General and Comparative Endocrinology 138 (2004) 58–69

www.elsevier.com/locate/ygcen

Research report

Environmental estrogens can aVect the function of mussel hemocytes through rapid modulation of kinase pathways

Laura Canesi,a,¤ Lucia Cecilia Lorusso,a Caterina Ciacci,a Michele Betti,a Massimiliano Zampini,b and Gabriella Galloc

a Istituto di Scienze Fisiologiche, Università di Urbino, “Carlo Bo,” Loc. Crocicchia, 61029 Urbino (PU), Italyb Istituto di Microbiologia e Scienze Biomediche, Università Politecnica delle marche, Ancona 60131, Italy

c DIBISAA, Università di Genova, Genova 16132, Italy

Received 26 February 2004; revised 13 April 2004; accepted 13 May 2004

Abstract

Estrogens and estrogenic chemicals can aVect several vertebrate non-reproductive functions, the immune response in particular.We have previously shown that in the hemocytes of the marine mollusc Mytilus the natural estrogen 17�-estradiol (E2) can aVect theimmune function through rapid tyrosine kinase-mediated signalling pathways converging on phosphorylation of both mitogen acti-vated protein kinases (MAPKs) and signal transducers and activators of transcription (STATs), whose activation plays a key role inthe immune response. In this work the eVects of synthetic estrogens (such as DES), estrogenic chemicals (such as Bisphenol A,Nonylphenol), and plant estrogens (genistein) on mussel hemocytes were evaluated. The results demonstrate that all the EDCs testedexert in vitro eVects similar to those of E2 on lysosomal membrane stability, although at concentrations 1000 times higher than thoseof the natural estrogen. When the eVects of DES, BPA, and NP on tyrosine kinase-mediated cell signalling were investigated, estro-genic compounds showed distinct eVects on the phosphorylation state of MAPK and STAT members. In particular, only DES, likeE2, induced p38 MAPK phosphorylation, whereas BPA and NP seem to have opposite eVects. Moreover, diVerent EDCs signiW-cantly decreased the tyrosine phosphorylation state of STAT3 and STAT5, showing a distinct eVect with respect to E2. Experimentswith speciWc kinase inhibitors showed that activation of p38 MAPK, but also of ERK MAPK and PI3-kinase, plays a key role inmediating the eVect of DES. On the other hand, the eVects of NP were partly mediated by ERK MAPK activation. BPA-inducedlysosomal membrane destabilisation was unaVected by either MAPK or PI3-K inhibitors. However, hemocyte pre-treatment withthe PKC inhibitor GF109203X prevented the eVects of both BPA and NP, this indicating that kinase pathways other than thoseinvolving MAPKs are also responsible for mediating the eVects of certain EDCs. Overall, the results support the hypothesis thatEDCs may rapidly modulate the function of mussel hemocytes through activation of transduction pathways involving diVerentkinase-mediated cascades. Moreover, the eVects of EDCs on the phosphorylation state of transcription factor STATs suggest thatthese compounds may lead to changes in gene expression secondary to modulation of kinase/phosphatases. Our data address to theimportance of investigating full range responses to estrogenic chemicals and may help understanding their basic mechanisms ofaction in ecologically relevant invertebrate species. 2004 Elsevier Inc. All rights reserved.

Keywords: Endocrine disrupting chemicals; Mytilus; Hemocytes; Immune function; Kinase-mediated cell signalling; MAPK; STAT

1. Introduction

Endocrine disrupting chemicals (EDCs) include bothnatural and synthetic steroid estrogens, as well as a vari-ety of estrogen-mimicking chemicals, such as alkylphe-

¤ Corresponding author. Fax: +39-0722-304-226.E-mail address: canesi@uniurb.it (L. Canesi).

0016-6480/$ - see front matter 2004 Elsevier Inc. All rights reserved.doi:10.1016/j.ygcen.2004.05.004

nols and alkylphenol ethoxylates, PCBs, dioxins, variouspesticides, and herbicides. There is considerable evidencefrom both laboratory and Weld studies that EDCexposure is at least partly responsible for disruption ofreproduction and development in vertebrates (reviewedby McLachlan, 2001; Rotchelle and Ostrander, 2003;Tapiero et al., 2002; Witorsch, 2002). Most EDCs havebeen demonstrated to bind both Wsh and mammalian

L. Canesi et al. / General and Comparative Endocrinology 138 (2004) 58–69 59

intracellular estrogen receptors (ERs) (Kuiper et al.,1998; McLachlan, 2001; Rotchelle and Ostrander, 2003;Tapiero et al., 2002). ERs, after estrogen binding andtranslocation into the nucleus, act as ligand-inducibletranscription factors speciWcally regulating expression oftarget genes (Hall et al., 2001; Moggs and Orphanides,2001; Nadal et al., 2001). This is generally referred to asgenomic or “classical pathway” of estrogen action; itinvolves direct participation of the ERs as transcriptionfactors without virtually any other previous signallingstep, and requires hours to days for an increase in geneexpression to occur. With this respect, EDCs are gener-ally found to be approximately 103–106 times less potentthan the naturally occurring 17�-estradiol (E2) (Kuiperet al., 1998; McLachlan, 2001; Rotchelle and Ostrander,2003).

In mammalian cells, estrogens can also trigger rapid(within seconds to minutes) eVects through non-genomicor “alternative pathways” (Kelly and Levin, 2001; Lösel etal., 2003; Nadal et al., 2001; Driggers and Segars, 2002;Segars and Driggers, 2002). These may be initiated ateither membrane or cytosolic locations and can result inboth direct local eVects (such as modiWcation of ionXuxes) and regulation of gene transcription secondary toactivation of kinase cascades (involving cAMP, MAPKs,PKC and PKA, PI-3K, etc.). Elicited responses dependupon the cell type studied and the conditions used; how-ever, rapid changes in phosphorylation state of mitogenactivated protein kinases (MAPKs) and in cytosolic [Ca2+]are among the most common events observed in non-genomic eVects of E2 (Lösel et al., 2003; Nadal et al., 2001;Driggers and Segars, 2002; Segars and Driggers, 2002).Recent data indicate that certain EDCs can also mimicthe rapid mechanisms of action of E2 involving both Ca2+

and kinase-mediated signalling (Klotz et al., 2000; Lee etal., 2003; Noguchi et al., 2002; Quesada et al., 2002).

Estrogens exert a broad spectrum of activities throughboth genomic and rapid mechanisms on a wide variety ofvertebrate cell and tissues (Moggs and Orphanides, 2001;Razandi et al., 2000), the immune system in particular(Bruce-Keller et al., 2000; Guo et al., 2002; Vegeto et al.,2001). A growing body of evidence indicates that alsoEDCs can aVect the immune function in mammals (AnsarAhmed, 2000; Kim and Jeong, 2003; Lee et al., 2003;Sakazaki et al., 2002; Sugita-Konishi et al., 2003).

In comparison to vertebrate systems, a limited num-ber of studies have examined the eVects of EDCs ininvertebrates. This is partly due to lack of knowledge oninvertebrate endocrinology and on the mechanisms ofaction of natural hormones (Depledge and Billinghurst,1999; Gagnè et al., 2001; Jobling et al., 2003; Matozzoet al., 2003; Ohelmann et al., 2000; Rotchelle and Ostra-nder, 2003); in particular, the possible eVects of estro-genic compounds on invertebrate immune cells arelargely unknown. We have recently shown that in mussel(Mytilus galloprovincialis Lam.) hemocytes, the cells

responsible for innate immunity (Renwrantz, 1990), thenatural estrogen 17�-estradiol (E2) rapidly aVects diVer-ent immune parameters (Canesi et al., 2004). Activationof kinase cascades, in particular those leading to thestress-activated p38 MAPK, but also to ERK MAPKsand PI3-Kinase, were mainly involved in mediating therapid eVects of E2 on the hemocyte function. Moreover,E2 induced signiWcant increase in the tyrosine phosphor-ylation of signal transducers and activators of transcrip-tion (STAT)-like proteins. Overall, the resultsdemonstrated that E2 can rapidly modulate the immunefunction of mussel hemocytes through activation oftyrosine kinase-mediated cascades.

These results prompted us to investigate whether alsoEDCs may aVect the function of Mytilus hemocytesthrough similar mechanisms. To this aim, the eVects ofdiVerent EDCs on lysosomal membrane stability and onthe phosphorylation state of MAPK- and STAT-likecomponents were evaluated. In particular, the eVects ofDES (diethylstylbestrol), BPA (Bisphenol-A; 4,4�isopro-pylidenediphenol), and NP (Nonylphenol) were investi-gated. DES has been used for decades as the syntheticestrogen of choice both in medicine and cattle feeding;DES and its metabolites are persistent and accumulatein the environment (McLachlan, 2001; Rotchelle andOstrander, 2003; Tapiero et al., 2002). Both BPA and NPhave been shown to bind estrogen receptors and toinduce estrogenic activity in a number of assays in mam-malian systems. BPA is a monomer used in the manufac-ture of polycarbonate plastics and in a variety of plasticproducts, including food can linings, compact disks,thermal (fax) papers, etc. NP consists of a mixture ofpredominantly para-substitued monoalkylphenolswidely used in the manufacture of non-ionic surfactants,lubricant additives, polymer stabilisers, and antioxi-dants. Both compounds can leach out of certain prod-ucts and are also detected at high levels in humansamples as well as in aquatic organisms (McLachlan,2001; Rotchelle and Ostrander, 2003; Tapiero et al.,2002).

2. Materials and methods

2.1. Animals

Mussels (Mytilus galloprovincialis Lam.) 4–5 cm long,were obtained from S.E.A. (Gabicce Mare, PU) and keptfor 1–3 days in static tanks containing artiWcial sea water(ASW) (1 L/mussel) at 16°C. Sea water was changed daily.

2.2. Hemolymph collection, preparation of hemocytemonolayers, and hemocyte treatment

Hemolymph was extracted from the posterior adduc-tor muscle of 8–20 mussels, depending on the experi-

60 L. Canesi et al. / General and Comparative Endocrinology 138 (2004) 58–69

ment, using a sterile 1 ml syringe with a 18 G1/2 in.needle. With the needle removed, hemolymph wasWltered through a sterile gauze and pooled in 50 ml Fal-con tubes at 4 °C. Hemolymph serum was obtained bycentrifugation of whole hemolymph at 200g for 10 minand the supernatant was sterilised through a 0.22�mpore size Wlter. Hemocyte monolayers were prepared aspreviously described (Canesi et al., 2002). BrieXy, aliqu-ots of 0.5 ml of hemolymph, corresponding to about 1–2 £ 106 cells, were seeded onto glass cover slips(40 £ 22 mm) placed in plastic culture dishes and incu-bated at 16 °C for 30 min to allow for cell attachment.Non-adherent hemocytes were subsequently removed bygently washing the preparations with sterilised ASW.Hemocyte monolayers were added with 1.5 ml of hemol-ymph serum and kept before use at 16 °C.

Hemocytes were incubated at 16 °C for diVerent peri-ods of time with diVerent EDCs (from stock solutions inethanol) suitably diluted in either ASW or hemolymphserum. Untreated and control vehicle hemocyte sampleswere run in parallel. In experiments with antiestrogens,Tamoxifen (Wnal concentration 100 nM from a 10 mMstock solution in ethanol) was added 10 min beforeEDCs.

2.3. Lysosomal membrane stability

Lysosomal membrane stability in control hemocytesand hemocytes incubated with diVerent concentrationsof EDCs for 30 min was evaluated by the neutral redretention assay as previously described (Canesi et al.,2003a) following the procedure of Lowe et al. (1995).Hemocyte monolayers on glass slides were incubatedwith 30�l of a NR solution (Wnal concentration 40 �g/mlfrom a stock solution of NR 20 mg/ml DMSO); after15 min the excess dye was washed out, 30�l of artiWcialsea water was added, and slides were sealed with a cover-slip. Every 15 min slides were examined under an opticalmicroscope and the percentage of cells showing loss ofdye from lysosomes in each Weld was evaluated. For eachtime point 10 Welds were randomly observed. The end-point of the assay was deWned as the time at which 50%of the cells showed sign of lysosomal leaking (the cytosolbecoming red and the cells rounded).

In experiments with kinase inhibitors, hemocyte mon-olayers were pre-treated for 20 min with PD98059 (forERK MAPK), SB203580 (for p38 MAPK) or wortman-nin (for PI3-kinase), and 30 min with GF109203X (forPKC) suitably diluted in ASW from 10 mM stock solu-tions in DMSO as previously described (Canesi et al.,2002). After pre-incubation, supernatants were discardedand hemocytes were incubated with EDCs or serum for30 min and then tested for the NR assay. Control hemo-cytes were maintained for 20 min in serum in the pres-ence of DMSO (0.1% Wnal concentration), as pre-treatedcells, then washed, and incubated with serum for 30 min.

2.4. Electrophoresis and Western blotting

Levels of phosphorylated MAPK- and STAT-likeproteins in whole cell extracts from hemocyte monolayerswere determined using phosphospeciWc antibodies as pre-viously described (Canesi et al., 2002, 2003b, 2004).Supernatants from each culture dish were discarded andeach hemocyte monolayer was lysed with 1 ml of ice-coldlysis buVer [50 mM Tris–HCl, pH 7.8, 0.25 M sucrose, 1%(w/v) sodium dodecyl sulfate (SDS), 1�g/ml pepstatin,10�g/ml leupeptin, 2 mM sodium orthovanadate, 10 mMNaF, 5 mM EDTA, 5 mM NEM (N-ethylmaleimide),40�g/ml phenylmethylsulphonyl Xuoride (PMSF), and0.1% Nonidet-P40] and sonicated for 45 s at 50 W. Sam-ples were boiled for 4 min and then centrifuged for 10 minat 14,000g to remove insoluble debris. Supernatants weremixed 1:1 (v:v) with sample buVer (0.5 M Tris–HCl, pH6.8, 2% SDS, 10% glycerol, 4% of 2-mercaptoethanol, and0.05% bromophenol blue) and samples (normalised forprotein content before loading to 30 �g of protein) wereresolved by 12% (for MAPKs), 7% (for STATs) SDS–polyacrylamide gel electrophoresis (Laemli, 1970). Pre-stained molecular-mass markers were run on adjacentlanes. The gels were electroblotted and stained with Coo-massie blue according to Towbin et al. (1979). Blots wereprobed with human recombinant speciWc anti-phospho-ERK1/2 or anti-phospho-p38 MAPK as primary anti-bodies (1:1000) and horseradish-peroxidase-conjugatedgoat anti-rabbit IgG (1:3000) was used as a secondaryantibody. Nitrocellulose membranes were stripped for30 min at 50 °C with stripping buVer (62.5 mM Tris–HCl,pH 6.7, containing 10 mM �-mercaptoethanol and 2%SDS) and reprobed with anti-actin antibodies (1:1000) asa loading control. Blots were also probed with the speciWcanti-phospho-STAT3 and anti-phospho-STAT5 (1:500)as primary antibodies, and horseradish-peroxidase-con-jugated goat anti-rabbit IgG (1:3000) was used as a sec-ondary antibody. After probing with anti-phospho-STATs, membranes were stripped and reprobed withanti-STAT3 or anti-STAT5 antibodies (1:500).

Immune complexes were visualised using anenhanced chemioluminescence Western blotting analysissystem (Amersham–Pharmacia) following the manufac-turer’s speciWcations. Western blots Wlms were digitised(Chemidoc–Biorad) and band optical densities werequantiWed using a computerised imaging system (Quan-tityOne). Relative optical densities (arbitrary units) werenormalised for the control band in each series.

2.5. Data analysis

The results are means of at least three experiments intriplicate §SD. Data from densitometric analyses ofWestern blots are means § SD of three independentexperiments. Statistical analysis was performed by usingthe Mann–Whitney U test with signiWcance at P 6 0.05.

L. Canesi et al. / General and Comparative Endocrinology 138 (2004) 58–69 61

2.6. Chemicals

All reagents were of analytical grade. Diethylstilbes-trol, 17�-ethynylestradiol, genistein, and Tamoxifenwere from Sigma (St. Louis, MO); Bisphenol A was fromGmbH Dr. Ehrenstorfer (Germany); 4-n-Nonylphenolwas from Riedel-de Haen (Germany).

Rabbit polyclonal anti-phospho-MAPK (ERK1/2)(Thr202/Tyr204), anti-phospho-p38 MAPK (Thr180/Tyr182) (human recombinant) antibodies and poly-clonal anti-phospho-STAT3 (Tyr705), and anti-phos-pho-STAT5 (Tyr694) antibodies were obtained fromNew England Biolabs (Beverly, USA). Mouse monoclo-nal anti-STAT3 (F-2) and anti-STAT5 (G-2) antibodieswere from Santa Cruz Biotechnology (CA). Anti-�-actinantibodies (rabbit polyclonal) were from Sigma. Pre-stained LMW and HMW molecular-mass markers werefrom Bio-Rad (CA). All other reagents were purchasedfrom Sigma (St. Louis, MO).

3. Results

3.1. EVect of synthetic and environmental estrogens onhemocyte lysosomal membrane stability

A number of EDCs, from the synthetic estrogensdiethylstylbestrol (DES) and 17�-ethynylestradiol (EE),to the plant estrogen genistein, were tested for their eVecton hemocyte lysosomal membrane stability and theresults are shown in Fig. 1. In control hemocytes, NRretention time was 110 § 10 min (mean § SD). In hemo-cytes incubated for 30 min with estrogenic compounds inthe low micromolar range and then tested for the NRassay, a signiWcant concentration-dependent destabilisa-tion of lysosomal membranes was observed (Fig. 1A).The results indicate that, in mussel hemocytes, the syn-thetic estrogens DES and EE have the same potency as

that of the estrogenic chemicals BPA and NP and of theplant estrogen genistein. A comparable eVect was previ-ously obtained with a 103 lower concentrations of thenatural estrogen E2 (Canesi et al., 2004). Hemocyte pre-treatment with the antiestrogen Tamoxifen (100 nM for10 min) signiWcantly inhibited the eVect of DES and BPAand prevented that of NP (Fig. 1B).

3.2. EVect of synthetic and environmental estrogens onphosphorylation of immunoreactive MAPK- and STAT-like proteins

In order to investigate the possible eVects of EDCson tyrosine kinase-mediated cell signalling in musselhemocytes, the phosphorylation state of MAPK andSTAT members was evaluated in protein extracts fromhemocytes incubated for diVerent periods of time (5–60 min) with the synthetic estrogen DES and the resultswere compared with those of the two estrogenic chemi-cals Bisphenol A (BPA) and Nonylphenol (NP). Phos-phorylation of MAPK-like and STAT-like Mytilusproteins were evaluated as previously described byelectrophoresis and Western blotting utilising speciWcanti-phospho-MAPK antibodies (Canesi et al., 2002)and anti-phospho-STAT antibodies (Canesi et al.,2003b).

Fig. 2 shows the eVects of DES on MAPK phosphory-lation. As reported in Fig. 2A, DES (25�M) induced arapid, large, and persistent increase in the level of phos-phorylated stress-activated p38, with a maximum at 15–30 min (about sevenfold with respect to controls;P6 0.05); after 60 min incubation with DES the level ofphosphorylated p38 was still about 3.6-fold higher thanin controls (P 60.05). DES also induced a small andtransient increase in the level of phosphorylated ERK2,with a maximum at 15 min (about 1.4-fold with respect tocontrols, as evaluated by densitometric band analysis;P6 0.05) followed by a decrease at 60 min. The eVects of

Fig. 1. EVect of EDCs on lysosomal membrane stability of mussel hemocytes. Hemocyte monolayers were incubated for 30 min with each compoundand then tested for NR retention time. (A) Dose-dependence; (B) eVect of Tamoxifen pre-treatment (100 nM, 10 min) on lysosomal membrane desta-bilisation induced by 25 �M EDCs. Data are means § SD of four experiments in triplicate. E, EDC-treated and T/E, Tamoxifen + EDC-treated.*P 6 0.05 EDC-treated vs controls; °P 6 0.05 T/E vs E.

62 L. Canesi et al. / General and Comparative Endocrinology 138 (2004) 58–69

DES were also evaluated in hemocytes pre-treated withTamoxifen (100 nM) (Fig. 2B). In Tamoxifen-treatedcells, the DES-induced phosphorylation of MAPKs wassigniWcantly aVected: the antiestrogen largely reduced theDES-mediated increase in p38 MAPK phosphorylation(the maximal increase with respect to controls at 15 minbeing reduced from sevenfold to about threefold in theabsence and in the presence of Tamoxifen, respectively).Moreover, in these conditions, DES induced a progres-sive and dramatic decrease in ERK2 phosphorylation(from ¡50% at 15 min to ¡90% at 60 min). In Figs. 2Aand B anti-actin blots are shown as loading controls.

In Fig. 3 the eVects of DES on the phosphorylationstate of diVerent STATs are reported. As shown inFig. 3A, DES induced a progressive and large decreasein the level of p-STAT3, with a minimum at 60 min(about ¡80%; P 6 0.05). On the other hand, DES rap-idly increased the level of phosphorylated STAT5between 5 and 30 min (with a maximum 2.2-fold increasewith respect to controls at 15 min; P 6 0.05) that was fol-lowed by a decrease at 60 min. The levels of unphosphor-ylated STAT3- and STAT5-like proteins did not changewith DES incubation. Tamoxifen pre-treatment did notsigniWcantly aVect the DES-induced changes in the levelof p-STAT3 or p-STAT5 (Fig. 3B). Similar, althoughsmaller eVects on MAPK and STAT phosphorylationwere observed with lower DES concentrations (5 �M)(not shown).

Fig. 4 shows the eVects of BPA on MAPK and STATphosphorylation. In hemocytes incubated with 25 �MBPA (Fig. 4A), the phosphorylation state of the stress-

activated p38 MAPKs was largely decreased (about¡50% with respect to controls from 15 to 60 min;P 6 0.05). On the contrary, in BPA-treated hemocytes, arapid and transient increase in the level of phosphory-lated ERK2 was observed, with a maximum at 5 min(about twofold with respect to controls; P 6 0.05). Asshown in Fig. 4B, BPA caused a transient increase in thelevel of p-STAT3, with a maximum at 15 and 30 min (upto threefold; P 6 0.05) followed by a decrease to controllevels at 60 min, whereas the level of total, unphosphory-lated STAT3-like protein did not change. On the otherhand, BPA induced a progressive and dramatic decreasein the level of p-STAT5, with a minimum at 60 min(¡80%; P 6 0.05), without aVecting the level of unphos-phorylated STAT5-like protein.

The eVects of NP on the phosphorylation of MAPKsand STATs are reported in Fig. 5. As shown in Fig. 5A,p38 MAPK phosphorylation did not show a clear timecourse in response to NP (25 �M); however, a signiWcantdecrease was observed at both 5 and 60 min (about¡50% with respect to controls; P 6 0.05). NP induced arapid and transient increase in the level of phosphory-lated ERK2 with a maximum at 15 min (about 1.6-foldwith respect to controls; P 6 0.05). Interestingly, asreported in Fig. 5B, NP induced a signiWcant and rapid(from 5 min) decrease in the level of both p-STAT3 andp-STAT5, without aVecting the level of unphosphory-lated STAT3- and STAT5-like proteins. In particular,the NP-induced decrease in p-STAT3 and p-STAT5were largest at 30 and 15 min, respectively (¡60% and¡80%, respectively; P 6 0.05).

Fig. 2. EVect of DES (25 �M) on MAPK phosphorylation in hemocyte monolayers. Protein extracts from hemocytes were subjected to 12% SDS–PAGE followed by Western blotting using polyclonal phosphospeciWc antibodies to MAPKs. (A) Time course of p38 and ERK2 MAPK phosphory-lation, respectively, in control hemocytes and hemocytes incubated with DES for diVerent periods of time; (B) results of the same experiments carriedout in hemocytes pre-treated with the antiestrogen Tamoxifen (100 nM, 10 min) (TAM/DES). Anti-actin blots are shown as loading controls. Bandswere detected using enhanced chemioluminescence reagents (see Section 2). Results are representative of three independent experiments. C, control.Insets: densitometric analysis of blots from three independent experiments (mean § SD). Relative increases in band optical densities (arbitrary units)were normalised for the control band in each series.

L. Canesi et al. / General and Comparative Endocrinology 138 (2004) 58–69 63

MAPK and STAT phosphorylation was also evalu-ated in hemocytes pre-treated with Tamoxifen and thenincubated with either BPA or NP. Overall, the diVerent

eVects of Tamoxifen pre-treatment on both BPA- andNP-induced phosphorylation state of MAPK and STATmembers was similar to that obtained with DES (data

Fig. 3. EVect of DES (25 �M) on STAT tyrosine phosphorylation in hemocyte monolayers. Protein extracts from control hemocytes and hemocytesincubated with DES for diVerent periods of time were subjected to 7% SDS–PAGE followed by Western blotting using polyclonal phosphospeciWcantibodies to phosphorylated STAT3 and STAT5. Blots were stripped and reprobed with antibodies to the corresponding total, unphosphorylatedSTATs. (A) Time course of STAT3 and STAT5 tyrosine phosphorylation, respectively, in control hemocytes and hemocytes incubated with DES fordiVerent periods of time; (B) results of the same experiments carried out in Tamoxifen-pre-treated hemocytes (TAM/DES). Bands were detectedusing enhanced chemioluminescence reagents (see Section 2). Results are representative of three independent experiments. C, control. Inset: densito-metric analysis of blots from three independent experiments (mean § SD). Relative increases in band optical densities (arbitrary units) were norma-lised for the control band in each series.

Fig. 4. EVect of BPA (25 �M) on MAPK and STAT phosphorylation in control hemocytes and hemocytes incubated with BPA for diVerent periodsof time. Protein extracts from hemocytes were subjected to either 12 or 7% SDS–PAGE followed by Western blotting using polyclonal phosphospec-iWc antibodies to MAPKs and STATs. C, control. (A) Time course of p38 and ERK2 and MAPK phosphorylation; anti-actin blots are shown asloading controls; (B) time course of STAT3 and STAT5 tyrosine phosphorylation; anti-STAT blots are shown as loading controls. Bands weredetected using enhanced chemioluminescence reagents (see Section 2). Results are representative of three independent experiments. Insets: densito-metric analysis of blots from three independent experiments (mean § SD). Relative increases in band optical densities (arbitrary units) were norma-lised for the control band in each series.

64 L. Canesi et al. / General and Comparative Endocrinology 138 (2004) 58–69

not shown). The level of phosphorylated MAPK- andSTAT-like proteins did not change in control hemocytesor in vehicle-treated hemocytes incubated for diVerenttimes in the absence of estrogens (data not shown).

The eVects of hemocyte treatment with diVerent EDCs(25�M) for diVerent periods of time on the phosphoryla-tion state of diVerent MAPK- and STAT-like membersare summarised in Table 1, where data indicate the meanpercent changes in phosphorylated protein bandsinduced by EDC treatment with respect to controls. AllEDCs induced a small but signiWcant and transientincrease in ERK2 MAPK phosphorylation, with a maxi-mum at 5–15 min. On the other hand, DES greatly stimu-lated p38 MAPK phosphorylation, whereas both BPAand NP showed an opposite eVect. STAT3 phosphoryla-tion was largely decreased by both DES and NP, whereasa transient increase was observed with BPA. On the con-trary, the level of p-STAT5 was increased by DES treat-ment and greatly decreased by both BPA and NP.

3.3. EVects of kinase inhibitors on EDC-inducedlysosomal membrane destabilisation

In order to investigate the possible relationshipbetween the eVects of EDCs on hemocyte membranedestabilisation and those on kinase-mediated cell signal-ling, the same experiments were carried out in hemocytespre-treated with diVerent kinase inhibitors as previouslydescribed (Canesi et al., 2002, 2004) and the results areshown in Fig. 6. Cell pre-treatment with SB203580, a spe-

ciWc inhibitor of p38 MAPK activation, completely pre-vented the lysosomal destabilisation induced by DES;signiWcant inhibition was also observed with the ERKMAPK inhibitor PD98059 and with wortmannin, theinhibitor of PI3-kinase (phosphatidylinositol 3-kinase).PD98059 signiWcantly reduced NP-induced lysosomal

Table 1Time course of EDC-induced changes in the phosphorylation state ofMAPK- and STAT-like members in mussel hemocytes

Data represent the results of densitometric analyses of blots (meanof three independent experiments). Relative band optical densities(arbitrary units) were normalised for the control band in each seriesand results are expressed as mean % change vs controls.

* P 6 0.05.

% change vs controls

Treatment 5 (min) 15 (min) 30 (min) 60 (min)

DESp-ERK2 MAPK ¡20 +40* +20 ¡65*

p-p38 MAPK +300* +580* +610* +260*

p-STAT3 ¡20 ¡40 ¡60* ¡80*

p-STAT5 +100* +120* +80* ¡40*

BPAp-ERK2 MAPK +100* +50* ¡10 0p-p38 MAPK 0 ¡50* ¡50* ¡50*

p-STAT3 +50* +210* +210* 0p-STAT5 ¡15 ¡25 ¡55* ¡80*

NPp-ERK2 MAPK +40 +60* +10 0p-p38 MAPK ¡40* 0 ¡15 ¡50*

p-STAT3 ¡60* ¡60* ¡90* ¡50*

p-STAT5 ¡40* ¡60* ¡55* ¡40*

Fig. 5. EVect of NP (25 �M) on MAPK and STAT phosphorylation in control hemocytes and hemocytes incubated with NP for diVerent periods oftime. Protein extracts from hemocytes were subjected to either 12% or 7 SDS–PAGE followed by Western blotting using polyclonal phosphospeciWcantibodies to MAPKs and STATs. C, control. (A) Time course of p38 and ERK2 MAPK phosphorylation; anti-actin blots are shown as loading con-trols; (B) time course of STAT3 and STAT5 tyrosine phosphorylation; anti-STAT blots are shown as loading controls. Bands were detected usingenhanced chemioluminescence reagents (see Section 2). Results are representative of three independent experiments. Insets: densitometric analysis ofblots from three independent experiments (mean § SD). Relative increases in band optical densities (arbitrary units) were normalised for the controlband in each series.

L. Canesi et al. / General and Comparative Endocrinology 138 (2004) 58–69 65

membrane destabilisation, whereas both SB203585 andwortmannin were ineVective. On the other hand, none ofthe inhibitors tested signiWcantly reduced the eVect ofBPA. Interestingly, cell pre-treatment with the PKCinhibitor GF109203X prevented the lysosomal destabili-sation induced by BPA and NP, without aVecting thatinduced by DES.

Hemocyte pre-treatment with diVerent inhibitors didnot aVect lysosomal membrane stability.

4. Discussion

We have recently shown that in mussel hemocytesnanomolar concentrations of the natural estrogen 17�-estradiol (E2) cause an increase in cytosolic [Ca2+] and inthe phosphorylation state of the components of tyrosinekinase-mediated signal transduction MAPK- andSTAT-like proteins (Canesi et al., 2004). Both eVects ofE2 on cell signalling were rapid, occurring from secondsto minutes, and were associated to both morphologicaland functional eVects, such as cell shape changes, releaseof lysosomal enzymes, lysosomal membrane destabilisa-tion, and stimulation of bactericidal activity, indicatingan E2-induced modulation of the hemocyte function.Hemocyte pre-treatment with speciWc kinase inhibitorsindicated that rapid activation of kinase cascades involv-ing p38 MAPK, ERK2 MAPK, and PI3-kinase, ratherthan changes in [Ca2+], were mainly responsible formediating the eVects of E2 in mussel hemocytes.

In order to compare the possible eVects of EDCs withthose of E2, we tested the eVect of diVerent compoundson the stability of lysosomal membranes. This parameterwas chosen on one hand since it is generally considered

Fig. 6. EVect of hemocyte pre-treatment with diVerent kinase inhibi-tors on the lysosomal membrane destabilisation induced by DES,BPA, and NP. Before incubation with diVerent EDCs, hemocytes werepre-treated for 20 min with SB203580 (20 �M), PD98059 (20 �M),wortmannin (0.1 �M), or for 30 min with GF109203X (2.5 �M). Dataare means § SD of four experiments in triplicate; signiWcant diVerencewith respect to controls *P 6 0.05 EDC-treated vs controls.

as representative of both hemocyte function in general(Lowe et al., 1995) and of the immune response in partic-ular (Hauton et al., 2001); on the other, because it wassigniWcantly aVected in a clear concentration-dependentmanner by nanomolar E2 through activation of kinasecascades (Canesi et al., 2004). The results here reportedshow that also diVerent environmental estrogens (DES,EE, BPA, NP, and genistein) signiWcantly reduced thestability of lysosomal membranes. The eVects of all com-pounds tested were dose-dependent in the micromolarrange, this indicating that in mussel hemocytes thepotency of environmental estrogens, including the syn-thetic estrogens DES and EE, was about 1000 timeslower than that of the natural estrogen.

Previous data showed that well-known EDCs such asPCB congeners aVected the lysosomal membrane stabil-ity in the same concentration range (Canesi et al., 2003a).Taken together, our results conWrm that lysosomalmembrane destabilisation represents a rapid response ofmussel hemocytes common to both natural and environ-mental estrogens. Moreover, the results here obtainedwith the antiestrogen Tamoxifen suggest that lysosomalmembrane destabilisation induced by EDCs, may bemediated by Tamoxifen-sensitive ER-like receptors.

Both PCBs and E2 (Canesi et al., 2003a, 2004) werealso shown to activate tyrosine kinase-mediated cell sig-nalling in mussel hemocytes. Therefore, we investigatedthe possibility that also the synthetic estrogen, DES, andtwo among the most well-known EDCs, BPA and NP,may aVect the hemocyte function by rapidly increasingthe phosphorylation state of MAPKs and STATs. Theresults indicate that DES induced a large (up to seven-fold with respect to controls) and persistent increase inphosphorylated stress-activated p38 MAPK. Similar,although smaller eVects were observed with lower DESconcentrations (5 �M) (data not shown). A smaller, tran-sient and signiWcant increase in the phosphorylationstate of ERK2 MAPK was also observed. In hemocytespre-treated with Tamoxifen the eVect of DES on bothp38 and ERK2 MAPK were largely decreased.

Both BPA and NP induced a transient increase in thephosphorylation of ERK2 MAPK similar to thatinduced by DES. The results conWrm previous dataobtained with E2 and PCBs (Canesi et al., 2003a,b, 2004),indicating activation of ERK MAPKs represents a com-mon eVect of both natural and environmental estrogensin mussel hemocytes. On the other hand, diVerent estro-genic compounds showed distinct eVects on the phos-phorylation of the stress-activated p38 MAPK. DESinduced a persistent p38 phosphorylation similar to thatpreviously observed with E2 (Canesi et al., 2004) andwith PCBs (Canesi et al., 2003a). On the contrary, bothBPA and NP did not increase, or even decreased at cer-tain time points, p38 MAPK phosphorylation.

In mussel hemocytes, activation of MAPKs, and ofp38 MAPK in particular, plays a key role in the immune

66 L. Canesi et al. / General and Comparative Endocrinology 138 (2004) 58–69

response (Canesi et al., 2002). Our results show thatMAPKs, that are extremely conserved signalling compo-nents involved in mediating the cellular response toendogenous and environmental stimuli, represent targetsfor diVerent EDCs in these cells. Since the MAPK signalcascade has emerged as a rapid, non-genomic pathwaymediating the response to estrogens in many mammaliancells (Kelly and Levin, 2001; Lösel et al., 2003; Nadalet al., 2001; Driggers and Segars, 2002; Segars and Drig-gers, 2002), and in mussel hemocytes (Canesi et al.,2004), our data demonstrate that MAPKs are alsoinvolved in rapid cell signalling by EDCs.

The possible relationship between activation of cyto-solic kinase cascades and the eVect of EDCs on musselhemocytes was investigated utilising speciWc kinaseinhibitors as previously described (Canesi et al., 2002).Cell pre-treatment with the p38 MAPK inhibitorSB203580 completely prevented the DES-induced lyso-somal membrane destabilisation, conWrming that p38activation plays a major role in mediating the eVect ofDES, like that of E2 (Canesi et al., 2004). Also the ERKMAPK inhibitor PD98059 and the PI3-kinase inhibitorwortmannin signiWcantly reduced the eVect of DES; inmussel hemocytes, like in mammalian cells, PI3-kinasewould lie upstream of p38 MAPK activation (Canesiet al., 2002). Overall, the results conWrm that the signal-ling components involved in mediating the eVects ofDES are similar to those of E2 (Canesi et al., 2004).

The results obtained with SB203580 and wortmanninindicate that neither p38 MAPK nor PI-3K activation isinvolved in mediating this response to BPA and NP,conWrming the results of Western blotting, althoughdata obtained with PD98059 indicate a role for ERKMAPK activation in mediating the eVect of NP. On theother hand, the eVect of BPA was unaVected by any ofthe inhibitors tested, indicating that other mechanismsare involved in mediating the BPA-induced lysosomalmembrane destabilisation. Protein kinase C (PKC) rep-resents a possible candidate: a role for PKC in mediatingthe rapid eVects of E2, but also of BPA and PCBs hasbeen recently shown in mammalian systems (Papaco-stantinou et al., 2003; Shin et al., 2002). PKC isoformshave been identiWed in mussel tissues (Mercado et al.,2002). When hemocytes were pre-treated with the PKCinhibitor GF109203X, the eVects of BPA and NP onlysosomal destabilisation were inhibited, whereas that ofDES was not. PKC inhibition by GF109203X has beendemonstrated in molluscan hemocytes (Walker andPlows, 2003). These results address to the role of PKC-dependent transduction pathways in mediating the rapideVects of certain EDCs also in invertebrate cells.

EDCs also aVected tyrosine phosphorylation ofSTAT components in mussel hemocytes. In response toDES treatment, the level of p-STAT5 was increased,whereas that of p-STAT3 was largely decreased. TheeVects of DES on STAT3 and STAT5 phosphorylation

were unaVected by Tamoxifen. The opposite eVect wasobserved with BPA, that lead to an increase in p-STAT3and to a decrease in p-STAT5. NP decreased the phos-phorylation state of both STAT3 and STAT5. On thecontrary, E2 was shown to increase STAT tyrosine phos-phorylation in mussel hemocytes (Canesi et al., 2004)like in mammalian cells (Bjornstrom and Sjoberg, 2002).

EDCs have been shown to aVect the phosphorylationstate and activity of transcription factors such as CREBand NF-AT (Lee et al., 2003; Quesada et al., 2002). Ourresults indicate that the phosphorylation state of STATsrepresents a common target for the rapid action ofdiVerent EDCs. STATs are the only transcription factorthat are directly activated by tyrosine phosphorylation:every STAT contains a conserved tyrosine residue nearthe C terminus that is speciWcally phosphorylated uponactivation by receptor associated Jaks (Janus tyrosinekinases); once phosphorylated, they dimerise and trans-locate from the cytosol to the nucleus; serine phosphory-lation is also required for maximal transcriptionalactivity (Horvath, 2000). Each STAT family membershows a distinct pattern of activation by growth factorsand cytokines and can regulate the transcription of par-ticular genes in each cell type (Horvath, 2000; Leamanet al., 1996).

STAT-like members have been shown to be rapidlytyrosine phosphorylated in response to both bacterialchallenge and heterologous interferons in mussel hemo-cytes (Canesi et al., 2003b), this indicating a conservedrole for these transcription factors in the activation ofthe immune response in both invertebrate and vertebrateimmunocytes. Overall, our data suggest that, in musselhemocytes, estrogenic compounds may aVect gene tran-scription secondary to modulation of tyrosine kinase/phosphatase-mediated cascades, possibly interferingwith the immune function.

In particular an eVect common to the action of thethree EDCs tested in the present work, and distinct fromthat of the natural estrogen, was the decrease in the levelof either p-STAT3 or p-STAT5, or both. This suggeststhe possibility that certain EDCs may modulate the phos-phorylation state of signalling components also throughactivation of phosphotyrosine phosphatases. SpeciWcprotein tyrosine phosphatases SHP-1 and SHP-2 areinvolved in regulating the transduction of signals relayedfrom the cell surface to the nucleus, including MAPKs,Jak/STATs, and PI3-kinase pathways, and play a criticalrole in cell survival, proliferation, and diVerentiation(Paling and Welham, 2002; Qu, 2002); SHP-1 has beenshown to be stimulated also by E2 in PC12 cells (Chenet al., 2001). However, their role in modulating signaltransduction has been much less investigated than that oftyrosine kinases in mammalian models; their role in cellsignalling is largely unknown in invertebrates.

We cannot at present explain the diVerences observedin the eVects of diVerent EDCs on phosphorylation of

L. Canesi et al. / General and Comparative Endocrinology 138 (2004) 58–69 67

diVerent components of tyrosine kinase-mediated cellsignalling in mussel hemocytes. In vertebrates, the activ-ity of most EDCs is weak, due to the interactions ofthese substances with the ER. ERs from diVerent verte-brate species exhibit diVerent ligand preferences and rel-ative binding aYnities for estrogenic compounds, due tothe variability in the amino acid sequence within therespective ER ligand binding domains (Matthews et al.,2000). In mussel hemocytes, indirect evidence for thepresence of both ER�- and ER�-like receptors has beenobtained by Western blotting of protein extracts withanti-human-ER antibodies (Canesi et al., 2004). More-over, only certain eVects of E2 (Canesi et al., 2004) andEDCs (this work) were prevented or strongly reduced bythe antiestrogen Tamoxifen, some others were not. How-ever, Tamoxifen is a potent inhibitor of PKC (Boyanet al., 2003); therefore, the eVects of the antiestrogenmay be due not only to interactions of this compoundwith putative ER-like receptors, but also to inhibition ofEDC-mediated PKC activation.

ER-like forms highly homologous to human ERs arepresent in the opisthobranch mollusc Aplysia; however,the transcriptional activity of the Aplysia ER seems to beconstitutive, and independent of estrogen activation(Thornton et al., 2003). As suggested by the Authors, anancient steroid receptor (AncSR1) which is shared byother invertebrate groups, may have retained this capac-ity, that seems to be lost in the Aplysia lineage. Recently,a 266 bp sequence showing 100% homology with thehuman ER� was found to be expressed by Mytilus gan-glionic tissue, although no information is at presentavailable on the role of this receptor in mussels (Stefanoet al., 2003). Such a receptor form may be responsible formediating the rapid eVects induced by 17�-estradiol bothin Mytilus ganglia (Stefano et al., 2003) and hemocytes(Canesi et al., 2004). The nature and estrogen-dependentactivation of ER-like receptors in molluscs, as well as inmost invertebrate species, is still largely unexplored.However, as underlined by Witorsch (2002), ER bindingper se may have limited inXuence on endocrine disrup-tion. The nature (estrogenic or antiestrogenic) or magni-tude of the response is a function of the substance itself,as well as of other factors, such as complexities withinthe various stages of the ER signalling, cross-talkbetween ER and other signalling pathways, and alterna-tive modes of estrogen actions.

Most EDCs are resistant to environmental degrada-tion and are considered ubiquitous contaminants in theaquatic environment. In the marine environment, theytend to associate with particular matter and sedimentsand, being lipophilic, can be highly bioconcentrated; inmussel tissues, concentrations of NP in the range of ng/gwere detected (Günther et al., 2001), up to 250 ng/g inmussels from the Adriatic Sea (Ferrara et al., 2001).Mytilus seems to be particularly resistant to the toxiceVects of EDC exposure, with a LC50 of 3.0 mg NP/L at

96 h (Granmo et al., 1989). The EDC concentrations uti-lised in our ‘in vitro’ study were in the same concentra-tion range; however, preliminary ‘in vivo’ experimentsconWrm the activity of these compounds on the hemo-cyte function and phosphorylation state of MAPKs andSTATs also at 103 lower EDC concentrations and longerexposure periods (6–24 h) (Canesi, unpublished results).Therefore, it may be hypothesised that environmentallymore realistic EDC concentrations may aVect also ‘invivo’ the immune function of these organisms throughmodulation of kinase-mediated cell signalling.

Most environmental regulation and testing programsfor endocrine disrupters are focused solely on verte-brates. This is largely due to the general lack of know-ledge on basic endocrine physiology of many invertebrategroups. Our data, indicating a rapid action of EDCs oncell signalling in molluscan cells, address to the impor-tance of investigating full range responses to estrogenicchemicals and may help understanding their basic mech-anisms of action in ecologically relevant invertebratespecies.

References

Ansar Ahmed, S., 2000. The immune system as a potential target forenvironmental estrogens (endocrine disrupters): a new emergingWeld. Toxicology 156, 191–206.

Bjornstrom, L., Sjoberg, M., 2002. Signal transducers and activators oftranscription as downstream targets on nongenomic estrogenreceptor actions. Mol. Endocrinol. 16, 2202–2214.

Boyan, B.D., Sylvia, V.L., Frambach, T., Lohmann, C.H., Dietl, J.,Dean, D.D., Schwartz, Z., 2003. Estrogen-dependent rapid activa-tion of protein kinase C in estrogen receptor-positive MCF-7breast cancer cells and estrogen receptor-negative HCC38 cells ismembrane-mediated and inhibited by Tamoxifen. Endocrinology144, 1812–1824.

Bruce-Keller, A.J., Keeling, J.L., Keller, J.N., Huang, F.F., Camondola,S., Mattson, M.P., 2000. AntiinXammatory eVects of estrogen onmicroglia activation. Endocrinology 141, 3646–3656.

Canesi, L., Betti, M., Ciacci, C., Scarpato, A., Citterio, B., Pruzzo, C.,Gallo, G., 2002. Signaling pathways involved in the physiologicalresponse of mussel hemocytes to bacterial challenge: the role ofstress-activated p38 MAPK. Dev. Comp. Immunol. 26, 325–334.

Canesi, L., Ciacci, C., Betti, M., Scarpato, A., Citterio, B., Pruzzo, C.,Gallo, G., 2003a. EVects of PCB congeners on the immune functionof Mytilus hemocytes: alterations of tyrosine kinase-mediated cellsignalling. Aquat. Toxicol. 63, 293–306.

Canesi, L., Betti, M., Ciacci, C., Citterio, B., Pruzzo, C., Gallo, G.,2003b. Tyrosine kinase mediated cell signaling in the activation ofMytilus hemocytes: possibile role of STAT proteins. Biol. Cell 95,608–618.

Canesi, L., Ciacci, C., Betti, M., Lorusso, L.C., Marchi, B., Burattini, S.,Falcieri, E., Gallo, G., 2004. Rapid eVect of 17beta-estradiol on cellsignaling and function of Mytilus hemocytes. Gen. Comp. Endocri-nol. 136, 58–71.

Chen, Z.J., Che, D., Vetter, M., Liu, S., Chang, C.H., 2001. 17beta estra-diol inhibits solubile guanylate cyclase activity trough a proteintyrosine phosphatases in PC12 cells. J. Steroid Biochem. Mol. Biol.78, 451–458.

Depledge, M.H., Billinghurst, Z., 1999. Ecological signiWcance of endo-crine disruption in marine invertebrates. Mar. Pollut. Bull. 39, 32–38.

68 L. Canesi et al. / General and Comparative Endocrinology 138 (2004) 58–69

Driggers, P.H., Segars, J.H., 2002. Estrogen action and cytoplasmic sig-naling cascades. Part II: the role of growth factors and phosphory-lation in estrogen signaling. Trends Endocrinol. Metab. 13, 422–427.

Ferrara, F., Fabietti, F., Delise, M., Piccioli Bocca, A., Funari, E., 2001.Alkylphenolic compounds in edible molluscs of the Adriatic Sea(Italy). Environ. Sci. Technol. 35, 3109–3112.

Gagnè, F., Blaise, C., Salazaar, S., Hansen, P.D., 2001. Evaluation ofestrogenic eVects of municipal eZuents to the freshwater musselElliptio complanata. Comp. Biochem. Physiol. C 128, 213–225.

Granmo, A., Ekelund, R., Magnusson, K., Berggren, M., 1989. Lethaland sublethal toxicity of 4-Nonylphenol to the common mussel(Mytilus edulis L.). Environ. Pollut. 59, 115–127.

Guo, Z., Krucken, J., Benten, P.M., Wuderlich, F., 2002. Estradiol-induced nongenomic calcium signaling regulates genotropic signal-ing in macrophages. J. Biol. Chem. 277, 7044–7050.

Günther, K., Dürbeck, H.W., Kleist, E., Thiele, B., Prast, H., Schwuger,M., 2001. Endocrine-disrupting Nonylphenols-ultra-trace analysisand time-dependent trend in mussels from the German bight.Fresenius J. Anal. Chem. 371, 782–786.

Hall, J.M., Couse, J.F., Korach, K.S., 2001. The multifaceted mecha-nisms of estradiol and estrogen receptor signaling. J. Biol. Chem.276, 36869–36872.

Hauton, C., Hawkins, L.E., Hutchinson, S., 2001. Response of hemo-cyte lysosomes to bacterial inoculation in the oysters Ostrea edulisL. and Crassostrea gigas (Thunberg) and the scallop Pecten maxi-mus (L.). Fish ShellWsh Immunol. 11, 143–153.

Horvath, C.M., 2000. STAT proteins and transcriptional responses toextracellular signals. Trends Biochem. Sci. 25, 496–502.

Jobling, S., Casey, D., Rodgers-Gray, T., Ohelmann, J., Schulte-Ohel-mann, U., Pawlowski, S., Baunbeck, T., Turner, A.P., Tyler, C.R.,2003. Comparative responses of molluscs and Wsh to environ-mental estrogens and estrogenic eZuent. Aquat. Toxicol. 65,205–220.

Kelly, M.J., Levin, E.R., 2001. Rapid actions of plasma membraneestrogen receptors. Trends Endocrinol. Metab. 12, 152–156.

Kim, J.Y., Jeong, H.G., 2003. Down-regulation of inducible nitric oxidesynthase and tumor necrosis factor-� expression by Bisphenol Avia nuclear factor-kB inactivation in macrophages. Cancer Lett.196, 69–76.

Klotz, D.M., Hewitt, S.C., Korach, K.S., Diaugustine, R.P., 2000. Acti-vation of a uterine insulin-like growth factor I signalling pathwayby clinical and environmental estrogens: requirement of estrogenreceptor alpha. Endocrinology 141, 3430–3439.

Kuiper, G.G.J.M., Lemmne, J.G., Carlsson, B., Corton, J.C., Safe, S.H.,Van der Saag, P.T., Van der Burg, B., Gustafsson, J., 1998. Interac-tion of estrogenic chemicals and phytoestrogens with estrogenreceptor �. Endocrinology 139, 4252–4263.

Leaman, D.W., Leung, S., Li, X., Stark, G.R., 1996. Regulation ofSTAT-dependent pathways by growth factors and cytokines.FASEB J. 10, 1578–1588.

Laemli, U.K., 1970. Cleavage of structural proteins during assembly ofthe head bacteriophage T4. Nature 227, 680–685.

Lee, M.H., Chung, S.W., Kang, B.Y., Park, J., Lee, C.H., Hwang, S.Y.,Kim, T.S., 2003. Enhanced interleukin-4 production in CD4+Tcells and elevated immunoglobulin E levels in antigen-primedmice by bisphenol A and Nonylphenol, endocrine disruptors:involvement of nuclear factor At-1 and Ca2+. Immunology 109,76–86.

Lowe, D.M., Fossato, V.U., Depledge, M.H., 1995. Contaminant-induced lysosomal membrane damage in blood cells of musselsMytilus galloprovincialis from the Venice Lagoon: an in vitro study.Mar. Ecol. Prog. Ser. 129, 189–196.

Lösel, R.F., Falkestein, E., Feuring, M., Schultz, A., Tillmann, H.-C.,Rossol-Haseroth, K., Wehling, M., 2003. Non genomic steroidaction:controversies, questions, and answers. Physiol. Rev. 83, 965–1016.

Matozzo, V., Deppieri, M., Moschino, V., Marin, M.G., 2003. Evalua-tion of 4-Nonylphenol toxicity in the clam Tapes philippinarum.Environ. Res. 91, 179–185.

Mercado, L., Cao, A., Barcia, R., Ramos-Martinez, J.I., 2002. Regula-tory properties of p105, a novel PKC isozyme in mantle tissue frommarine mussels. Biochem. Cell. Biol. 80, 771–775.

McLachlan, J.A., 2001. Environmental signalling: what embryos andevolution teach us about endocrine disrupting chemicals. Endocr.Rev. 22, 319–341.

Matthews, J., Celius, T., Halgren, R., Zacharewski, T., 2000. DiVeren-tial estrogen receptor binding of estrogenic substances: a speciescomparison. J. Steroid Biochem. Mol. Biol. 74, 223–234.

Moggs, J.G., Orphanides, G., 2001. Estrogen receptors: orchestratorsof pleiotropic cellular responses. EMBO Rep. 2, 775–781.

Nadal, A., Díaz, M., Valverde, M.A., 2001. The estrogen trinity: mem-brane, cytosolic and nuclear eVects. News Physiol. Sci. 16, 251–255.

Noguchi, S., Nakatsuka, M., Asagiri, K., Habara, T., Tarata, M., Koni-shi, H., Kudo, T., 2002. Bisphenol A stimulates NO synthesisthrough a non-genomic estrogen receptor-mediated mechanism inmouse endothelial cells. Toxicol. Lett. 135, 95–101.

Ohelmann, J., Schulte-Ohelmann, U., Tillman, M., Markert, B., 2000.EVects of endocrine disruptors on prosobranch snails (Mol-lusca:gasteropoda) in the laboratory. Part I: bisphenol A and octyl-phenol as xeno-estrogens. Ecotoxicology 9, 383–397.

Paling, N.R.D., Welham, M.J., 2002. Role of the protein tyrosine phos-phatase SHP-1 (Src homology phosphatase-1) in the regulation ofinterleukin-3-induced survival, proliferation and signalling. Bio-chem. J. 368, 885–894.

Papacostantinou, A.D., Goering, P.L., Umbreit, T.H., Brown, K.M.,2003. Regulation of uterine hsp90, hsp72 and HSF-1 transcriptionin B6C3F1 mice by � estradiol and bisphenol A: involvement of theestrogen receptor and protein kinase C. Toxicol. Lett. 144, 257–270.

Qu, C.-K., 2002. Role of SHP-2 tyrosine phosphatase in cytokine-induced signalling and cellular response. Biochim. Biophys. Acta1592, 297–301.

Quesada, I., Fuentes, E., Viso-Leon, M.C., Soria, B., Ripoll, C., Nadal,A., 2002. Low doses of the endocrine disruptor Bisphenol-A andthe native hormone 17�-estradiol rapidly activate transcription fac-tor CREB. FASEB J. 16, 1671–1673.

Razandi, M., Pedram, A., Levin, E.R., 2000. Estrogen signals to thepreservation of endothelial cell form and function. J. Biol. Chem.276, 38540–38546.

Renwrantz, L., 1990. Internal defence system of Mytilus edulis. In: Stef-ano, G.B. (Ed.), Neurobiology of Mytilus edulis. Manchester Uni-versity Press, Manchester, pp. 256–275.

Rotchelle, J.M., Ostrander, G.K., 2003. Molecular eVects of endocrinedisrupters in aquatic organisms. J. Toxicol. Environ. Health B 6,453–495.

Sakazaki, H., Ueno, H., Nakamuro, K., 2002. Estrogen receptor � inmouse splenic lymphocytes: possible involvement in immunity.Toxicol. Lett. 133, 221–229.

Segars, J.H., Driggers, P.H., 2002. Estrogen action and cytoplasmic sig-naling cascades. Part I: membrane-associated signaling complexes.Trends Endocrinol. Metab. 13, 349–354.

Shin, H.-J., Gye, M.-C., Chung, K.-H., Yoo, B.-S., 2002. Activity ofprotein kinase C modulated the apoptosis induced by polychlori-nated biphenyls in human leukemic HL-60 cells. Toxicol. Lett. 135,25–31.

Stefano, G.B., Zhu, W., Mantione, K., Jones, D., Salamon, E., Cho, J.J.,Cadet, P., 2003. 17-� estradiol downregulates gnglionic microglialcells via nitric oxide release: presence of an estrogen receptor �transcript. Neuroendocrinol. Lett. 24, 130–136.

Sugita-Konishi, Y., Shimura, S., Nishikawa, T., Sunaga, F., Naito, H.,Susuki, Y., 2003. EVect of bisphenol A on non-speciWc immunode-fenses against non pathogenic Escherichia coli. Toxicol. Lett. 136,217–227.

L. Canesi et al. / General and Comparative Endocrinology 138 (2004) 58–69 69

Tapiero, H., Ba, G.N., Tew, K.D., 2002. Estrogens and environmentalestrogens. Biomed. Pharmacother. 56, 36–44.

Thornton, J.W., Need, E., Crews, D., 2003. Resurrecting the ancestralsteroid receptor: ancient origin of estrogen signalling. Science 301,1714–1717.

Towbin, H., Stahelin, T., Gordon, J., 1979. Electrophoretic transferof protein from polyacrylamide gels to nitrocellulose sheets:procedure and some applications. Proc. Natl. Acad. Sci. USA 76,4350–4354.

Vegeto, E., Bonincontro, C., Pollio, G., Sala, A., Viappiani, S., Nardi,F., Brusarelli, A., Viviani, B., Ciana, P., Maggi, A., 2001. Estrogenprevents the lipolysaccharide-induced inXammatory response inmicroglia. J. Neurosci. 21, 1809–1818.

Walker, A.J., Plows, L.D., 2003. Bacterial lipopolysaccharide modu-lates proteine kinase C signalling in Lymnea stagnalis hemocytes.Biol. Cell 95, 527–533.

Witorsch, R.J., 2002. Endocrine disruptors: can biological eVects bepredicted? Regul. Toxicol. Pharmacol. 36, 118–130.