Aquatic Toxicology 75 (2005) 374–379

Putative midkine family protein up-regulation inPatellacaerulea (Mollusca, Gastropoda) exposed to sublethal

concentrations of cadmium

Silvana Vanuccia,∗,1, Daniela Minerdib,1,2, Kenji Kadomatsuc,Alessio Mengonid, Marco Bazzicalupod

a Department of Animal Biology and Marine Ecology, University of Messina, Salita Sperone 31, 98166 S Agata, Messina, Italyb Department of Animal Biology and Genetics, University of Florence, via Romana 19, 50125 Firenze, Italy

c Department of Biochemistry, University of Nagoya Graduate School of Medicine, 65 Tsurumai-cho, Showa-ku, Nagoya 466-8550, Japand Department of Animal Biology and Genetics, University of Florence, via Romana 19, 50125 Firenze, Italy

Received 9 February 2005; received in revised form 18 August 2005; accepted 30 August 2005

Abstract

A cDNA sequence of a putative midkine (MK) family protein was identified and characterised in the molluscPatella caerulea.The midkine family consists of two members, midkine and pleiotrophin (PTN), and it is one of the recently discovered cytokines.Our results show that this putative midkine protein is up-regulated in specimens ofP. caerulea exposed to sublethal cadmium

timereover,

lus gills)cancel

vetsto

l-

vi-

concentrations (i.e. 0.5 and 1 mg l−1 Cd) over a 10-day exposure period. Semiquantitative RT-PCR and quantitative RealRT-PCR estimations indicate elevated expression of midkine mRNA in exposed specimens compared to controls. MoRT-PCR Real time values were higher in the viscera (here defined as the part of the soft tissue including digestive gland pthan in the foot (i.e. foot plus head plus heart) of the limpets. At present, information on the functional signalling signifiof the midkine family proteins suggests that the up-regulation ofP. caerulea putative midkine family protein is a distress signalikely with informative value on health status of the organism and with potential prognostic capability.© 2005 Elsevier B.V. All rights reserved.

Keywords: Patella caerulea; Mollusc; Midkine; Pleiotrophin; Distress signal; Cadmium; RT-PCR Real time

∗ Corresponding author. Tel.: +39 347 5714315;fax: +39 090 393409.

E-mail address: silvana.vanucci@unime.it (S. Vanucci).1 These authors contributed equally to this work.2 Present address: Di.Va.P.R.A.-Plant Pathology, University of

Torino, Via L. da Vinci 44, 10095 Grugliasco, Italy.

In the last decade, coastal environments habeen subjected to increasing monitoring and efforhave been spent to study distress signals in orderdevelop biomarkers with warning prognostic capabiity (Moore, 2002). Cadmium is one of the most toxicand widespread heavy metals found in the marine enronment (Romeo et al., 1995); it promotes cellular

0166-445X/$ – see front matter © 2005 Elsevier B.V. All rights reserved.doi:10.1016/j.aquatox.2005.08.014

S. Vanucci et al. / Aquatic Toxicology 75 (2005) 374–379 375

oxidative stress and induction of anti-oxidant systemsin many marine molluscs (e.g.Leung and Furness,1999; Geret et al., 2002). Moreover, cadmium is recog-nised to be carcinogenic in mammals (IARC, 1999),exerting pronounced co-mutagenic effects. At present,these effects are explained not only by increasing thelevel of DNA strand breakage induced by oxidativestress, but also by impairing cell recovery (Hartwingand Schwerdtle, 2002; Fatur et al., 2003), and delayingthe onset of apoptosis (Waalkes et al., 2000). Recently,it has been showed that in the musselMytilus edulis cad-mium enhances genotoxicity by mechanisms similar tothose reported for mammals (Pruski and Dixon, 2002).

The midkine (MK) is a recently discovered family ofcytokines (see reviewsDeuel et al., 2002; Kadomatsuand Muramatsu, 2004) and consists of only two mem-bers, namely heparin-binding growth factors MK andpleiotrophin (PTN). Information on PTN and MK func-tion is mostly from mammals; MK and PTN sharereceptors and show similar biological activities thatinclude fibrinolytic, anti-apoptotic, mitogenic, trans-forming, angiogenic, chemotactic, and protooncogenicones. In adults, a characteristic property of MK/PTNgenes is a striking induction of gene expression inlimited emergency conditions, such as wound repairfollowing injuries. In stark contrast, MK and PTN arestrongly expressed in neurodegenerative diseases andin malignant tumours. The blood MK level in humancarcinomas is correlated with prognostic factors; thus,it is as a good marker for evaluating the progress ofc

ter-i ilypt wl-e tifi-c TNi portd sionf imeea stala a sente

ee alaG n

7 June 2003. To reduce size/age-related variability,limpets of similar wet flesh weight (1± 0.4 g with-out shell) were selected. Measurements taken in thesurface water layer in front of the breakwater gave anaverage salinity of 35 g l−1 and a temperature of 20◦C.Immediately after collection, limpets were transportedto the laboratory (Florence) and acclimated for 2 daysbefore being exposed to cadmium (Cd). Three hold-ing tanks, containing 20 specimens each, with eachspecimen singly placed in an open Petridish (diam-eter 6 cm, height 1 cm), were prepared and suppliedwith continuously aerated artificial seawater (35 g l−1

Reef CrystalsTM salt, Aquarium Systems). The threetanks were located in a temperature-controlled room(20± 0.5◦C) with a natural daylight cycle. Exposureconcentrations of cadmium were chosen with referenceto sublethal concentrations based on 96-h acute toxicitytests carried out onP. caerulea in the same labora-tory (Parenti, 2003). The three groups of limpets wereexposed for 10 days to 0 (control), 0.50 and 1.0 mg l−1

Cd, respectively. Cadmium concentrations were pre-pared using a 100 mg l−1 Cd stock solution obtainedby adding an appropriate quantity of CdCl2 (FlukaChemie®). Water was changed every second day toensure no build-up of toxic materials from animals,changing water quality. The animals were not fed forthe entire exposure period. Mortalities were recordeddaily and dead animals were discarded; at the end of the10th day mortalities were 4, 20 and 39% for 0 (control),0.5 and 1.0 mg l−1 Cd treated animals, respectively.F cad-m ells,a dis-s eadp ssuei is-c at−

es( dt u-t hegt tfw:5c neins( ex-p s

arcinomas (Ikematsu et al., 2003).In this work, we report on isolation and charac

sation of a cDNA sequence of a putative MK famrotein that is up-regulated inPatella caerulea exposed

o sublethal concentrations of cadmium. To our knodge, this is the first report regarding MK/PTN idenation in molluscs and on the involvement of MK/Pn response to heavy metal exposure. We also reata on the mRNA tissue/organ quantitative expres

eature in terms of both RT-PCR and RT-PCR Real tstimations. We chose the gastropod limpetP. caeruleas it is widely distributed over Mediterranean coareas, accumulates cadmium and is considered as

inel organism for the Mediterranean sea (Campanellat al., 2001).

Samples ofP. caerulea were collected from thxternal side of a breakwater at the Marina di Calera (42◦26′50′′N, 11◦26′00′′E, Tyrrhenian sea) o

-

ollowing exposure, the tissues of the control andium treated limpets were separated from the shnd for each specimen the whole soft body wasected into two parts: the foot (i.e. the foot plus hlus heart) and the viscera (i.e. the remain soft ti

ncluding digestive gland plus gills); foot (F) and vera (V) were frozen in liquid nitrogen and stored70◦C until analyses.Total RNA was extracted from 500 mg of tissu

viscera plus foot) ofP. caerulea specimens exposeo 1 mg l−1 cadmium using Trizol (Invitrogen) solion according to the manufacturer’s protocol. Teneration of cDNA was carried out by 3′RACE

echnique using a sense degenerated primer (M′-GTGTGGNAGCVVGTGC-3′) designed on cDNAonsensus sequences from gastropod metallothiothe original objective was focused on MT, with unected isolation of MK). SMART PCR cDNA wa

376 S. Vanucci et al. / Aquatic Toxicology 75 (2005) 374–379

obtained from 2.5�g of RNA following the manu-facturer’s instructions (SMART cDNA Synthesis Kit,Clontech). First strand synthesis was performed usingSuperScript III reverse transcriptase (Qiagen) accord-ing to manufacturer’s instructions. PCR was performedusing buffer (1.5 mM MgCl2), Taq DNA polymerase(Invitrogen, 2.5 U), primers (350 ng each), reverse tran-scriptase products (500 ng), dNTPs (2.5 mM of each)and sterilised water. cDNA amplification was per-formed following these parameters: 94◦C for 3 min(1 cycle), 94◦C for 1.5 min, 50◦C for 1.5 min, 72◦Cfor 1.5 min (30 cycles), 72◦C for 10 min (1 cycle).RACE–PCR product of about 500 bp was cloned inpGeM-T vector (Promega) and sequenced (GenBankaccession number AY678118).

Reverse transcription and PCR using specificprimers for Patella vulgata �2-tubulin cDNA (�-tubf: 5′-ACGTCGACAAGACCGACTTC-3′; �-tubr:5′-TGAAACCAGTTGGACACCAG-3′) and for P.caerulea putative MK cDNA (mdkf: 5′-AACAT-TAAGGGGAACCCAGG-3′; mdkr: 5′-TTGTTTTG-GTTTCGGTGTCA-3′) sequences were performed on2�g total RNA extracted from the viscera and fromthe foot ofP. caerulea specimens exposed to 0.5 and1 mg l−1Cd, and from control specimens. The house-keeping geneα2-tub was used as the internal referencegene. The analysis was carried out in a final vol-ume of 20�l containing 4�l of 5× buffer, 400�MdNTPs, 0.8�M of �-tubr and mdkr primers, 10 U ofRNase inhibitor (Amersham Biosciences), 100 U ofe nwsru o-c

h ai d).I nu-cS on-t ,0 ,6 s-c pa-rwf me

PCR analysis were�-tubf, �-tubr, mdkf and mdkr,already described above. A melting curve (55–95◦Cwith a heating rate of 0.5◦C for 10 s and a continuousfluorescence measurement) was recorded at the end ofevery run to assess amplification product specificity.The ‘Comparative threshold cycle (Ct)’ method wasused to calculate relative MK expression levels withthe�2-tubulin RNA as a reference (Rasmussen, 2001).

A 3′-end complete cDNA sequence of 538 bp wasobtained and analysed for the presence of open read-ing frames (ORFs). The analysis revealed the presenceof one not complete ORF of 417 bp, a 3′-untranslatedregion of 116 bp and a polyadenylation signal located70 bp downstream of the stop codon (TGA). Thededuced protein sequence contained 139 amino acidswith a predicted molecular weight of 15 kDa and anestimated pI of 10.52 (the nucleotide sequence has theGenBank accession number AY678118.). BlastX anal-ysis showed lowE-values with MK family sequencesbelonging to vertebrates and invertebrates. TheP.caerulea putative protein exhibited the highest degreeof sequence similarity to theAnopheles gambiaeMK/PTN protein (23.7% sequence identity, 47.5%sequence similarity). TheP. caerulea putative MK C-terminal half domain is conserved: 6 out of the 10conserved cysteine residues in vertebrates are presentin the limpet sequence (Fig. 1). Moreover, the fourresidues that characterise the heparin binding clusterI of the C-terminal half domain are also conserved,although the third amino acid arginine is converted tol C-t ise por-t ies( pe-c atea ervec inpr

t theP seds wash sn imalsw

c ,

nzyme and 2�g of total RNA. Reverse transcriptioas allowed to proceed for 60 min at 50◦C and wastopped by heating at 70◦C for 15 min. Amplificationeactions (45 s at 94◦C, 45 s at 62◦C, 45 s at 72◦C),sing primers�-tubf and mdkf, were allowed to preed for 25 cycles.

RT-PCR real time assay was carried out witCycler iQTM Real-Time PCR apparatus (Bio-Randividual reactions were assembled with oligoleotide primers (0.15 mM each), 12.5�l of 2X iQTM

YBR Green Supermix (Bio-Rad Laboratories; caining 100 mM KCl, 40 mM Tris–HCl, pH 8.4.4 mM dNTPs, 50 U�l−1 iTaq DNA polymerasemM MgCl2, 20 nM SYBR Green I, 20 nM fluoreein) plus an appropriate volume of each cDNA preation in a final volume of 25�l. The following programas run: 95◦C for 3 min (1 cycle), 94◦C for 20 s, 62◦C

or 20 s (45 cycles). The primers utilised for Real-Ti

ysine (Fig. 1). The domain corresponding to theerminal half domain of vertebrate MK and PTNvolutionally conserved, and this stresses the imance of its structure in retaining biological activite.g. neurite outgrowth, fibrinolysis). Cluster I is esially important for the recognition of heparin sulphs well as chondroitin sulphate proteoglycans, nell migration, and interaction with protein tyroshosphatase (Kadomatsu and Muramatsu, 2004andeferences therein).

Semiquantitative RT-PCR analyses showed tha. caerulea putative MK was overexpressed in expopecimens with respect to controls, and expressionigher in the viscera than in the foot (Fig. 2), whereao overexpression was observed in exposed anhen using�2-tubulin primers (data not shown).The RT-PCR Real time analysis (mdk correlation

oefficient: 0.996, equation:Y =−3.347X + 15.325

S. Vanucci et al. / Aquatic Toxicology 75 (2005) 374–379 377

Fig. 1. ClustalW alignment of the amino acid sequence deduced from thePatella caerulea putative midkine gene and from vertebrates andinvertebrates orthologous genes. The amino acids are indicated by single-letter code. Gaps were introduced for optimal alignment. Identicalresidues between two sequences are indicated by an asterisk (*); similar residues are indicated by one (low similarity) or two dots (highsimilarity). Conserved cysteins are indicated in bold, conserved residues of the heparin binding cluster I are marked with I. A:Danio rerio PTNsequence B60042; B:Xenopus tropicalis PTN sequence AAH61275.1; C:Mus musculus PTN sequence AAH02064.1, D:Mouse PTN sequenceI53001; E:Rattus norvegicus PTN sequence; F:Danio rerio midkine-related growth factor sequence NP571145; G:Bos taurus PTN sequenceNP 776380.1; H:Gallus gallus PTN sequence B60042; I:Patella caerulea putative MK/PTN sequence AY678118; L:Anopheles gambiaePTN/MK sequence XP315691.

378 S. Vanucci et al. / Aquatic Toxicology 75 (2005) 374–379

Fig. 2. Agarose gel of RT-PCR product (size: 146 bp) amplified usingspecific primers (mdkf/mdkr) forPatella caerulea putative midkinegene. Lanes 1 and 2, cDNA from the viscera (V) and from the foot (F)of one unexposed specimen; lanes 3 and 4, cDNA from V and F of thefirst 0.5 mg l−1 Cd-exposed specimen; lanes 5 and 6, cDNA from Vand F of the first 1 mg l−1 Cd-exposed specimen; lanes 7 and 8, cDNAfrom V and F of the second 0.5 mg l−1 Cd-exposed specimen; lanes9 and 10, cDNA from V and F of the second 1 mg l−1 Cd-exposed;lane 11, RT negative control; DNA marker (1 kbp plus, Invitrogen)is shown on the left.

PCR efficiency: 98.9%;α2-tub correlation coeffi-cient: 0.996, equation:Y =−3.276X + 16.665, PCRefficiency: 101.9%), performed on fourP. caeruleaspecimens exposed to cadmium (i.e. two specimens at0.5 and two specimens at 1 mg l−1 Cd, respectively),showed a dramatic MK over-expression in treated ani-mals with respect to controls;mdk transcript levels,normalised toα2-tub transcript levels and relative tounexposed controls, were also higher in the viscerathan in the foot (i.e. relative gene expressions in thefirst 0.5 mg l−1 Cd-exposed specimen were: 7900- and2340-fold in V and F, respectively; in the second0.5 mg l−1 Cd specimen: 9800- and 2730-fold, in Vand F, respectively; in the first 1 mg l−1 Cd specimen:25475- and 10603-fold, in V and F, respectively; in thesecond 1 mg l−1 Cd specimen: 18632- and 4534-fold,in V and F, respectively).

Our results indicate that cadmium induces over-expression of a MK family protein; a tissue/organ-dependent up-regulation seems also to emerge. Thehigher MK expression found in the viscera than in thefoot is consistent with observations that heavy metalsaccumulate mainly in the gills and digestive gland or inhepatopancreas in marine molluscs (Viarengo, 1989).The nature of the cytokine identified inP. caerulea has

intrinsically high informative value on the health sta-tus of the organism and warning prognostic capability.At present, we speculate that MK may be involved inrepairing cell injuries induced by cadmium oxidativestress and/or that MK may be activated by cadmiumrelated genotoxicity. This study is the basis for futureinvestigations that are needed to assess the expressionpattern of the MK-like protein and its use as a biomarkerof effect.

Acknowledgement

The authors thank Dr. Maria Ceccherini, Faculty ofAgriculture, Florence University, for facilities in usingCycler BIO-RAD instrument.

References

Campanella, L., Conti, M.E., Cubadda, F., Sucapane, C., 2001. Tracemetals in seagrass, algae and mollusc from an uncontaminatedarea in the Mediterranean. Environ. Pollut. 111, 117–126.

Deuel, T.F., Zhang, N., Yeh, H.-J., Silos-Santiago, I., Wang, Z.-Y.,2002. Pleiotrophin: a cytokine with diverse functions and a novelsignaling pathway. Arch. Biochem. Biophys. 397, 162–171.

Fatur, T., Lah, T.T., Filipic, M., 2003. Cadmium inhibits repair of UV-, methyl methanesulfonate- andN-methyl-nitrosourea-inducedDNA damage in Chinese hamster ovary cells. Mutat. Res. 529,109–116.

Geret, F., Serafin, A., Barreira, L., Bebianno, M.J., 2002. Effects ofida-,

H enicical

I ai, K.,atedan

I ns.ns,

K in in3.

L eind-

M e in

P ina’

cadmium on antioxidant enzyme activities and lipid peroxtion in the gill of the clamRuditapes decussates. Biomarkers 7242–256.

artwing, A., Schwerdtle, T., 2002. Interactions by carcinogmetal compounds with DNA repair processes: toxicologimplications. Toxicol. Lett. 127, 47–54.

kematsu, S., Nakagawara, A., Nakamura, Y., Sakuma, S., WakMuramatsu, T., Kadomatsu, K., 2003. Correlation of elevlevel of blood midkine with poor prognostic factors of humneuroblastomas. Br. J. Cancer 88, 1522–1526.

ARC, 1999. Overall Evaluation of Carcinogenicity to HumaIARC Monographs, vols. 1–73. IARC Scientific PublicatioLyon.

adomatsu, K., Muramatsu, T., 2004. Midkine and pleiotrophneural development and cancer. Cancer Lett. 204, 127–14

eung, K.M.Y., Furness, R.W., 1999. Induction of metallothionin DogwhelkNucella lapillus during and after exposure to camium. Ecotoxicol. Environ. Saf. 43, 156–164.

oore, M.N., 2002. Biocomplexity: the post-genome challengecotoxicology. Aquat. Toxicol. 59, 1–15.

arenti, A., 2003. Effetti letali e subletali del cadmio discioltoPatella caerulea (Molluschi, Gasteropodi): curve di mortalit

S. Vanucci et al. / Aquatic Toxicology 75 (2005) 374–379 379

e variazioni della frequenza cardiaca. Degree Thesis, FlorenceUniversity, p. 75.

Pruski, A.M., Dixon, D.R., 2002. Effects of cadmium on nuclearintegrity and DNA repair efficiency in the gills ofMytilus edulisL. Aquat. Toxicol. 57, 127–137.

Rasmussen, R., 2001. Quantification on the LightCycler instrument.In: Meuer, S., Wittwer, C., Nakagawara, K. (Eds.), Rapid CycleReal-Time PCR: Methods and Applications. Springer-VerlagPress, Heidelberg, Germany, pp. 21–34.

Romeo, M., Gnassia-Barelli, M., Lafaurie, M., 1995. Heavy met-als pollution in marine food chains. J. Eur. Hydrol. 26, 227–238.

Viarengo, A., 1989. Heavy metals in marine invertebrates: mecha-nisms of regulation and toxicity at the cellular level. CRC Critic.Rev. Aquat. Sci. 1, 295–317.

Waalkes, M.P., Fox, D.A., States, J.C., Patierno, S.R., Mc-Cabe,M.J.J., 2000. Metals and disorders of cell accumulation, modula-tion of apoptosis and cell proliferation. Toxicol. Sci. 56, 255–261.