalling domain, and one WWbp domain which
is characterized by several short PY- and PT-motifs and
which presumably mediates tyrosine phosphorylation in
WW domain–ligandinteractions
within the WWbp domain, this protein displays signifi-
cant homologies (47% identical, 68% similar residues) to
a predicted S. mansoni protein, WW domain-binding pro-
tein 2 (accession no. FN313948), of unknown function.
The KETc1 peptide also clearly maps to a genomic
region that encodes a 67 kDa protein with significant
homologies (46%, 62%) to a hypothetical protein of
S. mansoni (accession no. FN357512). Interestingly, how-
ever, the KETc1 encoding region is out of frame of the
actual protein-encoding sequence and should, actually,
not be present in E. multilocularis (and most probably all
immune-screeningsagainst
(Figure 6). Atleast
Figure 5
locularis larval stages. Displayed is the relative expression level of
the two identified Em95 isoforms (Em95, Em95-2), the two
EmY162 isoforms (EmY162a⁄b), and the newly identified Emoal
in primary cells (PC; black; representing the oncosphere–metaces-
tode transition), in the metacestode stage (MC; dark grey), as well
as in protoscoleces prior to (P); light grey) and after (P+; white)
activation by low-pH⁄pepsin treatment. The expression level of
Em95 in primary cells has been set to 100%. Note that the Em95
and Emoal isoforms are expressed in a PC-specific manner,
whereas the EmY162 isoforms display stronger expression in the
metacestode. Data were obtained by RNA-seq analyses on in vitro
cultivated parasite larval stages.
18
Expression of EG95⁄45W genes in Echinococcus multi-
Figure 6
E. multilocularis WBP-2 protein. Identified GRAM- and WWbp-domains are indicated below the sequence by full and dashed lines, respec-
tively. The GK1 peptide which has been used in Taenia vaccination trials (90) is printed in white on black background.
19
The Echinococcus multilocularis WBP-2 protein. Displayed is the deduced amino acid sequence of the GK1-peptide-containing
LOW RESOLUTION FIG
LOW RESOLUTION FIG
Volume 33, Number **, **, 2011Cestode genomics
? 2011 Blackwell Publishing Ltd, Parasite Immunology, 33, 1–21
13
1
2
3
4
5
6
7
8
9
10
11
12
13
14
15
16
17
18
19
20
21
22
23
24
25
26
27
28
29
30
31
32
33
34
35
36
37
38
39
40
41
42
43
44
45
46
47
48
49
50
51
52
1
2
3
4
5
6
7
8
9
10
11
12
13
14
15
16
17
18
19
20
21
22
23
24
25
26
27
28
29
30
31
32
33
34
35
36
37
38
39
40
41
42
43
44
45
46
47
48
49
50
51
52

Page 14

other cestodes). As briefly discussed by Rassy et al. (116),
the initial identification of KETc1 might have resulted
from a reading frame error of the employed kZAP vector
which, nevertheless, does not explain why this peptide
induces high levels of protection when used as an immu-
nogen against cysticercosis (90).
Apart from the characterization of parasite-specific
antigen families, the available genome information should
also facilitate the identification of parasite orthologs with
homologies to immunomodulatory host proteins or cestode
orthologs of trematode proteins with such activities. As
already outlined, for cell–cell communication, cestodes uti-
lize evolutionarily conserved signalling systems of the insu-
lin-, the epidermal growth factor-, and the transforming
growth factor-b (TGF-b)-pathways and respective parasite
receptors that are able to functionally interact with corre-
sponding host hormones and cytokines have already been
identified (72). This makes it likely that cestodes also
express cognate ligands of these signalling systems which,
provided that they are secreted, could activate the corre-
sponding host receptors to affect host physiology or the
immune response. In fact, in preliminary analyses, we could
already identify several genes on the genome of E. multiloc-
ularis that encode insulin-like peptides and cytokines with
significant homologies to members of the TGF-b⁄BMP
families (72). Particularly, regarding the prominent role of
TGF-b in inducing anti-inflammatory immune responses
(117), the parasite cytokines of the TGF-b⁄BMP family are
of considerable interest and are currently under study in our
laboratories concerning influences on immune effector cells
such as dendritic cells and Tcells.
Prominent examples of immunomodulatory factors from
schistosome eggs are the ‘interleukin 4 (IL-4)-inducing
principle’ IPSE, which stimulates basophils to express and
secrete the Th2-associated cytokines IL-4 and IL-13 (118),
as well as the Omega-1 component of schistosome egg
antigen, which drives Th2 immune responses in mice
(119).Although
E. multilocularis
component with similar activities as IPSE (120), we could
so far not identify any cestode gene that encodes an IPSE-
like peptide, indicating that the IL-4 inducing activity is
because of another component in these organisms. An
ortholog to Omega-1, on the other hand, is clearly
encodedby the
E. multilocularis
genomes and could, like its schistosome counterpart, be
involved in driving Th2 responses during AE and CE,
respectively. Another family of proteins that are of tremen-
dous interest concerning immunomodulation by helminths
are the so-called cystatins, which are cysteine protease
inhibitors that are secreted by nematodes and interfere
with host cell antigen processing and presentation (121).
Very recently, one of these molecules has been demon-
extractcontainsa
and
E. granulosus
strated to exploit activation and deactivation pathways of
MAPKs to induce regulatory macrophages in filarial
infections (122). Interestingly, the E. multilocularis genome
encodes at least one cystatin with homologies to those of
nematode parasites, and transcriptome data show that this
factor is specifically (and highly) expressed in the metaces-
tode stage that is representative for the chronic phase of
AE (data not shown). Because macrophages from E. mul-
tilocularis infected mice are impaired in their ability to
present antigen to lymph node T cells (123), respective
activities of the E. multilocularis cystatin would be of par-
ticular interest and are currently addressed in our (KB)
laboratory. Hence, not only for investigations on cestode
evolution and development, or for the design of effective
chemotherapeutics, but also for novel approaches into the
immunology of cestode infections, the currently ongoing
genome projects hold great potential.
DEVELOPMENTAL SIGNALLING SYSTEMS IN
FLATWORMS
Our laboratory (PDO) began developing the H. micros-
toma model to investigate the roles of developmental
regulatory genes in cestodes, with the aim of understand-
ing the complex life histories of parasitic flatworms from a
comparative evolutionary context. It has become clear that
metazoans share a surprisingly small number of signalling
systems used to pattern their bodies (e.g. Notch, Hedge-
hog, Wnt, TGF-b and Receptor Tyrosine Kinase) and the
presence of most of these systems in the earliest branching
metazoans suggests that complexity in contemporary
animal form has not arisen through invention of new
systems, but through modification of ancient, highly
conserved genetic programmes (124). Current knowledge
of the signalling systems that underpin flatworm morpho-
genesis is based primarily on the study of planarians, for
which availability of a draft genome of S. mediterranea
has greatly accelerated research on planarian regeneration
and stem cells and has helped to re-establish them as a
powerful model in developmental biology (29,125,126). In
particular, investigations of highly conserved signalling
systems such as the Wnt⁄b-catenin pathway have yielded
several important discoveries in recent years regarding the
cellular decision making used to pattern their bodies
during growth and regeneration (127). By contrast, the
developmental biology of parasitic flatworms, and of
parasitic organisms generally, has been largely ignored in
preference to research relating to disease processes (128).
Consequently, little is known about the genetic basis of
their morphogenesis or the extent to which they share the
same compliment of developmental systems and genes
found in free-living animals (124). Thanks to new genome
P. D. Olson et al.
Parasite Immunology
14
? 2011 Blackwell Publishing Ltd, Parasite Immunology, 33, 1–21
1
2
3
4
5
6
7
8
9
10
11
12
13
14
15
16
17
18
19
20
21
22
23
24
25
26
27
28
29
30
31
32
33
34
35
36
37
38
39
40
41
42
43
44
45
46
47
48
49
50
51
52
1
2
3
4
5
6
7
8
9
10
11
12
13
14
15
16
17
18
19
20
21
22
23
24
25
26
27
28
29
30
31
32
33
34
35
36
37
38
39
40
41
42
43
44
45
46
47
48
49
50
51
52

Page 15

data; however, we are now in a position to begin catalogu-
ing developmentally relevant genes and investigating their
roles in their complex life histories. Although we do not
focus here on immunology or a medically important
model species, elucidating signalling systems that regulate
basic developmental processes in parasitic flatworms has
obvious relevance to the design and evaluation of chemo-
therapeutic targets.
The segmented, or strobilate, condition that is the
hallmark of tapeworms is a derived trait that evolved as
an adaptation to reproduction, as opposed to locomotion,
and has been considered an evolutionary novelty by most
developmental biologists, suggesting it lacks homology
with known mechanisms in, e.g., annelid worms, flies or
mice (129,130). Using Hymenolepis as a classical model
for studying adult development in tapeworms, we have ini-
tiated investigations on the mechanisms of axial patterning
through investigation of Hox and Wnt regulatory genes
(128,131,132). Hox genes encode transcription factors that
establish anteroposterior (AP) polarity, regional differenti-
ation and axial elaboration by regulating gene expression
in spatially and temporally specific patterns, whereas Wnt
genes encode ligands involved in cell–cell communication
and have been hypothesized as the ancestral metazoan
patterning system (133) that evolved to work in concert
with Hox genes during embryogenesis (134). Together,
these gene families and their interacting partners are the
most important known regulators of axial patterning in
metazoans (134). Elucidating their roles in tapeworms will
provide a common means by which the mechanisms of
segmentation and larval metamorphosis can be compared
with other parasitic and free-living flatworms, and to more
distantly related animal groups.
Hox genes in flatworms
The Hox genes and their evolutionary cousins the ParaHox
genes (135,136) are notable not only for their universality
in regulating axial patterning in animals, but for their
‘colinear’ architecture, by which the order in which they
are arrayed in the genome corresponds to their spatial
domains of expression, anterior to posterior (137). Three
paralogy groups (anterior, central and posterior) are recog-
nized corresponding to these domains, and a total of 11
genes has been hypothesized to be the ancestral state in
lophotrochozoans, including duplication of their ancestral
posterior Hox ortholog, giving rise to the lophotrochozo-
an-specific Post-1 and Post-2 genes (138). Although the
presence of Hox genes in flatworms has been known since
some of the first searches for Hox orthologs outside flies
and mice (139), the first investigation to focus specifically
on Hox genes in a parasitic flatworm was in 2005 by Pierce
et al. (140) who examined S. mansoni. Their work indi-
cated that flatworms had both a reduced and a dispersed
complement of Hox genes, and subsequent empirical and
in silico investigations of the tapeworms H. microstoma,
Mesocestoides corti and E. multilocularis, the polyopistho-
cotylean ‘monogenean’ Polystoma spp. and additional
work on S. mansoni have now confirmed this to be true in
each of the major parasitic groups (128,141,142).
Table 4 shows the presence of genes encoding Hox
orthologs in the genomes of Hymenolepis and Echinococ-
cus spp., S. mansoni, polyopisthocotylean ‘monogeneans’,
and the planarian S. mediterranea. From these representa-
tives, it appears that flatworms have a core set of one
anterior gene (Hox1⁄Lab) and three central genes (Hox3,
Hox4⁄Dfd, Lox4⁄Abd-A). In addition, both characteristic
lophotrochozoan posterior
found, although those were initially thought to be missing
from flatworms (128,143). Planarians also show the pres-
ence of Hox5 orthologs and larger numbers of central and
posterior paralogs than found in parasitic flatworms,
although it must be noted that whereas some of the
homeobox sequences (e.g. Hox1, Hox4⁄Dfd and Hox8⁄-
Abda) show high levels of similarity to cognates outside
the group, other flatworm homeoboxes are divergent and
difficult to classify. Nevertheless, compared with other
major lophotrochozoan groups such as annelids and mol-
luscs, both free-living and parasitic flatworms show reduc-
tions in the numbers of Hox gene classes, and this may
relate to their lack of axial elaboration. Hymenolepis is
also oddly missing an ortholog of the central Hox3 gene
found in all other flatworms examined.
Hox genes (Post-1⁄2) are
Table 4 Hox transcription factors in the genomes of parasitic and
free-living flatworms
Gene
CestodaTrematoda ‘Monogenea’ Planarian
H.m. E.m. E.g.S.m.Pol. Scm.m.
Hox1⁄Lab
Hox3⁄Zen
Hox4⁄Dfd
Hox5
(Lox4)⁄AbdA
Hox9-14
(Post-1⁄2)
Total
11
1
1
1
1
1
1
1
1
1
1
1
1
1
1
1
3
5
1
1
2
1
2
1
2
1
2
1
?
56665+ 12
Hox nomenclature based on vertebrates (lophotrochozoans)⁄Dro-
sophila. Data based on searches of publically available genomes
and previous reports including empirical and in silico analyses
(128,140–142).
H.m., Hymenolepis microstoma; E.m., Echinococcus multilocularis;
E.g., Echinococcus granulosus; Pol., species of polyopisthocotylean
‘monogeneans’ (141); S.m., Schistosoma mansoni; Scm.m. Schmid-
tea mediterranea.
Volume 33, Number **, **, 2011Cestode genomics
? 2011 Blackwell Publishing Ltd, Parasite Immunology, 33, 1–21
15
1
2
3
4
5
6
7
8
9
10
11
12
13
14
15
16
17
18
19
20
21
22
23
24
25
26
27
28
29
30
31
32
33
34
35
36
37
38
39
40
41
42
43
44
45
46
47
48
49
50
51
52
1
2
3
4
5
6
7
8
9
10
11
12
13
14
15
16
17
18
19
20
21
22
23
24
25
26
27
28
29
30
31
32
33
34
35
36
37
38
39
40
41
42
43
44
45
46
47
48
49
50
51
52