Gene 526 (2013) 210–216

Contents lists available at SciVerse ScienceDirect

Gene

j ourna l homepage: www.e lsev ie r .com/ locate /gene

Molecular cloning and characterization of perlucin from the freshwater pearl mussel,Hyriopsis cumingii

Jing-Yun Lin a, Ke-Yi Ma a, Zhi-Yi Bai a, Jia-Le Li a,b,⁎a Key Laboratory of Freshwater Aquatic Genetic Resources certificated by Ministry of Agriculture, College of Fisheries and Life Science, Shanghai Ocean University, Shanghai 201306, PR Chinab E-Institute of Shanghai Universities, Shanghai Ocean University, Shanghai 201306, PR China

Abbreviations: qRT-PCR, quantitative real-time PCR;cDNA Ends; ISH, in situ hybridization; ORF, open readCRD, carbohydrate recognition domain.⁎ Corresponding author at: College of Fisheries and Life S

sity, 999 Hucheng Huan Road, Shanghai 201306, PR China.E-mail address: jlli2009@126.com (J.-L. Li).

0378-1119/$ – see front matter © 2013 Elsevier B.V. Alhttp://dx.doi.org/10.1016/j.gene.2013.05.029

a b s t r a c t

a r t i c l e i n f o

Article history:Accepted 15 May 2013Available online 31 May 2013

Keywords:PerlucinC-type lectinShell healingPearl formationHyriopsis cumingii

Perlucin is an important functional protein that regulates shell and pearl formation. In this study, we clonedthe perlucin gene from the freshwater pearl mussel Hyriopsis cumingii, designated as Hcperlucin. Thefull-length cDNA transcribed from the Hcperlucin gene was 1460 bp long, encoding a putative signal peptideof 20 amino acids and a mature protein of 141 amino acids. The mature Hcperlucin peptide contained six con-served cysteine residues and a carbohydrate recognition domain, similar to other members of the C-type lec-tin families. In addition, a “QPS” and an invariant “WND” motif near the C-terminal region were also found,which are extremely important for polysaccharide recognition and calcium binding of lectins. The mRNA ofHcperlucin was constitutively expressed in all tested H. cumingii tissues, with the highest expression levelsobserved in the mantle, adductor, gill and hemocytes. In situ hybridization was used to detect the presenceof Hcperlucin mRNA in the mantle, and the result showed that the mRNA was specifically expressed in theepithelial cells of the dorsal mantle pallial, an area known to express genes involved in the biosynthesis ofthe nacreous layer of the shell. The significant Hcperlucin mRNA expression was detected on day 14 postshell damage and implantation, suggesting that the Hcperlucin might be an important gene in shell nacreouslayer and pearl formation. The change of perlucin expression in pearl sac also confirmed that the mantletransplantation results in a new expression pattern of perlucin genes in pearl sac cells that are required forpearl biomineralization. These findings could help better understanding the function of perlucin in theshell and pearl formation.

© 2013 Elsevier B.V. All rights reserved.

1. Introduction

The shells and pearls of bivalves are a biomineralization product ofCaCO3 crystals, matrix proteins and other biopolymers (Mann, 2001;Wilbur, 1972). The shell of the pearl oyster consists of theperiostracum and two mineralized layers: the inner nacreous layerand the outer prismatic layer. Both layers are composed of calciumcarbonate and organic matrices but have different crystal polymor-phisms and microstructures that are controlled by the matrix pro-teins secreted from the epithelial cells of the mantle. Although theorganic matrix proteins constitute less than 5% by weight of thebiomineralized composite material, they do not only participate inthe construction of the organic nacre framework but also controlthe nucleation and growth of crystals and determine the polymorphspecificity of calcium carbonate as well (Marin et al., 2008; Wilt et

RACE, Rapid Amplification ofing frame; DIG, digoxigenin;

cience, Shanghai Ocean Univer-Tel./fax: +86 21 61900401.

l rights reserved.

al., 2003). Therefore, matrix protein is an important component forunderstanding the shell and pearl formation.

The triangle pearl mussel (Hyriopsis cumingii) is an importantfreshwater pearl mussel in China (Liu et al., 1979). The pearls formedinside these mussels have good qualities including color, cleannessand shape. Since the middle of the 1980s, owing to the success of ar-tificial propagation and the development of new technologies in in-ducing pearl production, thousands of farms have been establishedto culture H. cumingii in Zhejiang, Jiangsu, Jiangxi, Hubei, Hu'nanand Anhui provinces. In the world market, China produces 95% ofthe freshwater pearls, of which more than 70% are produced fromH. cumingii (Li et al., 2005; Wang et al., 2007). However, due to thelow quality of the pearl, the Chinese H. cumingii pearl output valueis only 10% of the total world pearl value (Li and Li, 2009). To improvethe pearl quality, it is necessary to explore the function of the matrixproteins in freshwater mussel shell and pearl formation.

Molecular biology techniques have been successfully applied to shellmatrix proteins to isolate and sequence the cDNAs of matrix proteinencoding genes, and a number of matrix proteins have been identifiedfrom the molluscan shell. For example, Aspein (Tsukamoto et al.,2004), Prismalin-14 (Suzuki and Nagasawa, 2007) and MSI31 (Sudoet al., 1997) occur in the prismatic layer; N16 (Samata et al., 1999),

211J.-Y. Lin et al. / Gene 526 (2013) 210–216

N19 (Yano et al., 2007) and MSI60 (Sudo et al., 1997) are found in thenacreous layer; and Nacrein (Miyamoto et al., 1996), MSI7 (Zhang etal., 2003) and EFCBP (Huang et al., 2007) occur in both layers. Manystudies have only focused on the marine bivalves (Wang et al., 2008;Yano et al., 2007), with few studies examining the freshwater bivalves.In our lab, we obtained one million high quality Expressed Sequencetags (ESTs) from the mantle and pearl sac from H. cumingii and identi-fied a perlucin gene fraction (Bai et al., 2013). Perlucin, a typicalC-lectin protein, is a carbohydrate-bindingmember of this large proteinfamily. This protein was originally isolated from the nacre of abaloneinner shell and was proved to promote calcium carbonate precipitationunder ambient conditions to nucleate the calcium carbonate crystalliza-tion and tomodify themorphology of calciumcarbonate crystals. There-fore, it is not only a constitutive protein of the nacre but also animportant functional molecule that regulates shell and pearl formation(Blank et al., 2003; Weiss et al., 2000).

To investigate the potential role of perlucin in freshwater musselshell and pearl formation, we cloned and reported for the first timethe complete perlucin transcript sequence from the freshwater mus-sel H. cumingii (designated as Hcperlucin). In the current study, weinvestigated the expression pattern of the Hcperlucin gene in varioustissues and mantle layers by quantitative real-time PCR (qRT-PCR)and in situ hybridization (ISH), respectively. In addition, we analyzedthe Hcperlucin gene expression during shell healing and early pearlformation. The cloning of the Hcperlucin gene could help us to under-stand shell and pearl formation in freshwater mussels.

2. Materials and methods

2.1. Ethics statement

All handling of mussels was conducted in accordance with theguidelines on the care and use of animals for scientific purposes setup by the Institutional Animal Care and Use Committee (IACUC) ofShanghai Ocean University, Shanghai, China.

2.2. Materials

Adult H. cumingii were obtained from the Weiwang pearl farm ofJinhua City, Zhejiang Province, China. A variety of tissues, includingmantle, gill, adductor muscle, liver, kidney, foot, hemocytes, gonadand intestine were collected and then immediately frozen in liquidnitrogen and stored at −80 °C until total RNA extraction.

2.3. Molecular cloning and sequencing

Total RNA was extracted from various tissues using RNAiso Plus re-agent (TaKaRa, Japan) and stored at −80 °C. The RNA quality and con-centration were determined using a Nanodrop 2000c (ThermoScientific, USA). RNA with a 260/280 ratio between 1.90 and 2.10 anda 260/230 ratio between 2.00 and 2.50 was considered satisfactory andused in this study. A 522 bp fragment of the perlucin gene was obtainedfrom a de novo transcriptomic library of H. cumingii (Bai et al., 2013) andused to clone the full-length cDNA of perlucin by the Rapid Amplificationof cDNA Ends (RACE)method using a gene-specific primer (Table 1). The5′-RACE was performed using the SMART RACE cDNA Amplification Kit

Table 1Primers used in the present study.

Gene Sense (5′-3′)

Perlucin GCAGGTGGAACGATTTAG

TGGATTTCTCATTGATATTCGTCCTGGTCGGAAGCCAAGGAATA

EF-1α GGAACTTCCCAGGCAGACTGTGC

(Clontech, Palo Alto, CA, USA) and an Advantage 2 cDNA PolymeraseMix (Clontech) following the manufacturer's instructions. Single-stranded cDNA for the RACE reaction was prepared from total RNA.The total volume of a 5′-RACE reaction was 50.0 μl. The reaction mix-ture contained 1.0 μl cDNA template, 5.0 μl 10 × BD Advantage 2 PCRBuffer, 1.0 μl 10 × UPM (Universal Primer A Mix), 1.0 μl gene-specificprimer (10 μM), 1.0 μl 50 mM dNTP Mix, 0.5 μl 50 × BD Advantage 2Polymerase Mix and 40.5 μl PCR-grade water. The cycle condition in-cluded one initial denaturation cycle of 94 °C for 3 min followed by35 PCR cycles of 94 °C for 30 s, 62 °C for 30 s and 68 °C for 1 min anda final extension step at 68 °C for 10 min. The 3′-RACE reaction wasperformed in a 50.0 μl volume containing 10.0 μl of 5 × PCR buffer,36.5 μl PCR-grade water, 0. 5 μl TaKaRa Ex Taq® HS (TaKaRa, Japan),0.5 μl gene-specific primer (10 μM), 0.5 μl M13 Primer M4 (10 μM)and 2.0 μl cDNA template using the following cycling parameters:35 cycles of a 30 s cDNA denaturation step at 94 °C, a 30 s annealingstep at 50 °C and a 1 min extension step at 72 °C. The 72 °C elongationstep was carried out 10 min. The PCR product was ligated into thepGEM®-T easy vector (Promega, USA), transformed into competentEscherichia coli DH5α cells, plated onto the LB-agar Petri dishes and in-cubated overnight at 37 °C. Positive clones containing the insert withthe expected size were identified by colony PCR. Six of the positiveclones were picked up and sequenced on an ABI PRISM 3730 Automat-ed Sequencer using BigDye terminator v3.1 (Applied Biosystems, USA).

2.4. Bioinformatics analysis

Nucleotide and amino acid sequence similarity searches wereperformed using the BLAST program (http://blast.ncbi.nlm.nih.gov/Blast.cgi). The open reading frame (ORF) of Hcperlucin cDNA was de-termined using the ORF Finder (http://www.ncbi.nlm.nih.gov/projects/gorf/), and the presence of signal peptide was further ana-lyzed using SignalP 3.0 Server (http://www.cbs.dtu.dk/services/SignalP/). Protein sequence similarity searches were performed withthe BLAST program in GenBank (http://www.ncbi.nlm.nih.gov/). Pro-tein multiple alignment analyses were performed using the ClustalXprogram.

2.5. Gene expression analysis by qRT-PCR

Total RNA was extracted from the mantle, gill, adductor muscle,liver, kidney, foot, hemocytes, gonad and intestine. A total of 600 ngRNA from each sample was reverse-transcribed with an iScript™cDNA Synthesis Kit (Bio-Rad, USA) according to the protocol providedby the manufacturer. The kit included a genomic DNA eliminationcolumn. The synthesized cDNA samples were stored at −20 °C priorto use in qRT-PCR assays. Primers were designed using the Primer3.0 program (http://frodo.wi.mit.edu/) and are listed in Table 1. Melt-ing curve analysis was carried out to select the optimum primer pairs.qRT-PCR was performed using a Bio-Rad CFX96 Real-Time PCR Detec-tion System (Bio-Rad, USA) in a 20.0 μl reaction mix containing 2.0 μlcDNA sample, 10.0 μl 2 × iScript reaction mix (Bio-Rad, USA), 0.5 μlof each primer (10 μM) and 7.0 μl nuclease-free water. The qRT-PCRcycling conditions were as follows: 95 °C for 30 s, 40 cycles of 95 °Cfor 5 s and 60 °C for 30 s. Finally, a melting curve analysis was under-taken. EF-1αwas adopted as a housekeeping gene. A negative control

Antisense (5′-3′) Application

3′-RACEAGACCGTGCCTGACAGATAAAG 5′-RACETGTTTTGTCTATATTGCCTTCTG ISHCGAACCATCAACCCAGTAACG qRT-PCRTCAAAACGGGCCGCAGAGAAT qRT-PCR

Fig. 1. Full-length cDNA and predicted protein sequence of the Hyriopsis cumingiiperlucin gene. Nucleotides and amino acids are numbered to the right of the sequence.The initiator codon (ATG) is shown in the square and the stop codon (TGA) is indicatedby an asterisk. Amino acid residues of signal peptide are underlined.

212 J.-Y. Lin et al. / Gene 526 (2013) 210–216

without any cDNA template was included for qRT-PCR analysis. The se-quences of Hcperlucin cDNA specific primers and EF-1α primers used inqRT-PCR are shown in Table 1. The triplicate fluorescence intensities ofeach sample, asmeasured by the crossing-point (CT) values, the relativeexpression levels were calculated using equation 2−ΔΔCT normalizedwith EF-1α (ΔCT = CT of target minus CT of EF-1α, ΔΔCT = ΔCT ofchallenged sample minus ΔCT of calibrator sample). One-way ANOVAwith Dunnett's T3 test were performed using SPSS 17.0 to determinewhether there were any significant differences. Data from theqRT-PCR experiments were expressed as the means ± SE. Significancewas accepted at the level of p b 0.05.

2.6. Preparation of digoxigenin (DIG)-labeled RNA probe and ISH

The cDNA fragments corresponding to nucleotides 719-1153 ofHcperlucin transcript were amplified and sub-cloned into thepGEM®-T Easy vector (Promega, USA), and then the recombinant plas-mid was linearized either with SacII (Promega, USA) or with SpeI(Promega, USA).

Digoxigenin (DIG)-labeled oligonucleotides for antisense andsense probes were synthesized by in vitro transcription with a DIGRNA Labeling Kit (Roche Diagnostics, Germany) using the SP6 or T7RNA polymerases (TaKaRa, Japan), respectively. ISH was carried outas described previously (Qiu et al., 2008). The mantle tissues werefixed overnight in 10 mM phosphate buffer (pH 7.4) containing 4%(w/v) paraformaldehyde and then dehydrated in ethanol and embed-ded in paraffin. Mantle tissue sections (5 μm) were hybridized withDIG-labeled antisense or sense RNA probes after deparaffinized inethanol and rehydrated. The DIG was visualized using colorimetricsubstrates NBT/BCIP (Roche Diagnostics, Germany) according to themanufacturer's instructions, and slides were observed under a lightmicroscope (Olympus).

2.7. Shell healing experiment and gene expression analysis of Hcperlucinby qRT-PCR

For expression analysis during shell healing, a rectangular piece ofshell (2 mm × 5 mm) was damaged from the middle ventral shellmargin of each mussel without hurting the mantle tissue. The shellrecovered completely after 28 days. Ninety damaged mussels weredivided into ten groups randomly, each of which contained nine indi-viduals. Mantles next to the damaged shell part were collected aftertimed intervals: 0, 2, 6, 12, 24, 48, 96 h and 7, 14, 28 days post shelldamaging (3 independent pools per time point, 6 mussels per pool).Meanwhile, three mantles without shell damaging were collected asa control for each time point. All mantles were frozen in liquid nitro-gen and stored at−80 °C until total RNA extraction. RNA preparation,qRT-PCR and data analysis were performed as described above.

2.8. Gene expression of Hcperlucin in the pearl sac during the early stagesof pearl formation

For expression analysis in the pearl sac during the early stages ofpearl formation, young mussels were reproduced in May 2012 andreached a shell length of approximately 10 cm before they weregrafted in November 2012. Twelve saibos were grafted into the leftand right mantles of each host mussel. After grafting, host musselswere cultured in the same pond. Pearl sacs (saibos) were collectedafter timed intervals: 0, 2, 6, 12, 24, 48, 96 h and 7, 14, 21, 28 dayspost grafting (3 independent pools per time point, 6 pearl sacs × 3mussels per pool). The pearl on days 14, 21 and 28 was collected aswell. The pearl sacs at 0 timed were collected to serve as control. Alltissues were quickly frozen in liquid nitrogen and kept in a −80 °Cfreezer. RNA preparation, qRT-PCR and data analysis were performedas described above.

3. Results

3.1. Cloning and sequence analysis of Hcperlucin cDNA

Two fragments of 929 and 514 bp were amplified by 3′-RACE and5′-RACE, respectively. The complete sequence of the Hcperlucin cDNAconsisted of a 5′ terminal untranslated region (UTR) of 48 bp, an openreading fragment (ORF) of 486 bp and a 3′ UTR of 926 bp with an18 bp poly (A) tail (Fig. 1). The sequence of Hcperlucin showed 33%sequence identity to the Haliotis diversicolor, 32% identity to Haliotisdiscus discus and 30% identity to Crassostrea gigas sequences. TheHcperlucin sequence was deposited in GenBank under accession no.KC436008.

3.2. Analysis of the deduced amino acid sequence

The Hcperlucin cDNA encoded a polypeptide of 161 amino acidswith a putative signal peptide of 20 amino acid residues and a matureprotein of 141 amino acids. The predicted Hcperlucin gene has a mo-lecular mass of 18 kDa and an isoelectric point of 6.57. Analysis of itsamino acid composition indicated the existence of six conserved cys-teine residues and one single carbohydrate recognition domain(CRD). The multiple sequence alignment of the Hcperlucin with 7other known carbohydrate-binding specificity revealed that almostall of the residues implicated in the calcium mediated carbohydratebinding were conserved in perlucin (Drickamer, 1988; Weis andDrickamer, 1996). Perlucin sequences also displayed conservation ofthe Cys- and Trp-rich regions. The “QPD” and the invariant “WND”

Fig. 2. Alignment of the perlucin from Hyriopsis cumingii and some selected members of the C-type lectin superfamily with predicted amino acid sequences. Residues identical in atleast 4 sequences including the new sequence are shaded. Star label six highly conserved cysteine residues; dots signify five highly conserved residues involved incalcium-dependent carbohydrate binding in C-type lectin group proteins (Mann, Weiss et al., 2000). The sequence motifs “QPS” and “WND”, which are proposed to determinecarbohydrate-binding specificity of C-type lectins are indicated by rectangular boxes. All the referred sequences are from NCBI: Hyriopsis cumingii (GenBank accession number:KC436008), Crassostrea gigas (EKC39512.1), Haliotis discus discus (ABO26590.1), Haliotis diversicolor (AEQ16377.1), Haliotis laevigata (P82596.3), Pinctada fucata (ACO36046.1),Ruditapes philippinarum (ACU83213), and Argopecten irradians (ADL27440.1).

213J.-Y. Lin et al. / Gene 526 (2013) 210–216

motif near the C-terminal region (labeled by a rectangular symbol inFig. 2) are extremely important for polysaccharide recognition andthe calcium binding of lectins. The presence of the “QPD” motif sug-gests specific binding to galactose while the motif “EPN” is indicativefor mannose binding (Drickamer, 1988; Iobst and Drickamer, 1994;Weis and Drickamer, 1996). However, in the corresponding site,“QPD” was replaced by “QPS”, suggesting that the specific carbohy-drate binding of the perlucin from H. cumingii cannot be predictedby amino-acid sequence comparisons and homology modeling.

Fig. 3. Perlucin gene expression in different tissues of Hyriopsis cumingii. Data werepresented as means ± SEM (n = 3).

3.3. Gene expression analysis by qRT-PCR and ISH

The Hcperlucin mRNA expression in nine tissues of adultH. cumingii was examined by qRT-PCR analysis. The housekeepinggene EF-1α was adopted as a positive control. As shown in Fig. 3,the mRNA of Hcperlucin was highly expressed in the mantle, adduc-tor, gill and hemocytes. These tissues, which are the main calciumcompartments, actively participate in calcium metabolism and shellformation.

To elucidate the function of perlucin involved in shell formation,we examined mantle tissue by ISH for the expression of HcperlucinmRNA. We found that Hcperlucin mRNA was specifically expressedin the epithelial cells of the dorsal mantle pallial (Fig. 4A), an areaknown to express genes involved in the biosynthesis of the nacreouslayer of the shell. Hybridization with a sense probe was adopted as anegative control, and no signals were detected (Fig. 4B).

3.4. Gene expression analysis of perlucin during shell healing

The expression level of Hcperlucin initially increased after shelldamage and then decreased and reached its lowest level at 48thhour post damage. Subsequently, the mRNA expression of Hcperlucinagain increased to the highest level on day 14 post damage beforelowering to a basal level until complete shell healing on day 28(Fig. 5). These results suggest that the oyster Hcperlucin participatesin shell healing.

3.5. Gene expression of perlucin in the pearl sac during the early stages ofpearl formation

After implantation, a lamellar crystal layer was produced on day14, and the nacre increased in thickness with the generation of

Fig. 4. In situ hybridization analysis of perlucin gene expression in the mantle ofHyriopsis cumingii. (A) The sections of the oyster mantle hybridized with the senseRNA probes labeled with digoxigenin, which was adopted as a negative control.(B) The sections of the oyster mantle hybridized with the anti-sense RNA probes la-beled with digoxigenin. (C) The enlarged view of rectangular box of (A). Threeoverlapping pictures of the same section are taken. Strong hybridization signals ap-pear in the epithelial cells of the mantle pallial (arrows). OF, outer fold; MF, middlefold; IF, inner fold; VM, ventral mantle pallial; and DM, dorsal mantle pallial.

214 J.-Y. Lin et al. / Gene 526 (2013) 210–216

nacreous layer on day 21, and the formation of a small pearl on day 28(Fig. 6). Significant expression of perlucin was detected by the 12thhour and 14th day post implantation (Fig. 7). In the earlier stages,

Fig. 5. The relative expression level of perlucin in mantle after shell damage. Data werepresented as means ± SEM. One asterisk denotes a significant (p b 0.05 by Dunnett'sT3 test, n = 3) difference from the control group. Two asterisks denotes a highly sig-nificant (p b 0.01 by Dunnett's T3 test, n = 3) difference from the control group.

the expression level of perlucin was up-regulated in the 12th hour,and decreased by day 7 post implantation. However, the perlucingene again reached a maximum expression on the day 14 before de-clining on day 28 (Fig. 7). These findings indicated that theHcperlucin gene may participate in pearl formation.

4. Discussion

Perlucin, a C-type lectin protein, was originally isolated from thenacre of abalone inner shell and was proved to promote calcium car-bonate precipitation, to nucleate the calcium carbonate crystalliza-tion, and to modify the morphology of calcium carbonate crystals(Dix, 1973; Weiss et al., 2000). In the current study, we first reportedthe identification and analysis of the C-type lectin perlucin gene fromthe freshwater pearl mussel H. cumingii. The full-length HcperlucincDNA was 1460 bp, which encoded a protein of 161 amino acidswith a putative signal peptide of 20 amino acid residues. The exis-tence of a signal peptide indicated that perlucin is an extracellularprotein, but no glycosylation site was detected in the Hcperlucingene, which was present in the Haliotis laevigata perlucin sequence(Mann et al., 2000). Hcperlucin contained six highly conserved cyste-ines and a carbohydrate recognition domain (CRD). The lectin domaincontained six cysteines that may form the disulfide bonds observed inother members of this superfamily (Mann and Siedler, 1999). Theconservation of Cys- and Trp-rich regions indicates that Hcperlucinis involved in Ca2+-dependent carbohydrate recognition of C-typelectins. The “QPD” and “EPN”motifs in perlucin suggested a specificityfor galactose and mannose (Drickamer, 1988; Iobst and Drickamer,1994; Weis and Drickamer, 1996), respectively. Solid phase assaysin H. laevigata showed a divalent metal ion-dependent binding ofperlucin to (neo) glycoproteins containing D-galactose or D-man-nose/D-glucose (Mann et al., 2000). Crystallographic analysis of sever-al C-type lectins indicted a compact globular structure (Weis et al.,1991) and the presence of Ca2+-binding sites for this family of lectins(Nielsen et al., 1997). These observations suggest that perlucin is afunctional C-type lectin with broad carbohydrate-binding specificityand may have an important role in shell and pearl formation. A de-tailed analysis on CRDs indicated that almost all of the residues impli-cated in calcium mediated carbohydrate binding are conserved inperlucin (Fig. 3). The similarities of the domain structures detectedbetween the bivalve and gastropod perlucins correlate with previousfindings on putatively conserved shell protein domains (Mann et al.,2012). Gastropod and bivalve nacres are the result of convergent evo-lution, and molecular mechanisms that guide the deposition of thevariants of nacre and its derivatives across the Mollusca are funda-mentally different (Jackson et al., 2010; Marie et al., 2011). Therefore,perlucin domain structure may be inherited from the last commonancestor of bivalves and gastropods or may result from independentrecruitments events.

Shell mineralization is a complicated process of calcium metabo-lism. Three main calcium compartments, namely the gill, hemolymphand mantle, are involved in the biomineralization process (Rousseauet al., 2003). During the process of biomineralization, calcium de-mand is very high for the formation of calcium carbonate precipitates.First, calcium ions are absorbed from the external environment viathe gill. Then, they are transported by the hemolymph and hemo-cytes. Finally, they cross the mantle tissues and precipitate on theorganic matrix surfaces that were pre-constructed by the mantlecells (Addadi et al., 2006; Mount et al., 2004; Rousseau et al., 2003).H. cumingii, like other bivalve mollusks, has an open circulatory sys-tem that circulates hemolymph through a number of cavities andsinuses. Meanwhile, the adductor is a key tissue for hemolymph trans-port. Hence, Hcperlucin was mainly expressed in the gill, adductor, he-mocytes andmantle, which implies that it might play an important rolein the shell formation process. The shell of the pearl oyster consists ofthe periostracum, the prismatic layer and the nacreous layer. The

Fig. 6. The photos show the implantation procedure and the pearls produced on days 14, 21 and 28.

215J.-Y. Lin et al. / Gene 526 (2013) 210–216

mantle, as the site of shell synthesis, is the most important tissueinvolved in biomineralization. Previous studies have revealed that dif-ferent regions of the mantle are responsible for the expression and se-cretion of the matrix proteins and act as regulators that are involvedin the formation of the three shell layers. Tyrosinases from the outer ep-ithelial cells of the middle fold are responsible for the periostracum for-mation (Zhang et al., 2006a); MSI31, Aspein, Prismalin-14 and KRMPfrom the outer and inner epithelial cells of the outer fold construct theprismatic layer (Sudo et al., 1997; Suzuki et al., 2004; Tsukamoto etal., 2004; Zhang et al., 2006b); and MSI60 expressed in the epithelialcells of the mantle pallial construct the nacreous layer (Sudo et al.,1997). In our study, strong hybridization signals appeared in the epithe-lial cells of the dorsal mantle pallial in an ISH experiment in the mantle,suggesting that Hcperlucin could be related to the formation of nacre-ous layers. The whole calcium transport and nacre formation processare strictly controlled by the regulation networks of these tissues andcells to synchronize the mollusk development and shell growth and torespond rapidly to shell damage (Rousseau et al., 2003).

To elucidate the role of Hcperlucin in shell formation, weperformed a shell damage experiment. Generally, the shell of a mus-sel has a certain growth rate; however, when shell damage occurs,mussel shell healing is induced, which accelerates the shell growthprocess. Correspondingly, the shell growth-related factors are in-creased and enhanced. Shell formation is a complex biomineraliza-tion process. After shell damage, the retraction and positioning ofthe mantle around the shell defect is in agreement with observation

Fig. 7. The relative expression level of perlucin in the pearl sac during the early stagesof pearl formation. Data were presented as means ± SEM. One asterisk denotes a sig-nificant (p b 0.05 by Dunnett's T3 test, n = 3) difference from the control group. Twoasterisks denotes a highly significant (p b 0.01 by Dunnett's T3 test, n = 3) differencefrom the control group.

from shell damage of Anodonta (Beedham, 1965). The formation of abrown non-mineralized layer represents the first step of the repair pro-cess, and the rapidity with which this layer forms differs according tothe species and habitat. In the freshwater bivalve Anodonta grandis(Saleuddin, 1967), the formation of the organic membrane takes placeafter 1 day, and the shell is fully re-mineralized after 6 months. In ourcase, the brown organic patch appeared 3 days after lesion, and themineralization process was completed after the 28th day. Therefore,the Hcperlucin expression level of the mussel in the mantle increasedin the 96th hour and reached a peak on the 14th day following shelldamage indicating that Hcperlucin participates in shell formation.

Nacre is the main component of the pearl. Pearl components aresecreted by the pearl sac, which is homologous to the cellular compo-sition of the outer epithelium of the mollusk mantle pallial zone (Dix,1973; Machii, 1968). Freshwater cultured pearl production is a com-plex biological process involving the implantation of a mantle graftfrom a donor pearl mussel into the mantle of a host mussel. Recently,microsatellite genetic markers and transcriptomes were used to ana-lyze the genotyping and biomineralization related genes within thepearl sac, respectively. This analysis demonstrated that the donormantle tissue is primarily responsible for the expression of biominer-alization genes in the pearl sac (Arnaud-Haond et al., 2007; McGintyet al., 2012). In this study, we collected the saibos, which did notfuse completely with host mussels cells within 28 days. The lamellarcrystal layer was produced on the 14th day, and with the growth ofthe nacre, a small pearl was produced on day 28. The highest expres-sion of Hcperlucin occurred on the 14th day suggesting that theHcperlucin gene participates in pearl formation. Meanwhile, thechange in perlucin expression in the pearl sac concomitant with thepearl formation also confirmed a new expression pattern of perlucingene in pearl sac cells.

Perlucin is a member of the C-type lectin family, which partici-pates in the immune response to different stressors, and defendsagainst invading pathogens (De Zoysa et al., 2011; Jia et al., 2011).Therefore, in both post shell damage and saibo implantation, theperlucin gene expression was quickly up-regulated within 96 h ofshell damage or implantation. Previous studies revealed that perlucinis secreted to regulate nacre entombment of the offending antigen asa means of inactivating non-self materials when pearl mussel cannoteject the invading parasites or particles. With the growth of nacre onthe offending particle, a beautiful pearl is produced (Wang et al.,2008). Further experimentation is required to understand the exactphysiological function of perlucin in the H. cumingii nacre formation.

5. Conclusion

We first acquired the full-length perlucin cDNA fromH. cumingii andanalyzed the mRNA expression pattern of Hcperlucin in different tis-sues, both during shell healing and pearl formation. Hcperlucinwas pri-marily expressed in the gill, adductor, hemocytes and mantle tissues,and specifically expressed in the epithelial cells of the dorsal mantle

216 J.-Y. Lin et al. / Gene 526 (2013) 210–216

pallial. This indicated that Hcperlucin is a nacre matrix protein, and re-sponsible for the shell nacre formation. The expression of Hcperlucinalso responded to the shell healing and pearl formation process. These re-sults demonstrate that perlucin has a positive effect on shell nacre andpearl formation. The change in perlucin expression in the pearl sac alsoconfirmed that mantle transplantation results in a new expression pat-tern of perlucin genes in pearl sac cells during pearl biomineralization.Future studies about structure-function relationship are required to un-derstand the physiological function of perlucin in nacre formation.

Acknowledgments

We thank the staff of the Weiwang Pearl Farm of Jinhua for theirhelp in the culturing and collecting of freshwater pearl mussels. Wealso thank Dr. Xiaojun Liu and Dr. Narayan Prasad Pandit especiallyfor their help with the manuscript preparation. This work was sup-ported by the National Technology Support Programme of China(2012BAD26B04) and Shanghai Universities Knowledge Service Plat-form (ZF1206).

References

Addadi, L., Joester, D., Nudelman, F., Weiner, S., 2006. Mollusk shell formation: a source ofnew concepts for understanding biomineralization processes. Chemistry 12, 980–987.

Arnaud-Haond, S., Goyard, E., Vonau, V., Herbaut, C., Prou, J., Saulnier, D., 2007. Pearlformation: persistence of the graft during the entire process of biomineralization.Mar. Biotechnol. 9, 113–116.

Bai, Z.Y., Zheng, H.F., Lin, J.Y., Wang, G.L., Li, J.L., 2013. Comparative analysis of thetranscriptome in tissues secreting purple and white nacre in the pearl mussel Hyriopsiscumingii. PLoS One 8, e53617.

Beedham, G.E., 1965. Repair of the Shell in Species of Anodonta. P. Zool. Soc. Lond. WileyOnline Library 107–123.

Blank, S., et al., 2003. The nacre protein perlucin nucleates growth of calcium carbonatecrystals. J. Microsc. 212, 280–291.

De Zoysa, M., et al., 2011. Microarray analysis of gene expression in disk abaloneHaliotis discus discus after bacterial challenge. Fish Shellfish Immunol. 30, 661–673.

Dix, T.G., 1973. Histology of the mantle and pearl sac of the pearl oyster Pinctadamaxima (Lamellibranchia). J. Malacol. Soc. Aust. 2, 365–375.

Drickamer, K., 1988. Two distinct classes of carbohydrate-recognition domains in ani-mal lectins. J. Biol. Chem. 263, 9557–9560.

Huang, J., Zhang, C., Ma, Z.J., Xie, L.P., Zhang, R.Q., 2007. A novel extracellular EF-handprotein involved in the shell formation of pearl oyster. BBA Gen. Subjects 1770,1037–1044.

Iobst, S.T., Drickamer, K., 1994. Binding of sugar ligands to Ca-2+-dependent animallectins: II. Generation of high-affinity galactose binding by site-directed mutagen-esis. J. Biol. Chem. 269, 15512–15519.

Jackson, D.J., et al., 2010. Parallel evolution of nacre building gene sets in molluscs. Mol.Biol. Evol. 27, 591–608.

Jia, X.W., Zou, Z.H., Wang, G.D., Wang, S.H., Wang, Y.L., Zhang, Z.P., 2011. Gene expres-sion profiling in respond to TBT exposure in small abalone Haliotis diversicolor. FishShellfish Immunol. 31, 557–563.

Li, J.L., Li, Y.S., 2009. Aquaculture in China—freshwater pearl culture. World Aquacult. 40, 60.Li, J.L., Qian, R.H., Bao, B.L., Wang, G.L., Qi, N.D., 2005. RAPD analysis on genetic diversity

among the stocks of Hyriopsis cumingii from the five large lakes of China. J. Shang-hai Fish. Univ. 14, 1–5.

Liu, Y.Y., Zhang, W.Z., Wang, Y.X., 1979. Freshwater Mollusk Economic Fauna of China. Sci-ence Press, Beijing.

Machii, A., 1968. Histological studies on the pearl sac formation. Bull. Pearl Res. Lab. 13,1489–1539.

Mann, K., Edsinger-Gonzales, E., Mann, M., 2012. In-depth proteomic analysis of a mol-lusc shell: acid-soluble and acid-insoluble matrix of the limpet Lottia gigantea. Pro-teome Sci. 10.

Mann, K., Siedler, F., 1999. The amino acid sequence of ovocleidin 17, a major protein ofthe avian eggshell calcified layer. Biochem. Mol. Biol. Int. 47, 997–1007.

Mann, K., Weiss, I.M., Andre, S., Gabius, H.-J., Fritz, M., 2000. The amino-acid sequenceof the abalone (Haliotis laevigata) nacre protein perlucin: detection of a functionalC-type lectin domain with galactose/mannose specificity. Eur. J. Biochem. 267,5257–5264.

Mann, S., 2001. Biomineralization: Principles and Concepts in Bioinorganic MaterialsChemistry. Oxford University Press, NY, Oxford.

Marie, B., Le Roy, N., Zanella-Cleon, I., Becchi, M., Marin, F., 2011. Molecular evolution ofmollusc shell proteins: insights from proteomic analysis of the edible musselMytilus. J. Mol. Evol. 72, 531–546.

Marin, F., Luquet, G., Marie, B., Medakovic, D., 2008. Molluscan shell proteins: primarystructure, origin, and evolution. Curr. Top. Dev. Biol. 209–276.

McGinty, E.L., Zenger, K.R., Jones, D.B., Jerry, D.R., 2012. Transcriptome analysis ofbiomineralisation-related genes within the pearl sac: host and donor oyster contri-bution. Mar. Genomics 5, 27–33.

Miyamoto, H., Miyashita, T., Okushima, M., Nakano, S., Morita, T., Matsushiro, A., 1996.A carbonic anhydrase from the nacreous layer in oyster pearls. Proc. Natl. Acad. Sci.U. S. A. 93, 9657–9660.

Mount, A.S., Wheeler, A.P., Paradkar, R.P., Snider, D., 2004. Hemocyte-mediated shellmineralization in the eastern oyster. Science 304, 297–300.

Nielsen, B.B., et al., 1997. Crystal structure of tetranectin, a trimeric plasminogen-binding protein with an alpha-helical coiled coil. FEBS Lett. 412, 388–396.

Qiu, G.F., Weber, G.M., Rexroad, C.E., Yao, J.B., 2008. Identification of RtGST-1, a novelgerm cell-specific mRNA-like transcript predominantly expressed in earlyprevitellogenic oocytes in rainbow trout (Oncorhynchus mykiss). Mol. Reprod.Dev. 75, 723–730.

Rousseau, M., Plouguerne, E., Wan, G., Wan, R., Lopez, E., Fouchereau-Peron, M.,2003. Biomineralisation markers during a phase of active growth in Pinctadamargaritifera. Comp. Biochem. Physiol. A 135, 271–278.

Saleuddin, A.S.M., 1967. The histochemistry of the mantle during the early stage ofshell repair. J. Molluscan Stud. 37, 371–380.

Samata, T., Hayashi, N., Kono, M., Hasegawa, K., Horita, C., Akera, S., 1999. A new matrixprotein family related to the nacreous layer formation of Pinctada fucata. FEBS Lett.462, 225–229.

Sudo, S., et al., 1997. Structures of mollusc shell framework proteins. Nature 387,563–564.

Suzuki, M., et al., 2004. Characterization of Prismalin-14, a novel matrix protein fromthe prismatic layer of the Japanese pearl oyster (Pinctada fucata). Biochem. J. 382,205–213.

Suzuki, M., Nagasawa, H., 2007. The structure-function relationship analysis ofPrismalin-14 from the prismatic layer of the Japanese pearl oyster, Pinctada fucata.FEBS J. 274, 5158–5166.

Tsukamoto, D., Sarashina, I., Endo, K., 2004. Structure and expression of an unusuallyacidic matrix protein of pearl oyster shells. Biochem. Biophys. Res. Commun. 320,1175–1180.

Wang, G.L., Yuan, Y.M., Li, J.L., 2007. SSR analysis of genetic diversity and phylogeneticrelationships among different populations of Hyriopsis cumingii from the five lakesof China. J. Fish. China 31, 152–158.

Wang, N., Lee, Y.H., Lee, J., 2008. Recombinant perlucin nucleates the growth of cal-cium carbonate crystals: Molecular cloning and characterization of perlucinfrom disk abalone, Haliotis discus discus. Comp. Biochem. Physiol. B 149,354–361.

Weis, W.I., Drickamer, K., 1996. Structural basis of lectin-carbohydrate recognition.Annu. Rev. Biochem. 65, 441–473.

Weis, W.I., Kahn, R., Fourme, R., Drickamer, K., Hendrickson, W.A., 1991. Structure ofthe calcium-dependent lectin domain from a rat mannose-binding protein deter-mined by MAD phasing. Science (New York, N.Y.) 254, 1608–1615.

Weiss, I.M., Kaufmann, S., Mann, K., Fritz, M., 2000. Purification and characterization ofperlucin and perlustrin, two new proteins from the shell of the mollusc Haliotislaevigata. Biochem. Biophys. Res. Commun. 267, 17–21.

Wilbur, K.M., 1972. Shell formation in mollusks. Chem. Zool. 7, 103–145.Wilt, F.H., Killian, C.E., Livingston, B.T., 2003. Development of calcareous skeletal ele-

ments in invertebrates. Differentiation 71, 237–250.Yano, M., Nagai, K., Morimoto, K., Miyamoto, H., 2007. A novel nacre protein N19 in the

pearl oyster Pinctada fucata. Biochem. Biophys. Res. Commun. 362, 158–163.Zhang, C., Xie, L.P., Huang, J., Chen, L., Zhang, R.Q., 2006a. A novel putative tyrosinase

involved in periostracum formation from the pearl oyster (Pinctada fucata).Biochem. Biophys. Res. Commun. 342, 632–639.

Zhang, C., Xie, L.P., Huang, J., Liu, X.L., Zhang, R.Q., 2006b. A novel matrix protein familyparticipating in the prismatic layer framework formation of pearl oyster, Pinctadafucata. Biochem. Biophys. Res. Commun. 344, 735–740.

Zhang, Y., et al., 2003. A novel matrix protein participating in the nacre framework for-mation of pearl oyster, Pinctada fucata. Comp. Biochem. Physiol. B 135, 565–573.