Results

ISJ 10: 29-37, 2013 ISSN 1824-307X

RESEARCH REPORT

A novel third complement component C3 gene of Ciona intestinalis expressed in the
endoderm at the early developmental stages

T Hibino1, M Nonaka2

1Faculty of Education, Saitama University, 255 Shimo-Okubo, Sakura-ku, Saitama City 338-8570, Japan
2Department of Biological Sciences, Graduate School of Science, The University of Tokyo, 7-3-1 Hongo,
Bunkyo-ku, Tokyo 113-0033, Japan

Accepted March 29, 2013

Abstract
The third complement component (C3) in ascidian was reported to function as an opsonin to

enhance phagocytosis and as a chemotactic factor for phagocytes, indicating that ascidian C3 works in
mesodermal cavity as a humoral factor like vertebrate C3s. In the basal Eumetazoa, Cnidaria lacking
mesodermal tissues, C3 was reported to work in an endodermal cavity. Evolution of structure and
function of C3 is still to be clarified. Here we report the identification of the third C3 gene, CiC3-3, in the
genome of an ascidian, Ciona intestinalis. Phylogenetic analysis using the entire amino acid sequences
of Eumetazoan C3s indicated that CiC3-3 possess a closer relationship to vertebrate C3, C4 and C5
than other ascidian C3s. Although CiC3-3 retained the α-β processing site and 6 cysteine residues in
the C3a region, it lacked the intra-molecular thioester bond and the catalytic histidine residue. Instead,
CiC3-3 had a unique insertion of about 70 residues long Lys/Arg-rich sequence. CiC3-3 was expressed
highly in the embryonic stages, but little in the adult in contradistinction to CiC3-1 and CiC3-2. The
expression of CiC3-3 in early embryonic stages was restricted to endoderm similar to cnidarian C3s.
Thus, the ascidian complement system could represent a unique evolutionary stage sharing a primitive
endodermal function with Cnidaria, and newly developed humoral function with vertebrates.

Key Words: thioester-containing protein (TEP); complement C3; innate immunity; immunogenetics; tunicate;
chordate

Introduction

The vertebrate complement system comprises

more than 30 proteins present in serum or on cell
surface, and plays a pivotal role in innate immunity.
This system is triggered by three different activation
pathways, the classical, alternative and lectin
pathways. These three pathways merge at the
proteolytic activation step of the complement
component 3 (C3) into C3a and C3b. Upon
proteolytic activation, C3 changes its conformation
exposing the intra-chain thioester bond at the
molecular surface. The exposed thioester bond of
C3b reacts with surface molecules of invading
microbes and makes a covalent bond, resulting in
covalent tagging of microbes with C3b. Covalently
attached C3b works as opsonin to induce
phagocytosis, and also induces assembly of the
___________________________________________________________________________

Corresponding author:
Masaru Nonaka
Department of Biological Sciences
Graduate School of Science
The University of Tokyo
7-3-1 Hongo, Bunkyo-ku, Tokyo 113-0033, Japan
E-mail: mnonaka@biol.s.u-tokyo.ac.jp

terminal components of complement (TCCs: C6~C9)
into a membrane-attack complex that can damage
the membrane of certain pathogens (Murphy, 2011).
The proteins possessing a similar domain structure
as vertebrate TCCs are present in ascidian and
amphioxus. However, these proteins may not be
activated through the complement system of
ascidian and amphioxus because they lack an
essential domain for interaction with C5b (Nonaka
and Kimura 2006).

The released smaller C3a fragment is an
anaphylatoxin to induce inflammation. The C3
subfamily including C3, C4 and C5 is a member of
thioester bond-containing protein (TEP) family,
together with the non-complement TEP subfamilies
such as the α2-macroglobulin (A2M) and CD109
subfamilies. The C3 subfamily members are
distinguished from the A2M and CD109 subfamily
members by the presence of the anaphylatoxin
(ANA) and C-terminal of C3, C4 ,C5 (C345C)
domains unique to the C3 subfamily (Sekiguchi et al.,
2012). Genes orthologous to vertebrate C3 have
been identified not only from invertebrate
deuterostome such as sea urchin (Al-sharif et al.,

Fig. 1 Phylogenetic tree of TEP family members constructed by the Neighbor-Joining method using the entire
amino acid sequences. Bootstrap values higher than 50 % are indicated in the tree. Accession numbers of each
entry are; human C3, C4A, C5, A2M, and CD109 (NP_000055, P0C0L4, AAA51925, P01023,and NP_598000),
carp C3H1, C4-1, and C5-1 (BAA36619, BAB03284, and BAC23057), ascidian, Halocynthia roretzi C3
(BAA75069), C. intestinalis CiC3-1, CiC3-2, CiA2M-like, and CiCD109-like (NP_001027684, CAC85958,
XP_002124325, and NP_001027688), sea urchin C3 and C3-2 (NP_999686 and Spbase: SPU_000997),
horseshoecrab C3 (AAQ08323), amphioxus C3 (BAB47146), coral C3 (AAN86548), cnidarian NvC3-1 and -2
(AB450038 and AB450040), and fly TEP1 (NP_523578).

1998; Hibino et al., 2006; Rast et al., 2006) and
amphioxus (Huang et al., 2008), but also from
protostomes such as horseshoe crab (Zhu et al.,
2005; Kawabata et al., 2009) and spider (Sekiguchi
et al. 2012) as well as cnidarian coral (Miller et al.,
2007) and sea anemone (Kimura et al., 2009). The
presence of C3 in Cnidaria indicated that the C3
gene has been established prior to the divergence of
cnidarian from bilaterian (Nonaka, 2011). In contrast
to its wide distribution, C4 and C5 has only been
identified in jawed vertebrate, suggesting that C4
and C5 were derived from a C3-like common
ancestor by gene duplication in the early stage of
jawed vertebrate evolution (Nonaka and Takahashi,
1992).

A tunicate, Ciona intestinalis (Urochordata) has
been an attractive research model for developmental
biology for more than a century (Satoh et al., 2003).
The recent accumulation of genome-wide sequence
information showed that not cephalochordate but
urochordate is a sister group of vertebrate, indicating
that C. intestinalis is one of the most important
species for understanding the origin and evolution of
vertebrates (Dehal et al., 2002, Putnam et al., 2008).

The C. intestinalis genome analysis revealed that
this animal possesses several genes for
complement components: two C3s, three Bf/C2s, 10
of C6/C7/C8/C9/perforin and so on. An ancestor of
the two C3-like genes seems to have diverged from
a common ancestor of vertebrate C3/C4/C5 and has
duplicated into two genes in the Ciona lineage
(Azumi et al., 2003a). Using C. intestinalis, in-depth
expressed sequence tag and large-scale oligo-DNA
microarray analyses have been advanced, which
identified gene expression profile during the life
cycle (Azumi et al., 2003b; Satou et al., 2002, 2003).
Interestingly, C3s, MASP, factor B (Bf), MBP and
two genes of complement C6-like were expressed
only in the adult stages. On the other hand, C1q-like
and two other genes of complement C6-like were
expressed in the middle of the embryonic stages and
maintained their expression level during the adult
stages (Azumi et al., 2007). The absence of C3
expression during the developmental stages could
be explained by one of the following two hypotheses:
(1) since C. intestinalis develops directly and
metamorphosis in a day after fertilization, protection
against infection which is considered to be the most

important physiological function of C3 is
unnecessary in this short period, or (2) unidentified
C3 is working under the developmental stage.

In this study, we report a novel C. intestinalis C3
gene, CiC3-3, that belonged to a different clade from
known ascidian C3 genes in a phylogenetic tree, and
that was specifically expressed in endoderm of the
embryos.

Materials and Methods

Adults and embryos

Adults of the ascidian C. intestinalis were
provided from Misaki Marine Biological Station, the
University of Tokyo through National Bio-Resource
Project (NBRP) of MEXT, Japan. The adults were
surgically dissected to draw eggs and sperm.
Fertilized eggs were incubated at devitellination
medium containing 0.065 % actinase E (Kaken Co.
Ltd.) and 1.3 % sodium thioglycolate in sea water at
a pH of approximately 10, to devitellinate chemically
(Satou et al., 2001). Devitellinated embryos were
reared at 18 °C in agar-coated plastic dished filled
with filtered sea water containing 50 µg/ml penicillin
and 100 µg/ml streptomycin.

Gene identification in C. intestinalis genome
database

Deduced amino acid sequences of all
computationally predicted proteins were downloaded
from the website, Ensembl C. intestinalis database
(http://www.ensembl.org/) (Hubbard et al., 2007),
and the Ghost Database: C. intestinalis genomic and
cDNA resources
(http://ghost.zool.kyoto-u.ac.jp/indexr1.html) (Satou
et al., 2005). A typical C3 protein contains multiple
domains; A2M_N, A2M_N2, A2M, ANATO,
A2M_comp, A2M_recep, C345C, whose profile
HMMs were downloaded from the Pfam website
(http://pfam.sanger.ac.uk/) (Bateman et al., 2004).
HMMER (Eddy et al., 1998) was used to identify
PFAM domain profile matches to C. intestinalis
protein models. The deduced amino acid sequences
of identified protein models were aligned with those
of known C. intestinalis C3, CiC3-1 and CiC3-2 and
Ciα2-macroblobulins to find out a novel C3 gene
model.

Cloning of a novel C3 gene of C. intestinalis

C. intestinalis cDNA was synthesized from the
adult tissues containing gills and blood cells, and
used as a template for PCR amplification. To confirm
the nucleotide sequences, especially exon-intron
boundary, of the novel identified C3 (CiC3-3) gene
model, RT-PCR was performed using primers that
were designed at the ends of 5’ UTR (forward
5’-TTGGAAAGCCGTACTATGCGACACG-3’) and 3’
UTR (reverse
5’-TGCTTTGGCAATATACACGTGGCAGT-3’) of the
gene model. Probably the nucleotide length of the
predicted gene model of 5.7 kbp was too long for
RT-PCR, the entire length of CiC3-3 could not be
amplified by using these primers. We then designed
other primers at the middle of the gene models
(forward 5’-TGGAACAATCGCTGCTGCTGTAA-3’,
reverse 5’-ATGCCTTCTGGGACCACATTCAA-3’).
ExTaq DNA polymerase (Takara) was used for this

PCR. The cDNA fragments were cloned into
pCR2.1-TOPO vector (Invitrogen) and were
sequenced using vector specific primers or gene
specific primers.

Domain Prediction and phylogenetic analysis

Domain structure of the CiC3-3 was predicted
by the SMART program
(http://smart.embl-heidelberg.de/) (Letunic et al.,
2002). The e-value for the domain confidence was
assessed by HMMER3 on the SMART program.
Multiple alignment of the amino acid sequences
among C. intestinalis and human C3s was done by
ClustalW on the MEGA5 program (Tamura et al.,
2011), as well as by eyes. Based on this alignment,
phylogenetic trees were constructed using full-length
amino acid sequence information or A2M_comp
domain region that was extracted using by the
SMART program. The neighbor-joining (NJ) method
(Saitou and Nei 1987) using MEGA5 excluding gaps
by pairwise deletion was performed. The reliability
for internal branches was assessed by the 1000
bootstrap replications.

Gene expression analysis using Ciona database and
whole mount in situ hybridization

The C. intestinalis protein database (CIPRO
2.5) integrates not only protein database, but also
transcriptome database including large-scale EST
analysis and DNA microarray data (Endo et al.,
2011). We extracted the expression data of the three
CiC3 genes from the website and integrated the data
in a graph. Whole mount in situ hybridization was
performed based on the previously described
protocol (Ogasawara et al., 2001; Satou et al., 2001)
with some modifications. For antisense or sense
ribonucleotide probe for CiC3-3, 544 bp of the 3’ end
of coding region that covers the full length of the
C345C domain and subsequent stop codon for
CiC3-3 was cloned into pTAC-2 with DynaExpress
TA PCR Cloning Kit, and the probes were
subsequently synthesized using Digoxigenin (DIG)
RNA labeling mix and T7 or SP6 RNA polymerase
(Roche). Embryos were fixed with 4 %
paraformaldehyde in 0.1 M MOPS (pH 7.5), 0.5 M
NaCl at 4 °C overnight. The developmental stages of
fixed embryos were determined following Hotta et al.
(2007). The fixed embryos were washed three times
with PBST (phosphate-buffered saline containing
0.1 % Tween-20), then partially digested with 2
μg/ml proteinase K in PBST for 20 min at 37 °C.
They were washed twice with PBST, subsequently
post-fixed with 4 % paraformaldehyde in PBST for 1
h at room temperature (RT), and then washed three
times with PBST. After prehybridization at 50 °C for
1 h, the embryos were hybridized at 50 °C for 24 h in
the following buffer. The hybridization buffer
contained 50 % formamide, 5x Denhardt’s solution,
100 μg/ml yeast RNA, 0.1 % Tween-20, and 0.2
μg/ml DIG-labeled RNA probes. After hybridization,
the embryos were washed three times with 2xSSC,
50 % formamide, 0.1 % Tween-20 at 50 °C for 15
min, then washed three times with 1xSSC, 50 %
formamide, 0.1 % Tween-20 (A) at 50 °C for 15 min.
Next they were washed twice with 1:1, (A): PBST at
RT for 10min, and then washed three times with
PBST at RT for 3 min. After the series of washing,

http://www.ensembl.org/
http://ghost.zool.kyoto-u.ac.jp/indexr1.html
http://pfam.sanger.ac.uk/
http://smart.embl-heidelberg.de/

the specimens were blocked with 0.5 % blocking
reagent (Roche) in PBST at RT for 30 min. They
were immersed in 1/2000 Anti-DIG-AP fab fragments
(Roche) diluted with PBST at RT for 6 h. The
embryos were washed 4 times with PBST for 10 min
and then washed twice with alkaline phosphatase
buffer (0.1 M Tris-HCl (pH 9.5), 50 mM MgCl2, 0.1 M
NaCl) for 10 min. For signal detection, the embryos
were incubated with NBT/BCIP in the alkaline
phosphatase buffer at RT overnight. The stained
embryos were dehydrated in a graded series of
ethanol, and then cleared in a 1: 2 mixture of benzyl
alcohol/benzyl benzoate.

Results

Identification of the third complement C3 gene in C.
intestinalis

To find gene candidates encoding multiple
domains of typical thioester containing protein (TEP)
superfamily from C. intestinalis, all of the deduced
amino acid sequences of the Fgenesh gene models
and the GENSCAN gene models that
computationally predicted from the C. intestinalis
genome were searched by local HMMER program
using profile HMMs containing the A2M_N, A2M_N2,
ANATO, A2M, A2M_comp and C345C domains. Out
of five gene models extracted by this analysis, four
gene models matched with the already reported TEP
genes. The other gene model, Fgenesh76597 or
GENSCAN101558, predicted a 3,759 bp open
reading frame corresponding to a 1,253 amino acid
sequence containing the A2M_N2, ANATO, A2M
and C345C domains. The same gene was contained
in the recently uploaded KH gene models (ver. 2012)
in Ghost Database: C. intestinalis genomic and
cDNA resources
(http://ghost.zool.kyoto-u.ac.jp/indexr1.html) (Satou
et al., 2005). This gene model,
KH.C12.243.v1.A.SL1-1, with a longer nucleotide
sequence than that of the Fgenesh/GENSCAN
model predicted a 1,873 amino acid sequence
containing the additional A2M_N domain at its
N-terminus. Based on these gene models, several
primers were constructed at the end or at the middle
of the sequences, and then a novel C3 gene
candidate was cloned and sequenced. The cloned
nucleotide sequences had a size of 5,946 bp (1,873
amino acid residues) that matched 98.7 %
(5666/5739) to that of the KH gene model. The novel
and the third complement C3 gene in C. intestinalis
was designated as CiC3-3.

Phylogenetic analysis using the Neighbor-joining
method

The deduced amino acid sequence of the
CiC3-3 gene was aligned with the known
Eumetazoan TEP superfamily genes using ClustalW
program (data not shown). The phylogenetic tree
was constructed based on the entire amino acid
sequences using the NJ method (Fig. 1). The
phylogenetic tree showed the presence of three
subfamilies, the C3, A2M and CD109 subfamilies,
supported with bootstrap percentages of 52, 98 and
99 %, respectively. In the C3 subfamily, CiC3-3 was
grouped with vertebrate C3/C4/C5 with a 85 %
bootstrap percentage, but not with other ascidian C3
lineage including CiC3-1, CiC3-2 and H. roretzi C3.
This result together with a long branch length of
CiC3-3 indicates that CiC3-3 is a highly derivative
ascidian C3, whose evolutionary origin is still to be
clarified by analyzing other ascidian species.

Structural features of CiC3-3

To reveal whether CiC3-3 conserves the
primary structure as well as domain structures of the
vertebrate C3 subfamily, the deduced amino acid
sequences of CiC3-3 were aligned with CiC3-1 and
-2, and human C3, C4 and C5. The SMART domain
search was also performed to find the multiple
domains (Fig. 2).

The alignment and domain search showed that
CiC3-3 possesses a signal peptide for secretion, the
α/β processing site (RXXR) for dividing into two
subunit chains, two Cys residues involved in an
inter-chain disulfide linkage between the α and β
chains and a possible activation cleavage site (TTR)
by the C3 convertase (Fig. 1, Table 1), suggesting
that CiC3-3 is processed into α- and β-chains held
together with the inter-chain disulfide bond similar to
mammal C3s. The C3a anaphylatoxin (ANA) region
of CiC3-3 contained the six Cys residues conserved
by most C3a analyzed thus far. Since CiC3-1 and -2
possess only four of them, CiC3-3 showed a higher
conservation in the C3a region. However, CiC3-3
lacked the thioester site, GCGEQ, and the catalytic
His residue for cleavage of thioester. Moreover
CiC3-3 also lacked the two Pro residues on both
sides of the thioester site which are conserved even
in human C5 lacking the thioester site. These
results suggest that the 3D structure around the
thioester site is markedly modified in CiC3-3
(highlighted in yellow and blue in Fig. 2,
summarized in Table 1).

Table 1

http://ghost.zool.kyoto-u.ac.jp/indexr1.html

Fig. 2 Sequence comparison of CiC3-3, CiC3-1, CiC3-2 and the human C3, C4B, and C5. The multiple sequence
alignment of CiC3-3 with human and other C. intestinalis C3s was performed with ClustalW. A2M_comp domain
region is boxed. Proteolytic cleavage sites are shown in bold letters. Conserved Cys residues in the C3a
anaphylatoxin region are marked (*). The inter-chain disulfide bridges between the α/β chains are shaded. The
thioester sites, catalytic His sites and KR-rich insertion are also annotated in each colored box.

Phylogenetic analysis of A2M_comp domain region
to reveal independent loss of thioester site and
catalytic His.

To reveal whether loss of the thioester site and
catalytic His occurred independently in CiC3-3 and
vertebrate C5 or not, we reconstructed a
phylogenetic tree with the deduced amino acid
sequences based on the A2M_comp domain located
at the C-terminal side of the thioester site. Although
the size of the usual A2M_comp domain is
approximately 260 amino acid residues long, this
domain of CiC3-3 expands to 336 residues due to an
insertion of approximately 70 amino acid residues
highly enriched in Lys and Arg. A similar Lys/Arg rich
insertion has already been reported from two
cnidarian C3s, Nv3-1, Nv3-2, although the insertion
of cnidarian C3 was observed at the different region,
much more C-terminal side. Therefore, the insertion
of the Lys/Arg rich sequence into C3 occurred at
least twice independently during the eumetazoa
evolution. The NJ tree constructed using the amino
acid sequences of the A2M_comp domain showed
the essentially the same topology as the tree based
on the full length information described above,
except that CiC3-3 is separated far from C3 family
(Fig. 3). The long branch of CiC3-3 indicates that the
primary structure of A2M_comp region of CiC3-3 is
highly divergent. Overall, these results indicate that
CiC3 possesses well conserved domain
organization similar to vertebrate C3/C4/C5 except
for the thioester site and the subsequent A2M_comp
domain.

Spatial and temporal expression of the CiC3-3
gene

To understand the gene expression pattern of
CiC3-3, we first analyzed transcriptome data on the
C. intestinalis protein database (CIPRO) (Endo et al.,

Fig. 3 Phylogenetic tree of TEP family members
constructed by the Neighbor-Joining method using
the amino acid sequences of A2M_comp domain.
Bootstrap values higher than 50 % are indicated in
the tree. The genes in this tree are same as Figure 1.

2011), and compared gene expressions among
CiC3-1, CiC3-2 and CiC3-3. Both CiC3-1 and
CiC3-2 were not expressed before the
metamorphosis except for very slight expression in
the tailbud stage, while both microarray and EST

A) B)

Fig. 4 Comparison of expression intensities among CiC3-1, CiC3-2 and CiC3-3. CiC3-1, -2 and -3 are shown in
light blue, light green and red, respectively. The bars represent the EST data, while the lines represent the
microarray data (labeled as CiC3-1M, -2M -3M). Y axes of Graphs A and B indicate relative expression levels. The
results of the EST and microarray analyses are shown on the left and right sides, respectively. A: Expression
profiles during the life cycle of C. intestinalis. The left side of the graph indicates the expression intensity for EST
data, while the right side of the graph denotes the expression intensity for microarray data. B: Expression profiles
of adult tissues. The bars denote expression level estimated by the EST analysis.

data showed that CiC3-3 was significantly
expressed from the gastrula to the tailbud stage (Fig.
4A). The expression of CiC3-3 disappeared by the
larva stage. After metamorphosis, CiC3-1 and CiC3-2
began to be expressed, whose intensities were
getting stronger during maturation. In contrast to
CiC3-1 and CiC3-2, CiC3-3 was not expressed by
juvenile and was slightly expressed from the young
adult to mature adult stages. The intensity of CiC3-3
expression in mature adults is approximately 1/5 of
CiC3-1 and 1/7 of CiC3-2 (Fig. 4A). The weak
expression of CiC3-3 was detected only in the blood
cells. CiC3-1 was ubiquitously expressed except for
ovary and endostyle, and CiC3-2 was expressed in
heart and blood cell (Fig. 4B). These expression
data indicate that CiC3-3 is expressed in a
contradistinctive manner from CiC3-1 and CiC3-2.

To identify the spatial expression pattern of
CiC3-3 during the development of C. intestinalis, we
next performed whole mount in situ hybridization
using RNA probes of CiC3-3. CiC3-3 began to be
expressed in the invaginated cells of the early
gastrulae (St. 11) (Fig. 5A). At the late gastrula stage
(St. 13), almost all of the invaginated cells expressed
CiC3-3. The CiC3-3 expression was then restricted
in the anterior end of the embryos (St. 14), especially
the anterior ventral side of the invaginated cells
strongly expressed CiC3-3 (Figs 5D, E). The strong
expression was observed in the endoderm of the
trunk region, and weak expression was observed in
the endoderm strand of the ventral midline of the tail
region (St. 16 and 19) (Figs 5F, G, H). At the mid
tailbud stage (St. 21) the CiC3-3 expression was
reduced and restricted only in the endoderm cells

around the endodermal cavity (Figs 5I, J). These
expression data indicates that CiC3-3 is specifically
expressed in the endoderm of embryos, and ceases
its expression before hatching into the larvae.

Discussion

It had been reported that the number of

complement C3 gene is one in H. roretzi, and two in
C. intestinalis (Nonaka et al., 1999; Marino et al.,
2002; Azumi et al., 2003). H. roretzi and C.
intestinalis belong to the orders, Pleurogona and
Enterogona, respectively, and are evolutionary far
apart to each other (Turon et al., 2004). The
phylogenic analysis of C3 genes have indicated that
the gene duplication event between CiC3-1 and
CiC3-2 occurred in the Enterogona lineage after the
divergence from Pleurogona (Marino et al., 2002).
The newly found CiC3-3 clustered with vertebrate
C3, C4 and C5, rather than with CiC3-1, CiC3-2 and
H. roretzi C3, although bootstrap percentage to
support this clustering was not very high. This
finding indicates the presence of two ancient C3
lineages in basal tunicates, the CiC3-1, CiC3-2 and
H. roretzi C3 lineage and the CiC3-3 lineage. Two
and three C3 genes were reported from the
genomes of a sea urchin, Strongylocentrotus
purpuratus, and an amphioxus, Branchiostoma
floridae, respectively (Hibino et al., 2006; Huang et
al., 2008). Thus all the basal deuterotomes whose
genomes have been elucidated so far contain
more than two C3 genes. Phylogenetic analysis
showed that the multiple C3 genes of each species
form species-specific cluster, indicating that gene

Fig. 5 Spatial expression pattern of CiC3-3 detected by whole mount in situ hybridization in the early gastrula

through the mid tailbud stage embryos. The scale bar in A indicates 20 µm, and the magnification is the same for
all pictures. A-E: anterior is toward to the top, F-J: anterior is toward to the left. F, H, I, J: ventral is toward to the
bottom. A: vegetal view of early gastrula (4.9 hpf, St. 11), B-C: vegetal view of late gastrulae (5.9 hpf, St.13), C: no
hybridization signal with sense probes, D: vegetal view of early neurulae (6.35 hpf, St. 14), E: lateral view of D,
vegetal is toward to the right, F: lateral view of late neurulae (7.4 hpf, St. 16), G: ventral view of F, H: lateral view of
early tailbud (9.3 hpf, St. 19), I, J: lateral view of mid tailbud (10 hpf, St. 21), arrow indicates endodermal cavity, J:
Diagram of mid-tailbud corresponding to plate I. Yellow, light blue and pink denote endoderm, nervous system and
notochord, respectively.

duplications occurred multiple times in each lineage.
CiC3-3 is exceptional in this aspect, suggesting a
unique evolutionary history of this gene.

CiC3-1 and CiC3-2 retain almost all domains
and structural features of vertebrate C3, suggesting
that they function as the central component of the
ascidian complement system. Actually, the C3a
fragment of CiC3-1 was demonstrated to induce
chemotaxis of C. intestinalis hemocytes in the same
way as vertebrate C3a (Pinto et al., 2003). In
contrast, CiC3-3 showed an unprecedentedly unique
structure. First of all, CiC3-3 lacked the thioester site
believed to be essential for covalent tagging of
invading microorganisms by usual C3. Unlike
vertebrate C5 which also lacks the thioester site but
retains the basic residues of the thioester domain,
CiC3-3 has a totally different sequence in this
domain. Especially, the insertion of the highly
Lys/Arg-rich sequence could have drastic structural
and functional consequence since it provides
extremely positive charge to this region. It is unlikely,
therefore, that CiC3-3 play a similar function as
vertebrate C3. However, the C3a region of CiC3-3
showed a higher conservation of Cys residues than
those of CiC3-1 and CiC3-2, implicating in
inflammatory process as anaphylatoxin.

In mammal, the C3 gene is mainly expressed in
hepatocytes and macrophages (Lambris, 1988). In
the ascidian H. roretzi, gastric caecum and blood
cells have been identified as the sites of C3 gene
expression (Nonaka et al., 1999). The paraffin
sections of the stomach in the adult of C. intestinalis
have shown that both CiC3-1 and CiC3-2 are
expressed only in the one type of blood cell, but not
in the wall of the stomach (Marino et al., 2002).
Gene expression profile during the life cycle of C.
intestinalis using the large-scale oligo-DNA
microarray showed that not only CiC3-1 and CiC3-2,
but also MASP, factor B, MBP and two genes of
complement C6-like were expressed only in the
adult stages (Azumi et al., 2007). Two other genes of
complement C6-like were expressed in the middle of
the embryonic stages and maintained their
expression level during the adult stages. In this
study, CiC3-3 showed a totally different temporal
expression pattern during the life cycle from the
other complement component genes of C.
intestinalis. CiC3-3 is prominently expressed during
the embryonic stages when the other complement
genes of C. intestinalis are hardly expressed. In
adult stages, in contrast, CiC3-3 is expressed at a
very low level, whereas the other complement genes
are expressed abundantly. Since interactions among
components are essential for complement activation,
these results suggest that CiC3-3 functions outside
of the complement system.

Whole mount in situ hybridization revealed that
CiC3-3 was first expressed in the invaginating
endoderm of the embryos. C. intestinalis develops in
a direct developing manner, and the larvae do not
undergo the differentiation of a functional gut. Thus,
endodermal expression of CiC3-3 does not
necessarily indicate digestive function. When the
gastrulation began, the expression started in the
invaginated region, and it was continuously seen
invaginating cells from the head endoderm through
the endoderm strand to the ventral blastopore. After

closure of the blastopore, it was strongly expressed
around the endodermal cavity in the trunk. This
expression pattern indicates the possibility that
CiC3-3 is involved in development of certain
embryonic region.

Complement C3 genes have been reported
from basic metazoans, cnidarian coral, Swiftia
exserta (Dishaw et al., 2005), and cnidarian sea
anemone, Nematostella vectensis (Kimura et al.,
2009). Another coral, Acropora millepora C3 is
expressed in undifferentiated endodermal cells of
the embryos and larvae (Miller et al., 2007), while N.
vectensis C3 is expressed in tentacles, pharynx, and
mesentery in an endoderm-specific manner.
Although all these cnidarian C3 possess the typical
domain structure of vertebrate C3 unlike CiC3-3, a
similar expression pattern during embryonic stages
could imply that CiC3-3 and cnidarian C3 play some
common developmental roles. If this is the actual
case, cnidarian C3 has dual roles in development
and immunity, which are divided into CiC3-3 and
CiC3-1, 2, respectively in ascidians.

References
Al-Sharif WZ, Sunyer JO, Lambris JD, Smith LC.

Sea urchin coelomocytes specifically express a
homologue of the complement component C3. J.
Immunol.160: 2983-2997, 1998.

Azumi K, De Santis R, De Tomaso A, Rigoutsos I,
Yoshizaki F, Pinto MR, et al. Genomic analysis
of immunity in a Urochordate and the
emergence of the vertebrate immune system:
"waiting for Godot". Immunogenetics 55:
570-581, 2003a.

Azumi K, Takahashi H, Miki Y, Fujie M, Usami T,
Ishikawa H, et al. Construction of a cDNA
microarray derived from the ascidian Ciona
intestinalis. Zool. Sci. 20: 1223-1229, 2003b.

Azumi K, Sabau SV, Fujie M, Usami T, Koyanagi R,
Kawashima T, et al. Gene expression profile
during the life cycle of the urochordate Ciona
intestinalis. Dev. Biol. 308: 572-582, 2007.

Bateman A, Coin L, Durbin R, Finn RD, Hollich V,
Griffiths-Jones S, et al. The Pfam protein
families database. Nucleic Acids Res. 32:
D138-D141, 2004.

Dehal P, Satou Y, Campbell RK, Chapman J,
Degnan B, De Tomaso A,et al. The draft
genome of Ciona intestinalis: Insights into
chordate and vertebrate origins. Science 298:
2157-2167, 2002.

Dishaw LJ, Smith SL, Bigger CH. Characterization of
a C3-like cDNA in a coral: phylogenetic
implications. Immunogenetics 57: 535-548, 2005.

Eddy SR. Profile hidden Markov models.
Bioinformatics 14: 755-763, 1998.

Endo T, Ueno K, Yonezawa K, Mineta K, Hotta K,
Satou Y, et al. CIPRO 2.5: Ciona intestinalis
protein database, a unique integrated repository
of large-scale omics data, bioinformatic
analyses and curated annotation, with user
rating and reviewing functionality. Nucleic Acids
Res. 39: D807-D814, 2011.

Hibino T, Loza-Coll M, Messier C, Majeske AJ,
Cohen AH, Terwilliger DP, et al. The immune
gene repertoire encoded in the purple sea
urchin genome. Dev. Biol. 300: 349-365, 2006.

Hotta K, Mitsuhara K, Takahashi H, Inaba K, Oka K,
Gojobori T, et al. A web-based interactive
developmental table for the ascidian Ciona
intestinalis, including 3D real-image embryo
reconstructions: I. From fertilized egg to
hatching larva. Dev. Dyn. 236: 1790-1805, 2007.

Huang S, Yuan S, Guo L, Yu Y, Li J, Wu T, et al.
Genomic analysis of the immune gene
repertoire of amphioxus reveals extraordinary
innate complexity and diversity. Genome Res.
18: 1112-1126, 2008.

Hubbard TJ, Aken BL, Beal K, Ballester B, Caccamo
M, Chen Y, et al. Ensembl 2007. Nucleic Acids
Res. 35: D610-D617, 2007.

Kawabata S, Koshiba T, Shibata T. The
lipopolysaccharide-activated innate immune
response network of the horseshoe crab. Inv.
Surv. J. 6: 59-77, 2009.

Kimura A, Nonaka M. Genomic view of the evolution
of the complement system. Immunogenetics 58:
701-713, 2006.

Kimura A, Sakaguchi E, Nonaka M.
Multi-component complement system of
Cnidaria: C3, Bf, and MASP genes expressed
in the endodermal tissues of a sea anemone,
Nematostella vectensis. Immunobiology 214:
165-178, 2009.

Letunic I, Goodstadt L, Dickens NJ, Doerks T, Schultz
J, Mott R, et al. Recent improvements to the
SMART domain-based sequence annotation
resource. Nucleic Acids Res. 30: 242-244, 2002.

Marino R, Kimura Y, De Santis R, Lambris JD, Pinto
MR. Complement in urochordates: cloning and
characterization of two C3-like genes in the
ascidian Ciona intestinalis. Immunogenetics 53:
1055-1064, 2002.

Miller DJ, Hemmrich G, Ball EE, Hayward DC,
Khalturin K, Funayama N, et al. The innate
immune repertoire in Cnidaria - ancestral
complexity and stochastic gene loss.Genome
Biol. 8: 13, 2007.

Murphy K. Janeway's Immunology 8th Edition,
Garland Science, 48-71, 2011.

Nonaka M. The complement C3 protein family in
invertebrates. Inv. Surv. J. 8: 21-32, 2011.

Nonaka M, Azumi K, Ji X, Namikawa-Yamada C,
Sasaki M, Saiga H, et al. Opsonic complement
component C3 in the solitary ascidian,
Halocynthia roretzi. J. Immunol.162: 387-391,
1999.

Nonaka M, Takahashi M. Complete complementary
DNA sequence of the third component of
complement of lamprey. Implication for the
evolution of thioester containing proteins. J.
Immunol.148: 3290-3295, 1992.

Ogasawara M, Minokawa T, Sasakura Y, Nishida H,
Makabe KW. A large-scale whole-mount in situ
hybridization system: Rapid one-tube
preparation of DIG-labeled RNA probes and

high throughput hybridization using 96-well
silent screen plates. Zool. Sci. 18: 187-193,
2001.

Pinto MR, Chinnici CM, Kimura Y, Melillo D, Marino
R, Spruce LA, et al. CiC3-1a-mediated
chemotaxis in the deuterostome invertebrate
Ciona intestinalis (Urochordata). J.
Immunol.171: 5521-5528, 2003.

Putnam NH, Butts T, Ferrier DE, Furlong RF,
Hellsten U, Kawashima T, et al.The amphioxus
genome and the evolution of the chordate
karyotype. Nature 453: 1064-71. 2008.

Rast JP, Smith LC, Loza-Coll M, Hibino T, Litman
GW. Review - Genomic insights into the
immune system of the sea urchin. Science 314:
952-956, 2006.

Saitou N, Nei M. The neighbor-joining method: a
new method for reconstructing phylogenetic
trees. Mol. Biol. Evol. 4: 406-425, 1987.

Satoh N, Satou Y, Davidson B, Levine M. Ciona
intestinalis: an emerging model for
whole-genome analyses. Trends Genet. 19:
376-381, 2003.

Satou Y, Takatori N, Yamada L, Mochizuki Y,
Hamaguchi M, Ishikawa H,et al. Gene
expression profiles in Ciona intestinalis tailbud
embryos. Development 128: 2893-2904, 2001.

Satou Y, Takatori N, Fujiwara S, Nishikata T, Saiga
H, Kusakabe T, et al. Ciona intestinalis cDNA
projects: expressed sequence tag analyses and
gene expression profiles during embryogenesis.
Gene 287: 83-96, 2002.

Satou Y, Kawashima T, Kohara Y, Satoh N. Large
scale EST analyses in Ciona intestinalis - Its
application as Northern blot analyses. Dev.
Genes Evol. 213: 314-318, 2003.

Satou Y, Kawashima T, Shoguchi E, Nakayama A,
Satoh N. An integrated database of the ascidian,
Ciona intestinalis: Towards functional genomics.
Zool. Sci. 22: 837-843, 2005.

Sekiguchi R, Fujito NT, Nonaka M. Evolution of the
thioester-containing proteins (TEPs) of the
arthropoda, revealed by molecular cloning of
TEP genes from a spider, Hasarius adansoni.
Dev. Comp. Immunol. 36: 483-489, 2012.

Tamura K, Peterson D, Peterson N, Stecher G, Nei
M, Kumar S. MEGA5: Molecular Evolutionary
Genetics Analysis Using Maximum Likelihood,
Evolutionary Distance, and Maximum
Parsimony Methods. Mol. Biol. Evol.28:
2731-2739, 2011.

Turon X, Lopez-Legentil S. Ascidian molecular
phylogeny inferred from mtDNA data with
emphasis on the Aplousobranchiata. Mol. Phyl.
Evol. 33: 309-320, 2004.

Zhu Y, Thangamani S, Ho B, Ding JL. The ancient
origin of the complement system. EMBO J. 24:
382-394, 2005.