Fish & Shellfish Immunology 47 (2015) 302e312

Contents lists available at ScienceDirect

Fish & Shellfish Immunology

journal homepage: www.elsevier.com/locate/fsi

Full length article

Molecular diversity and evolution of defensins in the manila

clam Ruditapes philippinarum
Qing Wang a, Linbao Zhang b, Dinglong Yang a, c, Qian Yu a, c, Fei Li a, Ming Cong a,
Chenglong Ji a, Huifeng Wu a, Jianmin Zhao a, *
a
Key Laboratory of Coastal Environmental Processes and Ecological Remediation, CAS, Shandong Provincial Key Laboratory of Coastal Zone Environmental
Processes, Yantai Institute of Coastal Zone Research, Chinese Academy of Sciences, Yantai 264003, PR China
b
South China Sea Fisheries Research Institute, Chinese Academy of Fishery Sciences, Guangzhou 510300, PR China
c
University of Chinese Academy of Sciences, Beijing 100049, PR China

a r t i c l e i n f o a b s t r a c t

Article history: Four types of defensins were identified in Manila clam and designated as Rpdef1, Rpdef2, Rpdef3 and
Received 22 April 2015 Rpdef4, which encoded a polypeptide of 49, 46, 45 and 42 amino acids, respectively. Sequence align-
Received in revised form ments indicated that Rpdef1 shared 46.9% identity with Rpdef2, 40.8% with Rpdef3, and 34.7% with
31 August 2015
Rpdef4. Analysis of transcript polymorphism showed that Rpdef3 accounted for about 60% frequency of
Accepted 2 September 2015
Available online 9 September 2015
Rpdefs occurrence in clams from three geographic origins (Dalian, Qingdao and Hangzhou). By quanti-
tative real-time RT-PCR (qRT-PCR) analysis, the transcripts of Rpdefs were mainly detected in hemocytes
and they responded sensitively to bacterial challenge in hemocytes. Evolutionary analysis indicated that
Keywords:
Defensin
all Rpdefs were under positive selection with positively selected basic amino acid residues detected in
Manila clam the C-terminal regions, which perhaps have a functional relevance by modifying the charge distribution
Diversity of Rpdefs. The results also showed some lineages with dN/dS > 1, suggesting positive selection pressures
Positive selection existed in some lineages of phylogeny tree constructed by mollusk defensins. Overall, our results suggest
that Rpdefs perhaps played important roles in host defense and positive selection is the major driving
force in generating high diversity of defensins in the Manila clam.
© 2015 Elsevier Ltd. All rights reserved.

1. Introduction ubiquitous families [1]. Defensins are a collection of small cationic

peptides with molecular weights of approximately 3e5 kDa [14].
Invertebrates exclusively depend on their innate immunity Generally, the animal defensin molecules can be classified into four
which consists of both cellular and humoral defenses [1,2]. The major groups according to their structure and origin: a-defensin, b-
former includes phagocytosis or encapsulation of pathogens with defensin, q-defensin and invertebrate defensin [15]. These defen-
subsequent pathogen destruction via enzyme activity and oxygen sins display similar structural features: the presence of a signal
metabolite release, while the latter includes various reactions peptide at the N-terminal region, followed by the mature peptide
mediated by molecules such as antimicrobial peptides (AMPs) and region which is characterized by 6e8 conserved cysteine residues
proteins [3,4]. In marine invertebrates, AMPs represent the major forming three or four disulfide bonds, and a C-terminal extension
humoral defense system against infection. The modes of action by rich in anionic residues [16]. Defensins have been found to be
which AMPs kill bacteria are diverse, and most of them related to widely distributed in marine invertebrate animals, especially in
plasma membrane disturbance and lethal alteration of microbial mollusks. Presently, multiple defensin molecules have been suc-
integrity [5]. In marine mollusks, several kinds of AMPs have been cessively identified from mussels, oysters, clam and abalone
characterized and studied, including defensins, mytilins, myticins, [14,17e20]. It has been shown that defensins from marine mollusks
mytimycin, big defensins and mytimacins [6e13]. are active against Gram-positive and Gram-negative bacteria and
Among the large number of AMPs, defensin is one of the most fungi, suggesting that they play important roles in innate immune
response of mollusks [6,7,17,18,21,22].
Due to their direct interaction with altered/new pathogens,
* Corresponding author. AMPs exhibit an extraordinary diversity in their structure and
E-mail address: [email protected] (J. Zhao).

http://dx.doi.org/10.1016/j.fsi.2015.09.008
1050-4648/© 2015 Elsevier Ltd. All rights reserved.
 Q. Wang et al. / Fish & Shellfish Immunology 47 (2015) 302e312 303

function [23,24]. Molecular diversity of AMPs such as myticin and whole experiment. Then sixty clams of three geographic origins (20
defensin has been detected in marine mussels and oysters individuals for each location) were immersed with high density of
[8,25,26]. Recently, the defensin from freshwater pearl mussel live Micrococcus luteus and Vibrio anguillarum with a final concen-
Hyriopsis cumingii has also been shown to contain six isoforms [27]. tration of 1  107 CFU mL1 respectively. After 24 h of challenge, the
Sequence diversification of AMPs by gene duplication has been hemocytes, digestive glands and gills of 45 individuals (15 in-
reported for both vertebrates and invertebrates [28,29]. Moreover, dividuals for each location) were sampled and stored in liquid ni-
an increasing number of studies suggested that the evolution of trogen before use.
AMPs is driven by positive selection in both vertebrates and in- For bacterial challenge experiment, adult clams (shell-length:
vertebrates [26,30e32]. ~3.5e4.5 cm) were purchased from a local culturing farm (Yantai)
Although the knowledge on mollusk defensins has been much and acclimatized for 7 days. Then the clams were exposed to
reported, the information on evolutionary pattern of mollusk V. anguillarum at a final concentration of 1  107 CFU mL1. At 12 h,
defensins is still very limited. In this study, four isoforms of 24 h and 48 h intervals following the challenge, the hemocytes of
defensins have been characterized from the Manila clam and their four individuals were sampled and stored in liquid nitrogen.
biochemical properties and structures have been predicted. More- Meanwhile, the hemocytes, gill, digestive gland, mantle and foot of
over, the evolutionary patterns of these defensins from Manila clam four untreated clams were also sampled to determine the tis-
and other mollusks have also been discussed. sueedistribution profiles of Rpdefs.

2. Materials and methods 2.2. Total RNA extraction and sequence amplification

2.1. Animal culture and challenge Frozen tissues were pulverized under liquid nitrogen, and sub-
jected to total RNA extraction using the TRIzol Reagent (Invitrogen,
For mRNA polymorphism characterization, the clams Ruditapes USA). The extracted RNA was then treated with RQ1 RNase-Free
philippinarum (shell length: ~3.0e4.0 cm) were purchased from DNase (Promega, USA) to remove DNA contamination. Single-
culture farms at three different sites (Dalian, Qingdao and Hang- stranded cDNA was synthesized from the total RNA with M-MLV
zhou) (Fig. 1). The clams are acclimated for two week before reverse transcriptase (Promega, USA).
commencement of the experiment. They were maintained in The EST sequences from cDNA library constructed from Manila
filtered seawater at 20e22
C and 30‰ salinity throughout the clam hemocytes (unpublished) were used to construct a blast

Fig. 1. The sampling sites of manila clam Ruditapes philippinarum along the coast of China.
304 Q. Wang et al. / Fish & Shellfish Immunology 47 (2015) 302e312

Table 1
Primers used in the present study.

Primer Sequence (50 -30 ) Sequence information

P1 (reverse) TGGTGCTGTGATGAGTTCTAT0 50 RACE primer

P2 (reverse) TGCACCTCTGACGTAATGT 50 RACE primer
P3 (forward) GGTTTGGTTGCCCTGAAGATGA 30 RACE primer
P4 (forward) GAACTCATCACAGCACCAACA 30 RACE primer
P5 (forward) CAACAGGTTTAGCACTCAACGG Polymorphism detection primer
P6 (reverse) AAACTTGCTTGCGTGTTGGTGC Polymorphism detection primer
P7 (forward) TTGATGCCGGGTTTGGTTG Real time primer for Rpdef1
P8 (reverse) CAACCGTAACAAGTGCACCT Real time primer for Rpdef1
P9 (forward) CCGAAAATGGCTGCCCTAAT Real time primer for Rpdef2
P10 (reverse) GTTACACAGGCACGACAAGT Real time primer for Rpdef2
P11 (forward) AGGACGATGATTGCTTTTACTGT Real time primer for Rpdef3
P12 (reverse) CCCAGCAATCGTTACACCTG Real time primer for Rpdef3
P13 (forward) CGTTGATGGCTGTCGTGTAT Real time primer for Rpdef4
P14 (reverse) CGAGCAAGCGTAACACCTG Real time primer for Rpdef4
P15 (forward) CTCCCTTGAGAAGAGCTACGA Real time primer for b-actin
P16 (reverse) GATACCAGCAGATTCCATACCC Real time primer for b-actin

database using the makeblastdb program available from the NCBI 3D structure of defensins was predicted with Phyre 2 server (Pro-
website. The oyster defensin sequence (GenBank accession no. tein Homology/analogY Recognition Engine V 2.0), and visualized
CAJ19280) was used as a query sequence in a tblastn search (default using the PyMOL software (DeLano, The PyMOL Molecular Graphics
parameters, version 2.2.28þ) against this database to identify ho- System, 2002, http://www.pymol.org). Multiple alignments were
mologues within the database. Then a putative Manila defensin performed with the ClustalW program (http://www.ebi.ac.uk/
sequence was identified and subjected to blastx searches against the Tools/msa/clustalw2/). A maximum likelihood (ML) phylogenetic
NCBI nucleotide database, to confirm its identity. To generate the tree based on the nucleotide sequences of mollusk defensins was
full-length cDNA of R. philippinarum genes (Rpdefs), two reverse constructed using PhyML 3.0 [35]. For phylogenetic analyses, the
primers P1 and P2, and two forward primers P3 and P4 (Table 1), optimum evolutionary models were selected using the jModelTest
were designed based on the EST sequence. The nested PCR strategy program [36]. For ML analysis, 100 bootstraps were used to esti-
was applied to the 30 and 50 RACE. For transcript polymorphism mate the node reliability.
detection, two specific primers (P5 and P6, Table 1) designed in the
50 UTR and 30 UTR were employed to clone the full coding sequence of 2.5. Testing for positive selection
Rpdefs. The PCR profile and subsequent sequencing were conducted
as described previously [33,34]. A total of 126 positive Rpdef clones The nucleotide sequences encoding amino acids of Rpdefs were
were bi-directionally sequenced respectively. used to construct a Maximum likelihood (ML) tree using appro-
priate nucleotide substitution model. The reliability of interior
2.3. Quantitative real-time PCR (qRT-PCR) assay branches of each phylogeny was assessed with 1000 bootstraps.
The phylogeny was used to estimate nonsynonymous to synony-
Quantitative real-time PCR (qRT-PCR) was carried out in an ABI mous substitution rate ratio (u ¼ dN/dS) by the maximum likeli-
7500 Real-time Detection System by using the SYBR ExScript qPCR hood (ML) method implemented in CODEML program of the PAML
Kit (Takara, Japan) as described previously [33]. The PCR amplifi- 4.4 software package [37]. Positive selection can be inferred from a
cation was carried out in a total volume of 50 mL, containing 25 mL of higher proportion of nonsynonymous than synonymous sub-
2  SYBR Green PCR Master Mix, 20 mL of the diluted cDNA, 1 mL of stitutions per site (dN/dS > 1). Likelihood ratio tests (LRTs) were
each of primers (10 mmol/L), and 3 mL of DEPC-treated water. The used to determine whether any codon positions were subjected to
thermal profile for qPCR was 50
C for 2 min, 95
C for 10 min positive selection as indicated by u > 1.
followed by 40 cycles of 95
C for 15 s and 60
C for 1 min. All re- To test for heterogeneous selective pressure at amino acid sites,
actions were run in triplicate. Dissociation curve analysis of the site-specific models were tested: M0 (one-ratio) against M3
amplicons was performed at the end of each PCR reaction to (discrete), M1a (nearly neutral) against M2a (positive selection),
confirm that only one PCR product was amplified and detected. The M7 (beta) against M8 (beta & u). The assumption and parameters of
expressions of Rpdefs were analyzed using the 2DDCT method with each model were as describe previously [33]. The branch model
b-actin gene as the internal control. The primers used to quantify was also conducted using the likelihood ratio test between the one-
the expression of Rpdefs were listed in Table 1. ratio model and the free-ratio model results to detect positive se-
lection acting on particular lineages of the phylogenetic tree. The u
2.4. Sequence analysis, structure prediction and phylogenetic value in the one-ratio model was fixed whereas the value in the
analysis free-ratio model was estimated. The LRTs between nested models
were conducted by comparing twice the difference of the log-
The searches for nucleotide and protein sequence similarities likelihood values (2DL) between two models with the c2 distribu-
were performed with the BLAST algorithm (http://www.ncbi.nlm. tion. The Naive Empirical Bayes (NEB) method and Bayes empirical
nih.gov/blast). The deduced protein sequences were analyzed Bayes (BEB) method were used to calculate the posterior proba-
with ExPASy (http://www.expasy.org/). Signal peptide was pre- bility that each codon was from the site class of positive selection
dicted by SignalP 4.0 server (http://www.cbs.dtu.dk/services/ under models M3, M2a and M8 respectively [38].
SignalP/). Prediction of putative disulfide bonds was performed
using Scratch Protein Predictor (http://scratch.proteomics.ics.uci. 2.6. Statistical analysis
edu/), DISULFIND web-server (http://cassandra.dsi.unifi.it/) and
DiANNA web server (http://clavius.bc.edu/~clotelab/DiANNA/). The SPSS 16.0 software (SPSS Inc., USA) was used for statistical
 Q. Wang et al. / Fish & Shellfish Immunology 47 (2015) 302e312 305

analysis. All data were given in terms of relative mRNA expression alignments indicated that Rpdef1 shared 46.9% identity with
as means ± SE (n ¼ 4). One-way analysis of variance (ANOVA) was Rpdef2, 40.8% with Rpdef3, and 34.7% with Rpdef4. The transcript
performed on all data and P < 0.05 was considered statistically polymorphisms of these defensins have been shown in
significant. Supplemental figures.

3. Results 3.2. Structure prediction

3.1. Sequence analysis of Rpdefs All of the Rpdefs have eight cysteine residues, which are pre-
dicted to form four disulfide bonds. However, the linkage patterns
The 126 cDNA sequences of Rpdefs were deposited in GenBank of the disulfide bridges predicted by different servers were not the
under accession no. JX096678-JX096804. These sequences coded same. Three-dimensional structure of Rpdef1 and Rpdef3 was
for defensins that fell in four defensin categories, which were predicted based on the template of Crassostrea gigas defensin (PDB
named as Rpdef1, Rpdef2, Rpdef3 and Rpdef4, respectively. The ID: 2B68) with confidence ¼ 99.9% (coverage ¼ 86%) and
complete coding sequence of Rpdef1, Rpdef2, Rpdef3 and Rpdef4 confidence ¼ 92.1% (coverage ¼ 84%), respectively. However, 3-D
was of 219 bp, 210 bp, 207 bp and 195 bp in length, which encoded structure of Rpdef2 and Rpdef4 was predicted based on the tem-
a polypeptide of 72, 69, 68 and 64 amino acids, respectively. The plate of actinomycin (PDB ID: 2RU0) and scorpion toxin (PDB ID:
putative signal peptide of Rpdef1, Rpdef2 and Rpdef3 was identified 1SXM), with confidence ¼ 63.3% (coverage ¼ 57%) and
at the N-terminal sequence with the first 24 amino acids, while the confidence ¼ 83.0% (coverage ¼ 55%), respectively. All the predicted
putative signal peptide of Rpdef4 comprised of the first 21 residues. 3-D structures of Rpdefs possessed characteristic features of
The mature peptide of Rpdef1, Rpdef2, Rpdef3 and Rpdef4 con- defensins, which comprised of one a-helix and two antiparallel b-
sisted of 49, 46, 45 and 42 amino acids, which had a theoretical sheets (Fig. 2). Rpdef1 and Rpdef3 had similar speculated 3-D
isoelectric point (pI) of 6.86, 8.23, 8.50 and 8.73 with an increasing structures as C. gigas defensin, while Rpdef2 and Rpdef4 dis-
predicted net charge of 0, þ2, þ3 and þ4. The amino acid played unique 3-D structure respectively.

Fig. 2. Three-dimensional structures of Rpdefs and their counterparts. Cgdef (PDB ID: 2B68), actinomycin (PDB ID: 2RU0), scorpion toxin (PDB ID: 1SXM).
306 Q. Wang et al. / Fish & Shellfish Immunology 47 (2015) 302e312

Table 2 The phylogeny tree was generated using the models GTRþG for
Sequences used for multiple alignments and phylogenetic analysis. mollusk defensins determined by the Akaike information criterion.
Peptide Species MW Da pI Accession numbers Phylogenetic analysis showed that Rpdef2 and Rpdef3 first clus-
MGD1 Mytilusgalloprovincialis 4091 9.03 P80571
tered together, then segregated with Rpdef1, Rpdef4 and McDef
MGD2 Mytilusgalloprovincialis 4126 8.81 AAD52660 successively and formed as a subclade (Fig. 4). The subclade of
MedefA Mytilusedulis 4151 9.15 P81610 defensins from Manila clam first rooted with defensins from fresh
MedefB Mytilusedulis 4271 9.18 P81611 water mussels Hyriopsis schlegelii and Hyriopsis cumingii, further
Cgdef Crassostreagigas 4642 8.73 CAJ19280
clustered with defensins from oysters and mussels and formed a
Cgdefh1 Crassostreagigas 4763 8.50 ABD66301
Cgdefh2 Crassostreagigas 4677 8.51 ABD66302 clade with RpdefB (a defensin deposited in the Genbank database)
Cvdef Crassostreavirginica 4265 9.18 P85008 from the Manila clam. At last, these defensins grouped with
Hsdef Hyriopsisschlegelii 4877 7.78 AEJ86348 defensins from the abalone and zebra mussel. These results indi-
Hssdef Haliotis discus discus 4902 7.85 ACZ15982
cated that RpdefB diverged from Rpdefs and McDef before the
Dpdef Dreissenapolymorpha 5684 6.27 ACZ02692
RpdefB Ruditapesphilippinarum 5287 7.79 AEK78067
divergence of Manila clam with pacific oyster and Mediterranean
MCdef Ruditapesphilippinarum 4975 8.71 Adhya et al., 2012 mussel.
Rpdef1 Ruditapesphilippinarum 5435 6.86 AFP50047
Rpdef2 Ruditapesphilippinarum 5430 8.23 AFP49990
Rpdef3 Ruditapesphilippinarum 5330 8.50 AFP49946 3.4. mRNA polymorphism and diversity of Rpdefs
Rpdef4 Ruditapesphilippinarum 4749 8.72 AFP49977
Hcdef1 Hyriopsiscumingii 4264 8.50 Ren et al., 2011 A total of 126 clones were sequenced and virtually translated
Hcdef2 Hyriopsiscumingii 4275 8.51 Ren et al., 2011
into amino acid sequences. Rpdef3 accounted for 60% frequency of
Hcdef3 Hyriopsiscumingii 4906 8.32 Ren et al., 2011
Hcdef4 Hyriopsiscumingii 4880 8.66 Ren et al., 2011 occurrence of the clones, with Rpdef1, Rpdef2 and Rpdef4 ac-
Hcdef5 Hyriopsiscumingii 7112 8.48 Ren et al., 2011 counting for 17%, 13% and 10% of the clones (Table 3). Multiple
Hcdef6 Hyriopsiscumingii 4602 8.50 Ren et al., 2011 alignments indicated that Rpdef1 had 8 different kinds of amino
acid sequences of the 22 sequences, as for Rpdef2 with 4 out of 17
sequences, Rpdef3 with 16 out of 75 sequences and Rpdef4 with 5
3.3. Multiple alignment and phylogenetic relationships out of 12 sequences. The above results suggested that the tran-
script variation of Rpdef1 and Rpdef4 was larger than that of
The sequences used for multiple alignments and phylogenetic Rpdef2 and Rpdef3. The phylogeny trees were also constructed to
analysis were shown in Table 2. Multiple alignments indicated that compare the sequence variance between these four Rpdef genes
eight cysteine residues and C-terminal motif (-RRSIQ-) were highly with PhyML 3.0. The models employed for phylogeny analyses
conserved in Rpdefs (Fig. 3a), whereas only four cysteine residues (HKYþG for Rpdef1, HKYþGþI for Rpdef2, GTR for Rpdef3, and
were conserved in mollusk defensins (Fig. 3b). Rpdefs had eight HKYþI for Rpdef4) were determined by the Akaike information
cysteine residues like defensins from the Manila clam (McDef), criterion. The phylogeny analysis indicated that the Rpdef1 se-
pacific oyster (Cgdef, Cgdefh1 and Cgdefh2), Mediterranean mussel quences are more divergent than those of Rpdef2, Rpdef3 and
(MGD1, MGD2) and triangle-shell pearl mussel (Hcdef5, Hcdef6), Rpdef4 (Fig. 5).
while some mollusk defensins contain a pattern of six conserved With regard to clams from Dalian, frequencies of occurrence
cysteine residues (Fig. 3b). were extremely different, with Rpdef3 accounting for 82% of the

Fig. 3. (a) Multiple alignments of Rpdef1 with Rpdef2, Rpdef3 and Rpdef4. (b) Multiple alignments of Rpdefs with other mollusk defensins deposited in GenBank. The black shadow
region indicates positions where all sequences share the same amino acid residue. Gaps are indicated by dashes to improve the alignment. The GenBank accession numbers and the
species are shown in Table 2.
 Q. Wang et al. / Fish & Shellfish Immunology 47 (2015) 302e312 307

Fig. 4. Phylogenetic tree constructed by maximum likelihood method based on the nucleotide sequences of defensins from mollusks. Numbers at the forks indicate the bootstrap
values (in %) out of 100 replicates. The sequences used to construct phylogeny trees of defensins are shown in Table 2.

clones and Rpdef4 present in only one clone. Other coding se- 3.6. Temporal expression profiles of Repdefs mRNAs in hemocytes
quences have been found in 3e5 clones. However, as for clams from post bacterial challenge
Qingdao and Hangzhou, Rpdef3 accounted for 43% and 48% of the
clones; Rpdef2 and Rpdef4 showed the least frequency of occur- Following bacterial challenge, the expression levels of all Rpdefs
rence (Table 3). in hemocytes (Fig. 7) increased significantly at 12 h post challenge
As concerned to tissue distribution of the different Rpdef tran- (P < 0.05). For Rpdef1, the expression level returned to the original
scripts, 52 clones were present in gills, whereas 36 and 38 clones level at 24 h and 48 h post challenge (Fig. 7A). However, the tran-
were present in hemocytes and digestive gland respectively. As for scripts of Rpdef2 and Rpdef3 were significantly inhibited at 24 h
Rpdef1, 50% clones were present in hemocytes, while for Rpdef2 and 48 h following challenge (P < 0.05) (Fig. 7B, C). The expression
and Rpdef3, the clones were almost averagely present in three level of Rpdef4 was significantly up-regulated at 12 h and 24 h, and
tissues. The clones of Rpdef4 were predominantly present in gills down-regulated at 48 h post challenge (P < 0.05) (Fig. 7D).
and digestive gland (Table 3).
3.7. Evolutionary analysis of Rpdefs and other mollusk defensin
genes

3.5. Tissue-specific expressions profiles of Rpdefs mRNAs

Phylogeny-based codon substitution models were used to
identify the codons under positive selection. For all Rpdefs, the
The tissue distribution of Repdef mRNAs was investigated by
M0eM3 comparison revealed that M3 was not a better fit to the
qRT-PCR with b-actin as internal control. During the qRT-PCR as-
data than M0 (2DІ ¼ 130; P > 0.05). The comparison of null models
says, only one peak was detected at the corresponding melting
(M1a and M7) with their corresponding alternative models (M2a
temperature in the dissociation curve analysis, suggesting that the
and M8) rejected the null models (Table 4), thus indicated that the
PCR was specifically amplified. The transcripts of Rpdefs were
variants of Rpdefs have evolved under positive selection with the
detected in all the tissues examined, including gills, digestive gland,
positively selected sites R63 and R64 falling in the C-terminus. For
hemocytes, mantle and foot. The transcripts of Rpdefs with the
Rpdef1, Rpdef2 and Rpdef4, the comparison of null models with
exception of Rpdef4 were dominantly expressed in hemocytes,
their corresponding alternative models agreed the null models (P
moderately expressed in gills, mantle and foot (Fig. 6A, B, C), and
>0.05), suggesting no amino acid residue under positive selection.
least detected in digestive gland. However, the expression level of
However, the amino acid residues R63, R64, S65, I66 and Q67 in Rpdef3
Rpdef4 mRNA was high in gills and hemocytes, moderate in mantle
were detected under positive selection (Table 5).
and digestive gland, and least in foot (Fig. 6D).
The M1eM0 comparison revealed that model M1 was better fit
to the data than M0 (P < 0.05), indicating the u ratios were
Table 3 different among lineages of the phylogeny tree. The lineage-
Number of clones of Rpdefs in different location and different tissues. specific selection test showed that the u values along most
Genes Number of clones
examined lineages (28 out of 39) were less than 1, whereas there
were also some branches and internal branches with high u
Location Tissue
values (Table 6). As shown in Fig. 6, the Rpdef3 lineage and Rpdefs
Dalian Qingdao Hangzhou Hemcoytes Gill Digestive gland branch (formed by Rpdef1, Rpdef2, Rpdef3 and Rpdef4) had u
Rpdef1 5 11 6 11 8 3 ratios >1, indicating Rpdef3 lineage and Rpdefs branch undergo
Rpdef2 3 6 8 6 6 5 positive selection. In the clade of oysters, the u values for the
Rpdef3 40 20 15 18 32 25 Cgdefh2 lineage and Cgdefs branch (formed with Cgdefh1 and
Rpdef4 1 9 2 1 6 5
Cgdefh2) were more than 1. There were also three lineages with
308 Q. Wang et al. / Fish & Shellfish Immunology 47 (2015) 302e312

Fig. 5. Phylogenetic tree constructed by maximum likelihood method based on the nucleotide sequences of Rpdefs.

very high u ratios in the clade of fresh water pearl mussel, 4. Discussion
implying positive selection on these lineages. However, the u ratio
value for the Mediterranean mussel branch was <1, implying no Marine mollusks account for a large quantity of current global
positive selection on this branch. The Manila clam branch aquaculture output. With the rapid development of intensive
(without RpdefB) had high u ratios >1, revealing that this branch mariculture, some cultured species have been seriously affected by
is under positive selection pressure during evolution (Fig. 8). In diseases and mortalities in recent years [39,40]. Therefore, basic
addition, there were also some clades from different taxon knowledge on the innate immunity of commercially important
exhibited high u ratios. The site-specific models were used to test mollusks is urgently needed, especially characterization of the
for heterogeneous selective pressure at amino acid sites. The immune-associated molecules and their functions. AMPs constitute
M1aeM2a comparison revealed that M2a was better fit to the data an important first-line defense of the immune system in mollusks.
(P < 0.01). LRTs also gave significantly better results for M8 Among these naturally occurring antibiotic peptides, defensins
(P < 0.01). Under the M2a model, 23 amino acids within the form a unique family of cysteine-rich cationic and structured
mature peptide regions are under positive selection, while 17 polypeptides, serving as effector molecules of innate immunity
amino acid residues are positively selected under the M8 model. [41]. Although Mollusca are the largest and most diverse phylum of
We considered a site under positive selection if the BEB posterior animals next to arthropods, AMPs such as defensins have been
probability is > 0.95. Under M2a and M8 model, 14 and 6 posi- characterized only in a few farmed species, such as oysters, mussels
tively selected sites were detected with BEB posterior probability and abalones. Therefore, there is still a great potential to unveil new
>0.95 respectively (Table 6). AMP molecules in this phylum [20].
 Q. Wang et al. / Fish & Shellfish Immunology 47 (2015) 302e312 309

Fig. 6. Tissue-specific expression profiles of Rpdefs mRNAs measured by qRT-PCR. The mRNA expression level is calculated relative to b-actin expression. Each symbol and vertical
bar represents the mean ± SE (n ¼ 4). A-Rpdef1, B-Rpdef2, C-Rpdef3, D-Rpdef4.

Fig. 7. Temporal expression profiles of Rpdefs (A-Rpdef1, B-Rpdef2, C-Rpdef3, D-Rpdef4) in hemocytes post V. anguillarum challenge. The mRNA expression level is calculated
relative to b-actin expression. Each symbol and vertical bar represents the mean ± SE (n ¼ 4). Significant difference from control is indicated with an asterisk at P < 0.05.

Table 4
Comparison among different nested models to test for positive selection among codons of all Rpdef sequences.

Model lnL Estimates of parameters 2DІ Positively selected sites

M1a (nearly neutral) 1702 P0 ¼ 0.25618, (p1 ¼ 0.74382) 46 Not allowed

M2a (positively selection) 1679 p0 ¼ 0.22939, p1 ¼ 0.73902, (p2 ¼ 0.03159), u0 ¼ 0.08282, (u1 ¼ 1), u2 ¼ 14.61681 P < 0.01 63R*, 64R*
M7 (beta) 1703 p ¼ 0.43916, q ¼ 0.33694 46 Not allowed
M8 (beta & w > 1) 1680 p0 ¼ 0.96839, (p1 ¼ 0.03161), p ¼ 0.14181, q ¼ 0.04183, us ¼ 14.21207 P < 0.01 63R*, 64R*
310 Q. Wang et al. / Fish & Shellfish Immunology 47 (2015) 302e312

Table 5
Comparison among different nested models to test for positive selection among codons of Rpdef1, Rpdef2, Rpdef3 and Rpdef4.

Model lnL Estimates of parameters 2D І Positively selected sites

Rpdef1 M1a (nearly neutral) 700.29 p0 ¼ 0.25658, (p1 ¼ 0.74342) 4.62 Not allowed
M2a (positively 697.98 p0 ¼ 0.08753, p1 ¼ 0.69312, (p2 ¼ 0.21935), u0 ¼ 0.00000, P > 0.05 4M, 10F, 28P, 29D, 32Y, 40D, 50D, 51A,
selection) (u1 ¼ 1), u2 ¼ 7.59516 52W, 53T, 55R, 56H, 65K, 67R
M7 (beta) 700.34 p ¼ 0.14734, q ¼ 0.04084 4.82 Not allowed
M8 (beta & w > 1) 697.93 p0 ¼ 0.78255, (p1 ¼ 0.21745), p ¼ 0.10482, q ¼ 0.02625, P > 0.05 4M, 10F, 28P, 29D*, 30D, 32Y, 40D, 50D,
us ¼ 6.67735 51A, 52W, 53T, 55R, 56H, 65K, 67R
Rpdef2 M1a (nearly neutral) 319.19 p0 ¼ 0.00001, (p1 ¼ 0.99999) 5.00 Not allowed
M2a (positively 316.69 p0 ¼ 0.00000, p1 ¼ 0.67091, (p2 ¼ 0.32909), u0 ¼ 1.00000, P > 0.05 all amino acids
selection) (u1 ¼ 1), u2 ¼ 999.00000
M7 (beta) 319.19 p ¼ 2.02546, q ¼ 0.00500 5.00 Not allowed
M8 (beta & w > 1) 316.69 p0 ¼ 0.66973, (p1 ¼ 0.3302), p ¼ 0.00500, q ¼ 1.57842, P > 0.05 all amino acids
us ¼ 999.00000
Rpdef3 M1a (nearly neutral) 733.46 p0 ¼ 0.78832, (p1 ¼ 0.21168) 63.74 Not allowed
M2a (positively 701.59 p0 ¼ 0.56707, p1 ¼ 0.35845, (p2 ¼ 0.07449), u0 ¼ 0.07666, P < 0.01 63R*, 64R*, 65S*, 66I*, 67Q*
selection) (u1 ¼ 1), u2 ¼ 21.89229
M7 (beta) 734.99 p ¼ 0.05393, q ¼ 0.06892 66.66 Not allowed
M8 (beta & w > 1) 701.69 p0 ¼ 0.92554, (p1 ¼ 0.07446), p ¼ 0.06396, q ¼ 0.07578, P < 0.01 63R*, 64R*, 65S*, 66I*, 67Q*
us ¼ 20.39602
Rpdef4 M1a (nearly neutral) 320.09 p0 ¼ 0.00001, (p1 ¼ 0.99999) 0.50 Not allowed
M2a (positively 319.84 p0 ¼ 0.40603, p1 ¼ 0.00000, (p2 ¼ 0.59397), u0 ¼ 0.00000, P > 0.05 14A, 19D, 46N
selection) (u1 ¼ 1), u2 ¼ 2.45542
M7 (beta) 320.09 p ¼ 1.18224, q ¼ 0.00500 0.50 Not allowed
M8 (beta & w > 1) 319.84 p0 ¼ 0.40603, (p1 ¼ 0.59397), p ¼ 0.00845, q ¼ 1.78436, P > 0.05 14A, 19D, 24H, 26Y, 27H, 34Y, 42Y, 46N
us ¼ 2.45543

Table 6
Parameter estimates and log-likelihood values under different models of variable u ratios among sites. Site numbers and amino acids refer to the Rpdef1 sequence.

Model Model code lnL Estimates of parameters 2Dl P Value Positively selected sites

Branch One ratio model 3427.1 u ¼ 0.39891 57.2 P < 0.05 NA

model (df ¼ 38)
Free-ratio 3398.5 u estimated independently for each
model branch (see Fig. 4)
Site model M1a 3426.9 P0 ¼ 0.01754 (p1 ¼ 0.98246) 51.2 P < 0.01 Not allowed
M2a 3401.3 p0 ¼ 0.01756, p1 ¼ 0.63873, 28P*, 31E*, 32Y**, 33E, 34C, 35H**, 37H, 38C*, 40D**, 41S,
(p2 ¼ 0.34371), u0 ¼ 0.00000, 43G, 44C, 47G, 48Y**, 49C*, 50D, 51A**, 52W*, 53T**, 54L**,
(u1 ¼ 1), u2 ¼ 4.82950 55R**,56H**, 57R
M7 3363.9 p ¼ 0.06961, q ¼ 0.09195 126 P < 0.01 Not allowed
M8 3426.9 p0 ¼ 0.01754, p ¼ 0.00500, 28P, 31E, 32Y**, 33E, 35H**, 38C, 40D*, 43G, 48Y, 49C, 50D,
q ¼ 0.00500, (p1 ¼ 0.00500), 51A**, 52W, 53T, 54L** 55R,56H**
w ¼ 1.00000

Fig. 8. Positive selection analyses of defensins across the mollusk phylogeny. The numbers shown along each branch are the dN/dS values for the entire gene along that branch.

In this study, we have characterized four types of defensins from identity with Rpdef2, 40.8% with Rpdef3, and 34.7% with Rpdef4,
the commercially important Manila clam. Presently, six defensin whereas it had only 18.4% identity with RpdefB. Like other mollusk
isoforms including McDef and RpdefB (deposited in the Genbank defensins, all of these defensins from Manila clam with the exception
database) have been isolated from this species. The amino acid of Rpdef1 had a net positive charge, and their theoretical pIs were
alignments indicated that Rpdef1 shared 65.3% with McDef, 46.9% more than 8.0 (Table 2). However, Rpdef1 and defensin from Dreissena
 Q. Wang et al. / Fish & Shellfish Immunology 47 (2015) 302e312 311

polymorpha (Dpdef) had a theoretical pI of less than 7, and possessed a was the major driving force in the generation of diverse AMPs [47].
net charge of 0 and 1, respectively. Although the vast majority of Consistent with earlier studies on molecular evolution of several
AMPs are cationic in nature, a significant number of anionic AMPs AMPs [25e31], our results suggested that all Rpdefs evolved
have also been reported. As for the non-cationic defensins, perhaps through positive selection. Positively selected codons R63 and R64
they can dock to the bacterial membrane via the positively charged had been detected in the C-terminal regions of all these Rpdefs. In
and hydrophobic loop near their C-terminus [42]. addition, the amino acid residues S65, I66 and Q67 in the C-terminal
Rpdef3 sequences accounted for the most percentage of the region of Rpdef3 were also detected under positive selection.
Rpdefs sequences, suggesting this defensin might play a more Similarly, the majority of positively selected sites are located in the
important role than other Rpdefs in this clam. The occurrence C-terminal region of myticin-C [25,48]. Based on the predicted
frequencies of different Rpdefs varied in clams from different tertiary structure of Rpdefs, the positively selected amino acids fall
geographical locations. The transcript variation was perhaps related in the predicted C-terminal coils (data not shown), which are highly
to pathogen load in clams and other possible environmental factors exposed at the surface of Rpdefs. Moreover, the positively charged
of different sites, which may have forced the defensin genes to R63 and R64 perhaps increase the binding ability of peptide to the
diversify [43]. The transcript sequences of Rpdef2 and Rpdef3 were bacteria. It had been demonstrated that positively charged amino
almost averagely present in three tissues, while the transcript se- acids on the surface of the defensin could greatly improve their
quences of Rpdef1 were mainly present in hemocytes and Rpdef4 antibacterial activity, probably by promoting a better binding to the
were mainly detected in gills and digestive gland. These above re- cell wall or membrane of target bacteria [49,50]. Therefore, the
sults indicated different Rpdefs perhaps played different immune positively selected amino acids might have a functional relevance
functions in specific tissues. In addition, multiple alignment and by modifying the charge distribution of Rpdefs. Further investiga-
phylogeny analysis showed that Rpdef1 had a higher diversity than tion on the functional roles of C-terminal domain in defensins will
other Rpdefs, while Rpdef3 had the lowest diversity. The diver- be crucial to interpret these observations [48].
gence of Rpdef1 was likely due to purifying selection acting in a The results of the present study provide strong evidence that
long term manner, while the polymorphism reduction of Rpdef3 positive selection is the major driving force in generating high di-
was probably caused by a recent selective pressure [26], which has versity in Rpdefs. Zhu et al. [47] also reported that many members
been detected by evolutionary analysis in the present study. of the CS-ab superfamily exhibited molecular diversity and diverse
In this study, the constitutive expression of all Rpdefs was biological functions, and suggested that positive Darwinian selec-
mainly detected in hemocytes, which are primarily responsible for tion is the major driving force in generating such diversity. The
the defense against pathogens in bivalves. Similarly, it has been direct involvement of these peptides with the altered pathogens in
reported that oyster defensin (Cgdefh2) and mussel defensins a changing environment is probably the cause of such adaptive
(MGD1 and MGD2) were abundantly expressed in hemocytes molecular evolution of these AMPs [25].
[14,16,21]. As for Rpdef4, the highest expression was found in gills,
which is continuously exposed to environmental stress factors such 5. Conclusions
as toxic substances and pathogens. These types of tis-
sueedistribution profiles suggested these defensins could respond We characterized the complete coding sequences and predicted
promptly to bacterial challenge. However, it was reported that the 3-D structures of four defensins from the Manila clam. Next, we
MCdef transcript was expressed at the highest level in adductor studied the diversity of defensins from three geographical clam
muscle of the Manila clam [32]; and the abalone defensin transcript populations and found that Rpdef3 accounted for about 60% fre-
was highly expressed in the mantle and hepatopancreas [19]. Thus, quency of Rpdefs occurrence in these clams. The expression profiles
these results demonstrated that these defensins may play different post bacterial challenge suggested that Rpdefs were involved in the
immune roles in different tissues of these bivalves. host defense in this clam. Furthermore, it was found that Rpdef3
In order to further understand the possible biological functions and all Rpdefs were under positive selection with positively
of the Rpdefs, their mRNA expression patterns post bacterial chal- selected amino acid residues detected in the C-terminal regions,
lenge were examined at different time intervals in hemocytes. The which perhaps have a functional relevance by modifying the charge
expression of all Rpdefs was up-regulated remarkably after 12 h distribution of Rpdefs. Therefore, positive selection could be the
bacterial challenge compared to that of the control, indicating that major driving force in generating high diversity of defensins in the
the transcripts of Rpdefs were induced to fight against the bacteria. Manila clam.
The up-regulation of defensin transcripts was also observed in disk
abalone [19], Manila clam (MCdef) [32] and freshwater pearl Acknowledgements
mussel Hyriopsis schlegelii [44]. Then the expression levels of all
Rpdefs except Rpdef4 returned to original level or were inhibited at We thank Dr. Xiutang Yuan, Dr. Tianlong Qiu and Dr. Changkao
24 h and 48 h post challenge. The possible reason of down- Mu for their help in sample collection. This research was supported
regulation was that there was no need to produce large amount by grants from NSFC (Grant No. 41476126), the Strategic Priority
of defensins with the progressive clearance of the invasive bacteria. Research Program of the Chinese Academy of Science (Grant No.
In fact, the expression of oyster defensin (Cgdefh2) and mussel XDA11020405) and Key Research Program of the Chinese Academy
defensin (MGD2) was also found to be significantly down-regulated of Sciences (Grant No. KZZD-EW-14).
in hemocytes after bacterial infection [14,16]. Above all, the tran-
scriptional up-regulation of Rpdefs against bacterial challenge as Appendix A. Supplementary data
well as their highly constitutive expression in hemocytes indicated
that they may play important roles in innate immunity. Supplementary data related to this article can be found at http://
It has been shown that genes involved in the immune system of
dx.doi.org/10.1016/j.fsi.2015.09.008
various animals typically show a faster rate of amino acid sub-
stitutions and have evolved under positive selection [45,46]. Due to
References
their direct interaction with altered/new pathogens, AMPs exhibit
an extraordinary diversity in their biochemical and biological } derha
[1] K. So €ll, Invertebrate immunity, Dev. Comp. Immunol. 23 (1999)
functions. It has been reported that positive Darwinian selection 263e266.
312 Q. Wang et al. / Fish & Shellfish Immunology 47 (2015) 302e312

[2] E.S. Loker, C.M. Adema, S.M. Zhang, T.B. Kepler, Invertebrate immune systems- antimicrobial peptide variants in the Mediterranean mussel, Mytilus gallo-
not homogeneous, not simple, not well understood, Immunol. Rev. 198 (2004) provincialis, Peptides 29 (2008) 1094e1101.
10e24. [26] P. Schmitt, Y. Gueguen, E. Desmarais, E. Bache re, J. de Lorgeril, Molecular di-
[3] M. Zasloff, Antimicrobial peptides of multicellular organisms, Nature 415 versity of antimicrobial effectors in the oyster Crassostrea gigas, BMC Evol.
(2002) 389e395. Biol. 10 (2010) 23.
[4] L. Song, L. Wang, L. Qui, H. Zhang, Bivalve Immunity, in: K. Soderhall (Ed.), [27] Q. Ren, M. Li, C.Y. Zhang, K.P. Chen, Six defensins from the triangle-shell pearl
Advances in Experimental Medicine and Biology e Invertebrate Immunity, mussel Hyriopsis cumingii, Fish. Shellfish Immunol. 31 (2011) 1232e1238.
Springer, New York, USA, 2010, p. 708, 44e65. [28] A. Patil, A.L. Hughes, G. Zhang, Rapid evolution and diversification of
[5] A.J. Otero-Gonz alez, B.S. Magalha ~es, M. Garcia-Villarino, C. Lo
pez-Abarrategui, mammalian alpha-defensins as revealed by comparative analysis of rodent
D.A. Sousa, S.C. Dias, et al., Antimicrobial peptides from marine invertebrates and primate genes, Physiol. Genomics 20 (2004) 1e11.
as a new frontier for microbial infection control, FASEB J. 24 (2010) [29] N. Pujol, O. Zugasti, D. Wong, C. Couillault, C.L. Kurz, H. Schulenburg,
1320e1334. J.J. Ewbank, Anti-fungal innate immunity in C. elegans is enhanced by evolu-
[6] M. Charlet, S. Chernysh, H. Philippe, C. Hetru, J.A. Hoffmann, P. Bulet, Innate tionary diversification of antimicrobial peptides, PLoS Pathog. 4 (7) (2008)
immunity. Isolation of several cysteine-rich antimicrobial peptides from the e1000105.
blood of a mollusc, Mytilus edulis, J. Biol. Chem. 271 (1996) 21808e21813. [30] J.A. Tennessen, Molecular evolution of animal antimicrobial peptides: wide-
[7] G. Mitta, F. Hubert, T. Noe €l, P. Roch, Myticin, a novel cysteine-rich antimi- spread moderate positive selection, J. Evol. Biol. 18 (2005) 1387e1394.
crobial peptide isolated from haemocytes and plasma of the mussel Mytilus [31] A. Padhi, B. Verghese, S.K. Otta, B. Varghese, K. Ramu, Adaptive evolution after
galloprovincialis, Eur. J. Biochem. 265 (1999) 71e78. duplication of penaeidin antimicrobial peptides, Fish. Shellfish Immunol. 23
[8] A. Pallavicini, M.M. Costa, C. Gestal, R. Dreos, A. Figueras, et al., High sequence (2007) 553e566.
variability of myticin transcripts in hemocytes of immune-stimulated mussels [32] M. Adhya, H.D. Jeung, H.S. Kang, K.S. Choi, D.S. Lee, M. Cho, Cloning and
suggests ancient host-pathogen interactions, Dev. Comp. Immunol. 32 (2008) localization of MCdef, a defensin from Manila clams (Ruditapes philippinarum),
213e226. Comp. Biochem. Physiol. B 161 (2012) 25e31.
[9] M.M. Costa, S. Dios, J. Alonso-Gutie rrez, A. Romero, B. Novoa, et al., Evidence of [33] Q. Wang, L. Zhang, J. Zhao, L. You, H. Wu, Two goose-type lysozymes in Mytilus
high individual diversity on myticin C in mussel (Mytilus galloprovincialis), galloprovincialis: possible function diversification and adaptive evolution,
Dev. Comp. Immunol. 33 (2009) 162e170. PLoS One 7 (9) (2012) e45148.
[10] M.G. Parisi, H. Li, M. Toubiana, N. Parrinello, M. Cammarata, P. Roch, Poly- [34] Q. Wang, C. Wang, C. Mu, H. Wu, L. Zhang, J. Zhao, A novel C-type lysozyme
morphism of mytilin B mRNA is not translated into mature peptide, Mol. from Mytilus galloprovincialis: insight into innate immunity and molecular
Immunol. 46 (2009) 384e392. evolution of invertebrate C-type lysozymes, PLoS One 8 (6) (2013) e67469.
[11] J. Zhao, L. Song, C. Li, D. Ni, L. Wu, L. Zhu, et al., Molecular cloning, expression [35] S. Guindon, J.F. Dufayard, V. Lefort, M. Anisimova, W. Hordijk, O. Gascuel, New
of a big defensin gene from bay scallop Argopecten irradians and the antimi- algorithms and methods to estimate maximum-likelihood phylogenies:
crobial activity of its recombinant protein, Mol. Immunol. 44 (2007) 360e368. assessing the performance of PhyML 3.0, Syst. Biol. 59 (2012) 307e321.
[12] R.D. Rosa, A. Santini, J. Fievet, P. Bulet, D. Destoumieux-Garzo n, E. Bache
re, Big [36] D. Posada, jModelTest: phylogenetic model averaging, Mol. Biol. Evol. 25
defensins, a diverse family of antimicrobial peptides that follows different (2008) 1253e1256.
patterns of expression in hemocytes of the oyster Crassostrea gigas, PLoS One [37] Z. Yang, PAML 4: phylogenetic analysis by maximum likelihood, Mol. Biol.
6 (9) (2011) e25594. Evol. 24 (2007) 1586e1591.
[13] M. Gerdol, G. De Moro, C. Manfrin, P. Venier, A. Pallavicini, Big defensins and [38] Z. Yang, W.S.W. Wong, R. Nielsen, Bayes empirical Bayes inferences of amino
mytimacins, new AMP families of the Mediterranean mussel Mytilus gallo- acid sites under positive selection, Mol. Biol. Evol. 22 (2005) 1107e1118.
provincialis, Dev. Comp. Immunol. 36 (2012) 390e399. [39] X.M. Guo, S.E. Ford, F.S. Zhang, Molluscan aquaculture in China, J. Shell Res. 18
[14] M. Gonzalez, Y. Gueguen, G. Desserre, J. de Lorgeril, B. Romestand, E. Bachere, (1999) 19e31.
Molecular characterization of two isoforms of defensin from hemocytes of the [40] pin, I. Arzul, B. Morga, N. Faury, T. Renault, Detection and
A. Segarra, J.F. Pe
oyster Crassostrea gigas, Dev. Comp. Immunol. 31 (2007) 332e339. description of a particular Ostreid herpesvirus 1 genotype associated with
[15] W. Xu, M. Faisal, Defensin of the zebra mussel (Dreissena polymorpha): mo- massive mortality outbreaks of Pacific oysters, Crassostrea gigas, in France in
lecular structure, in vitro expression, antimicrobial activity, and potential 2008, Virus Res. 153 (2010) 92e99.
functions, Mol. Immunol. 47 (2010) 2138e2147. [41] T. Ganz, Defensins and host defense, Science 286 (1999) 420e421.
[16] G. Mitta, F. Vandenbulcke, P. Roch, Original involvement of antimicrobial [42] R. Lai, L.O. Lomas, J. Jonczy, P.C. Turner, H.H. Rees, Two novel non-cationic
peptides in mussel innate immunity, FEBS Lett. 486 (2000) 185e190. defensin-like antimicrobial peptides from haemolymph of the female tick,
[17] J.K. Seo, J.M. Crawford, K.L. Stone, E.J. Noga, Purification of a novel arthropod Amblyomma hebraeum, Biochem. J. 379 (2004) 681e685.
defensin from the American oyster, Crassostrea virginica, Biochem. Biophys. [43] M. Sonthi, M. Toubiana, A. Pallavicini, P. Venier, P. Roch, Diversity of coding
Res. Commun. 338 (2005) 1998e2004. sequences and gene structures of the antifungal peptide mytimycin (MytM)
[18] Y. Gueguen, A. Herpin, A. Aumelas, J. Garnier, J. Fievet, J.M. Escoubas, et al., from the Mediterranean mussel, Mytilus galloprovincialis, Mar. Biotechnol. 13
Characterization of a defensin from the oyster Crassostrea gigas. Recombinant (2011) 857e867.
production, folding, solution structure, antimicrobial activities, and gene [44] K. Peng, J.H. Wang, J.Q. Sheng, L.G. Zeng, Y.J. Hong, Molecular characterization
expression, J. Biol. Chem. 281 (2006) 313e323. and immune analysis of a defensin from freshwater pearl mussel, Hyriopsis
[19] M. De Zoysa, I. Whang, Y. Lee, S. Lee, J.S. Lee, J. Lee, Defensin from disk abalone schlegelii, Aquaculture 334 (2012) 45e50.
Haliotis discus discus: molecular cloning, sequence characterization and im- [45] J. Trowsdale, P. Parham, Mini-review: defense strategies and immunity
mune response against bacterial infection, Fish. Shellfish Immunol. 28 (2010) related genes, Eur. J. Immunol. 34 (2004) 7e17.
261e266. [46] T.B. Sackton, B.P. Lazzaro, T.A. Schlenke, J.D. Evans, D. Hultmark, et al., Dy-
[20] G.A. Díaz, Defensins and cysteine rich peptides: two types of antimicrobial namic evolution of the innate immune system in Drosophila, Nat. Genet. 39
peptides in marine mollusks, Inv Surv. J. 7 (2010) 157e164. (2007) 1461e1468.
[21] G. Mitta, F. Vandenbulcke, F. Hubert, P. Roch, Mussel defensins are synthesised [47] S. Zhu, B. Gao, J. Tytgat, Phylogenetic distribution, functional epitopes and
and processed in granulocytes then released into the plasma after bacterial evolution of the CSab superfamily, Cell Mol. Life Sci. 62 (2005) 2257e2269.
challenge, J. Cell Sci. 112 (1999) 4233e4242. [48] M. Vera, P. Martínez, L. Poisa-Beiro, A. Figueras, B. Novoa, Genomic organi-
[22] C. Gestal, M. Costa, A. Figueras, B. Novoa, Analysis of differentially expressed zation, molecular diversification, and evolution of antimicrobial peptide
genes in response to bacterial stimulation in hemocytes of the carpet-shell myticin-C genes in the mussel (Mytilus galloprovincialis), PLoS One 6 (8)
clam Ruditapes decussatus: identification of new antimicrobial peptides, (2011) e24041.
Gene 406 (2007) 134e143. [49] T. Schneider, T. Kruse, R. Wimmer, I. Wiedemann, V. Sass, U. Pag, et al.,
[23] P. Nicolas, D. Vanhoye, M. Amiche, Molecular strategies in biological evolution Plectasin, a fungal defensin, targets the bacterial cell wall precursor lipid II,
of antibacterial peptides, Peptides 24 (2003) 1669e1680. Science 328 (2010) 1168e1172.
[24] C.A. Semple, P. Gautier, K. Taylor, J.R. Dorin, The changing of the guard: mo- [50] P. Schmitt, M. Wilmes, M. Pugnie re, A. Aumelas, E. Bache re, H.G. Sahl, et al.,
lecular diversity and rapid evolution of beta-defensins, Mol. Divers 10 (2006) Insight into invertebrate defensin mechanism of action: oyster defensins
575e584. inhibit peptidoglycan biosynthesis by binding to lipid II, J. Biol. Chem. 38
[25] A. Padhi, B. Verghese, Molecular diversity and evolution of myticin-C (2010) 29208e29216.