690 L. CANESI ET AL.JOURNAL OF EXPERIMENTAL ZOOLOGY 286:690–698 (2000)

© 2000 WILEY-LISS, INC.

JEZ 0771

Immunoelectron Microscope Analysis of EpidermalGrowth Factor Receptor (EGFR) in Isolated Mytilusgalloprovincialis (Lam.) Digestive Gland Cells:Evidence for Ligand-Induced Changes in EGFRIntracellular Distribution

LAURA CANESI,1* MANUELA MALATESTA,2 SERAFINA BATTISTELLI,2

CATERINA CIACCI,1 GABRIELLA GALLO,1 ANDGIANCARLO GAZZANELLI2

1Istituto di Scienze Fisiologiche, Loc. Crocicchia, Urbino, Italy2Istituto di Istologia e Analisi di Laboratorio, Università di Urbino, Urbino(PE), 61029, Italy

ABSTRACT In mammalian cells, the binding of epidermal growth factor (EGF) to its receptor(EGFR), a glycoprotein with intrinsic tyrosine kinase activity, leads to the pleiotropic responses toEGF. Among these, a negative feedback response by stimulation of receptor internalization andlysosomal degradation, this attenuating signal transduction.

In this work, data are reported on the identification of specific EGFRs in isolated digestivegland cells from the marine mussel (Mytilus galloprovincialis Lam.) by immunoelectron micros-copy. In control digestive cells, EGFR immunoreactivity was mainly associated with cytoplasmicmembrane structures and, to a lesser extent, the cell membrane. The presence of EGFR-like re-ceptors was confirmed by Western blotting of digestive gland cell extracts with two different mono-clonal antibodies that recognize either intracellular or extracellular epitopes.

The addition of mammalian EGF resulted in significant time and temperature-dependent changesin EGFR subcellular distribution in mussel cells. In cells exposed to EGF for 0–15 min at 4°C, thedistribution of EGFR was not significantly different from that of the control cells. On the otherhand, at 18°C, an increased labelling along the cell membrane was observed after 5–10 min afterEGF addition, with a concomitant decrease in the cytoplasmic signal. Moreover, after 20 min ofexposure to EGF, ligand binding apparently resulted in EGFR compartmentation within the lysos-omes. These observations were confirmed by quantitative analysis of EGFR labelling at differenttimes of EGF exposure. Similar results were obtained utilizing the two different monoclonal antibod-ies. The results indicate that, in mussel digestive cells, the binding of heterologous EGF to specificreceptors induces a negative feedback response by stimulating the lysosomal degradation of EGFR,thus suggesting the presence of mechanisms responsible for receptor downregulation similar to thoseobserved in mammalian cells. J. Exp. Zool. 286:690–698, 2000. © 2000 Wiley-Liss, Inc.

In mammalian cells, the epidermal growth fac-tor (EGF) signal is mediated by the EGF receptor(EGFR), a 170 KD glycoprotein with intrinsic pro-tein kinase activity (Carpenter, ’87). Earlier reports(Hesketh et al., ’85; Carpenter and Cohen, ’90;Moolenar et al., ’93) indicate that ligand bindingleads to receptor dimerization and autophos-phorylation of tyrosine residues: these events arecrucial in triggering the pleiotropic responses toEGF, such as transient increases in cytosolic [Ca2+]and intracellular pH, phosphorylation of a varietyof cytosolic and nuclear substrate proteins, DNAand protein synthesis, eventually leading to cell

*Correspondence to: Laura Canesi, Istituto di Scienze Fisiologiche,Università di Urbino, Loc. Crocicchia, 61029 Urbino (PS), Italy.E-mail: canesi@uniurb.it

Received 11 March 1999; Accepted 28 October 1999

proliferation and/or differentiation. EGF binding toEGFR also induces a negative feedback response(receptor downregulation) by stimulation to recep-tor internalization and targeting of the lysosomalcompartment; in the presence of EGF, most of theEGFR appears to be delivered to lysosomes andrapidly degraded, this attenuating signal transduc-tion (Beguinot et al., ’84; Wells et al., ’90; Wiley etal., ’91; Opresko et al., ’95; Sorkin, ’96).

IMMUNOELECTRON MICROSCOPY OF EGFR IN MYTILUS CELLS 691

Genetic studies have demonstrated the presenceof specific nucleotide sequences closely homologousto EGFR and of their gene products also in inver-tebrate cells (Greenwald, ’85; Livneh et al., ’85).These studies indicated that EGFR belongs to awide family of signalling proteins that have beenconserved in evolution. The presence of both EGF-responsive cells and EGF-like peptides has beensuggested also in marine invertebrates (Odinstovaet al., ’93). Recently, the utilization of cells iso-lated from the tissues of the mussel (Mytilus) hasproven particularly suitable for investigating thepossible effects of heterologous (mammalian)growth factors and the mechanisms involved intyrosine-kinase-mediated signal transduction inmarine invertebrate cells (Canesi et al., ’97). Thisapproach allowed the identification of EGF respon-sive cells in mussel tissues: mammalian EGF in-duced a cytosolic Ca2+ transient and stimulationof DNA synthesis in isolated digestive gland cells.Both responses were abolished by cell pretreat-ment with inhibitors of EGFR tyrosine kinase ac-tivity, suggesting that ligand binding to specificEGF receptors and activation of their intrinsic ty-rosine kinase were crucial to the mussel cell re-sponse to heterologous EGF.

In this work, the presence of EGFR in isolatedmussel digestive gland cells was investigated byimmunoelectron microscopy. Moreover, the pos-sible effects of ligand binding on the subcellulardistribution of EGFR were evaluated.

MATERIALS AND METHODSAnimals

Mussels (Mytilus galloprovincialis Lam.) ob-tained from Cattolica (RN, Italy) were cleanedfrom epibiota and kept in an aquarium at 15°Cin static tanks containing aerated artificial sea-water (1 litre/mussel) for three days. Seawaterwas changed daily.

Cell isolationDigestive glands were excised from three to five

mussels, cut into pieces, and washed in calcium-magnesium-free saline (CMSF; 1100 mOsm, pH7.3, containing 20 mM HEPES buffer, 500 mMNaCl, 12.5 mM KCl, and 5 mM ethylenediamine-tetraacetic acid). Aliquots of about 1 g (w/w) tis-sue were transferred to a flask containing 25 mlof dissociating solution (0.02% pronase w/v inCMFS). The tissue was then dissociated by gentlestirring on a magnetic stirrer for 20 min at 18°C.The dissociated cell suspension was filtered

through 250- and 60-µm diameter nylon filtersand centrifuged at 100g for 10 min at 10°C inconical 50 ml Falcon plastic tubes. The pelletswere resuspended in 25 ml of physiological sa-line (PS; 1100 mOsm, pH 7.3, containing 20 mMHEPES buffer, 436 mM NaCl, 10 mM KCl, 10mM CaCl2, and 53 mM MgSO4) and again centri-fuged at 100 g for 10 min. The final pellets wereresuspended in 10 ml of filtered sterilized Leibo-vitz L-15 medium (supplemented with 350 mMNaCl, 7 mM KCl, 4 mM CaCl2, 8 mM MgSO4 and40 mM MgCl2, and pH 7.3 containing 1% genta-mycin and 1% penicillin), suitably diluted for cellcounting in a Thoma chamber, and kept overnightat 18°C.

EGF exposureAliquots of 1 ml cell suspension (about 106 cells/

ml) were exposed to EGF (from mouse submaxil-lary gland, Calbiochem, CA) at the final concen-tration of 200 nM. Control and EGF treated cellswere kept for different periods of times either at4°C or 18°C (see results and figure captions) un-der gentle stirring and then fixed for electron mi-croscopy.

Immunoelectron microscopyControl and EGF-treated cell suspensions were

mixed 1:1 (v/v) with a fixative solution (8%paraformaldehyde and 1% glutaraldehyde dilutedin 0.5 M phosphate-buffered saline [PBS], pH 7.3).After fixation (1 hr at 4°C), the cell suspensionswere centrifuged at 450g for 10 min. The super-natants were discarded, and the cell pellets werewashed in PBS. Free aldehydes were blocked with0.5 M NH4Cl in PBS for 1 hr at 4°C; then cellswere washed again with PBS and embedded in1.7% (w/v) agar-agar. Samples were dehydratedthrough graded concentrations of ethanol and em-bedded in LRWhite resin. After polymerizationwith UV light, ultrathin sections were cut andplaced on nickel grids coated with a Formvar-car-bon layer and then processed for immunocy-tochemistry. All samples were treated with twodifferent mouse monoclonal anti-EGFR antibod-ies (clone F4 and clone 29.1, Sigma, Buchs, Swit-zerland). Sections were floated for 3 min innormal goat serum (NGS), diluted at 1:100 withPBS, and then incubated for 17 hr at 4°C, withthe primary antibody diluted at 1:50 in a solu-tion containing 0.1% bovine serum albumin(Fluka, Buchs, Switzerland) and 0.05% Tween 20in PBS. After rinsing, sections were floated againin NGS and then reacted for 20 min at room tem-

692 L. CANESI ET AL.

perature with the secondary 12 nm-gold-conju-gated antibody (Jackson ImmunoResearch Labo-ratories, Inc., West Grove, Pennsylvania), dilutedat 1:10 in PBS. Sections were then rinsed, airdried, and finally stained with lead citrate. Ascontrols, some grids were treated with the incu-bation mixture without the primary antibody andthen processed as described above. Grids were ob-served in a Zeiss EM 902 electron microscope op-erating at 80 kV.

Quantitative evaluation of labellingThe presence of EGFR in the immunolabelled

sections from the same experiment was assessedby evaluating the labelling density over selectedcellular compartments (cytoplasmic areas devoidof mitochondria, cell membrane, and lysosomes)in control cells and in cells exposed to EGF for 5,20, and 30 min. The measurements, made by oneoperator who was blind as to the experimentalcondition, were carried out on 20 randomly se-lected electron micrographs (final magnification20,000×) from each sample utilizing a computer-ised image analysis system (Image Pro-Plus forWindows 95). The labelling densities in each cellcompartment were expressed as the number ofgold grains per µm (in the case of the cell mem-brane) or per µm2 (in the case of cytoplasm andlysosomes).

Electrophoresis and Western blottingAliquots of cell suspensions (10 ml containing

approximately 106 cell/ml) were rapidly sedi-mented at 100g, resuspended in 1 ml of ice-coldlysis buffer (50 mM Tris-HCl pH 78, 0.25 M su-crose, 1% [w/v] SDS, 1 µg/ml pepstatine, 10 µg/ml leupeptine, 2 mM sodium orthovanadate, 10mM NaF, 5 mM EDTA, 5 mM NEM (N-ethylm-aleimide), 40 µg/ml PMSF [phenylmethylsul-phonyl fluoride], and 0.1% Nonidet-P40) andsonicated for 45 sec at 50 W. Samples were boiledfor 4 min and then centrifuged for 10 min at14,000g to remove insoluble debris. Supernatantswere mixed 1:1 (v:v) with a sample buffer (0.5 MTris-HCl pH 6.8, 2% SDS, 10% glycerol, and 4%β-mercaptoethanol), and samples (30 µg of pro-tein) were resolved by 8% SDS-polyacrylamide gelelectrophoresis (Laemli, ’70). The gels werestained with Coomassie blue and electroblottedaccording to Towbin et al. (’79). The blots wereprobed with the mouse monoclonal anti-EGFRantibodies (either F4 or 29.1, 1:1000 dilution).Horseradish-peroxidase-conjugated goat anti-mouse IgG (1:3000 dilution) was used as a sec-

ondary antibody. Immune complexes were visu-alized using an enhanced chemioluminescenceWestern blotting analysis system (Amersham-Pharmacia Corp.) following the manufacturer’sspecifications.

Statistical analysisThe differences in labelling density in each cel-

lular compartment between the different experi-mental conditions were evaluated utilizing theMann-Whitney U test. Statistical significance wasset at P ≤ .05.

RESULTSCell isolation

Mussel digestive gland consists of numerousblindly ending tubules whose epithelium is com-posed of two mature cell types, the digestive cellsand the basophilic cells. Digestive cells, whosemain function is intracellular digestion of food,contain a well-developed lysosomal vacuolar sys-tem where intracellular digestion takes place,mainly consisting of large heterolysosomes, smalllysosomes, and morphologically heterogeneousresidual bodies. Basophilic cells are less numer-ous, with abundant rough endoplasmic reticu-lum and Golgi typical of secretive cells, but theirfunction is still unclear (Langton, ’75; Morton,’83). A detailed ultrastructural characterizationof isolated mussel digestive gland cells has beenrecently reported (Robledo and Cajaraville, ’97).Our microscopic observations indicate that theisolation procedure utilized in the present study(see Materials and Methods) gives a cell prepa-ration on which most cells showed morphologi-cal features typical of digestive cells, whereasonly few basophilic cells were identified (about5% of the total cell population).

EGFR immunoreactivity indigestive gland cells

Immunocytochemical analysis demonstrated thepresence of EGFR in isolated mussel digestivegland cells (Fig. 1): the same immunolabeling pat-tern was observed separately utilizing two differ-ent anti-EGFR monoclonal antibodies directedagainst either intracellular or extracellular epi-topes. In control digestive cells, EGFR immunore-activity was mainly distributed in the cytoplasm,although it also occurred along the cell membrane(Fig. 1a). The cytoplasmic signal was mostly as-sociated with membrane structures such as thewell-developed vacuolar system (Fig. 1a, inset), theRER cisternae (Fig. 1b), and the Golgi complex

IMMUNOELECTRON MICROSCOPY OF EGFR IN MYTILUS CELLS 693

(Fig. 1c). Large heterolysosomes were also weaklylabelled (Fig. 1a), whereas nuclei and lysosomeswere devoid of gold grains. Within each cell prepa-ration, different digestive cells showed different

labelling densities: in general, the larger the celland the greater the number of heterolysosomes,the stronger the signal.

The few basophil cells showed a clearly differ-

Fig. 1. Control mussel digestive cells immunolabelled withanti-EGFR antibodies. a. The labelling is distributed throughthe cytoplasm, mainly associated with membrane structures(arrows). A detail of labelled membranes (thin arrows) isshown in the inset. A weak signal also occurs along the cell

membrane (arrowheads) and in the heterolysosomes (HL). b.RER cisternae show a weak but specific labelling. c. A Golgicomplex (arrow) showing a few gold grains in peripheral cis-ternae (arrowheads). Bars = 0.5 µm.

694 L. CANESI ET AL.

ent labeling pattern, with weak cytoplasmic la-belling (not shown).

Electrophoresis and Western blotting withanti-EGFR antibodies

To confirm the specificity of the two anti-EGFRantibodies utilized in immunoelectron microscopy,immunoblotting of cell extracts fractionated by 8%SDS-polyacrylamide gel electrophoresis was ap-plied. The results (Fig. 2) indicate that only a ma-jor protein band, corresponding to the MW ofabout 170 KD, was recognized by both the F4 and29.1 anti-EGFR antibodies.

Effects of EGF exposure on subcellularEGFR distribution

Microscopic observationsEGF exposure resulted in time- and tempera-

ture-dependent modifications in the labelling pat-tern of digestive cells, whereas basophil cells didnot undergo any change. In digestive cells exposedto EGF and kept for 0–15 min at 4°C, the distri-bution of EGF receptors was similar to that ob-served in control cells, even though, after about15 min of exposure, a few cells displayed an in-creased labelling along the cell membrane (notshown). On the other hand, when EGF-treatedcells were kept at 18°C, time-dependent changesin the labelling pattern were observed: in diges-tive cells exposed to EGF for 5 and 10 min, anincrease in labelling intensity along the cell sur-face and a corresponding decrease in the cytoplas-mic signal were observed (Fig. 3a). Moreover, after10 min from EGF addition, some cells showed astrong labelling within lysosomes (Fig. 3b). After20 min, labelling was mainly localized within ly-

Fig. 2. Western blotting of digestive gland cell extractswith anti-EGFR antibodies. Cell lysates, prepared as describedin Materials and Methods, were subjected to 8% SDS-poly-acrylamide gel electrophoresis and immunoblotting with theanti-EGFR antibodies. Lane A, clone F4; lane B, clone 29.1.

Fig. 3. Digestive cells exposed to EGF (200 nM) for dif-ferent periods of time at 18°C. Immunolabelling with anti-EGFR antibodies. a. Digestive cells exposed to EGF for 5 min:the cell surface shows a strong signal (arrows), whereas thecytoplasm appears weakly labelled. b. Digestive cells exposed

to EGF for 10 min: the signal is concentrated within smalllysosomes (stars). c. Digestive cells exposed to EGF for 20min: a small lysosome shows a strong signal (arrow), whereasthe cell surface is devoid of gold grains (arrowhead). Bars =0.5 µm.

IMMUNOELECTRON MICROSCOPY OF EGFR IN MYTILUS CELLS 695

sosomes in most digestive cells, whereas the cellsurface and cytoplasm contained only a few goldgrains (Fig. 3c). Finally, after 30 min of exposureto EGF, digestive cells showed a labelling patternsimilar to that observed in control cells except forthe lysosomal signal, which seemed still strongerthan in controls (not shown).

Control digestive cells did not show any time-or temperature-dependent change in receptor dis-tribution. EGF-treated digestive cells showed avariability in labelling densities similar to thatobserved in control cells, the larger ones display-ing higher amounts of EGFR: nuclei were alwaysdevoid of gold grains, and large heterolysosomesalways showed a constant, weak labelling, as incontrol cells (not shown). In all immunocytochemi-cal experiments, control samples incubated in theabsence of anti-EGFR antibodies showed only anegligible signal (Fig. 4).Quantitative analysis of EGFR distribution

On the basis of the electron microscopic obser-vations, the quantitative evaluation of EGFR la-belling was carried out on those cell compartmentsthat either showed significant labelling in controlcells or seemed to be mainly involved in EGFR

redistribution in cells exposed to EGF, i.e., cyto-plasm, the cell membrane, and lysosomes. Theresults of the quantitative evaluation of the anti-EGFR labelling on these three cell compartmentsin control cells and in cells exposed to EGF for 5,20, and 30 min are shown, respectively, in Fig-ure 5a–c.

In control cells, the EGF receptors were mainlydistributed in the cytoplasm, whereas the cellmembrane and lysosomes showed weak labelling;such an intracellular distribution did not show anychange with time. After 5 min of exposure to EGF,the EGFR density showed a significant decreasein the cytoplasm (Fig. 5a), whereas a drastic in-crease was observed on the cell membrane (Fig.5b); lysosomes almost showed no labelling (Fig.5c). After 20 min of exposure, the density of EGFRin the cytoplasm remained lower than in controlcells (Fig. 5a), whereas it returned to control val-ues in the cell membrane (Fig. 5b). On the otherhand, a remarkable increase in the amount of la-belling was observed in the lysosomes (Fig. 5c).Finally, after 30 min of EGF exposure, the den-sity of EGFR was not significantly different fromthat of control cells both in the cytoplasm and inthe cell membrane (Fig. 5a,b). The lysosomal la-belling decreased, but it was still higher than incontrol cells (Fig. 5c).

The background levels (0.05 ± 0.01 grains/µm2),evaluated in the field outside the cells, were sig-nificantly lower (P ≤ 0.05) than those found in anyother of the cellular compartments considered.

DISCUSSIONImmunoelectron microscopy utilizing mono-

clonal antibodies directed against mammalianEGFR demonstrates the presence of specific EGFreceptors in mussel digestive gland cells. This wasconfirmed by Western blotting of digestive glandcell extracts. The fact that both the monoclonalantibodies directed against either an external car-bohydrate (clone 29.1, Sigma) or a cytoplasmicepitope (residues 985–996) (clone F4, Sigma) ofthe mammalian EGFR were able to recognizeEGFR-like molecules in mussel cells utilizing bothtechniques confirms, also in these organisms, theevolutionary conservation of the receptor molecule,or at least of the epitopes recognized by the twoantibodies.

Immunoelectron microscopy indicated that thetwo main cell types in mussel digestive glandshowed a different expression of EGFR, with di-gestive cells always showing strong immunoreac-tivity, whereas immunolabelling was weak in

Fig. 4. Control sample subjected to the immunocytochemi-cal procedure in the absence of the anti-EGFR antibody. Thedigestive cell shows only a negligible labelling (arrow), thusdemonstrating the absence of nonspecific binding by the sec-ondary gold-conjugated antibody. HL, heterolysosome. Bar =0.5 µm.

696 L. CANESI ET AL.

basophil cells. An even weaker labelling wasshown by circulating hemocytes, whereas it wastotally absent in cerebral ganglia (unpublished re-sults). Taken together, these observations indicatethat EGFR-immunoreactivity is a characteristicfeature of digestive gland cells.

Control digestive cells showed EGFR immunore-

activity in the cytoplasm, mostly associated withmembrane structures; labelling was also observedalong the cell membrane. Also in mammalian cells,the existence of different EGFR pools has beendemonstrated, depending on the state of receptorphosphorylation and/or oligomerization and asso-ciation with the cytoskeleton (Berkers et al., ’91;Sorokin et al., ’94); in A-431 cells, only a minorfraction (15%) of the total EGFR is located at theplasma membrane (Carpentier et al., ’86). In theabsence of ligand, unoccupied receptors tend tobe nonspecifically internalized along with themembrane proteins and continuously recycled bythe physiological endocytic process (Felder et al.,’90; Wiley, ’91; French et al., ’94; Lamaze andSchmidt, ’95; Sorkin, ’96).

In the control digestive cells of mussels, the highproportion of EGFRs associated with cytoplasmicstructures may be related in part to their asso-ciation to cytoskeletal structures, in part to theintense endocytic processes which physiologicallyoccur in this cell type, and, therefore, to the pre-sumably high membrane turnover rate. In thissituation, a large fraction of intracellular recep-tors would not undergo activation by ligand bind-ing because they are not accessible to EGF. Thiscould partly account for the results previously ob-tained, indicating that mussel digestive cells wereresponsive to concentrations of EGF higher thanthose routinely utilized in mammalian systems(Canesi et al., ’97).

In mussel digestive cells, but not in basophilcells, the addition of heterologous (mammalian)EGF resulted in significant modification of theimmunolabelling pattern. In studies on receptor-mediated endocytosis in mammalian cells, low-ering the temperature has been widely utilizedto inhibit the endocytic process (Beguinot et al.,’84; Lai et al., ’89; Sorkin, ’96). Therefore, in mus-sel cells, the distributions of the EGFR afterEGF addition at 4°C and at 18°C (the tempera-ture at which mussels were usually maintainedin the aquarium), were compared. As previouslydemonstrated in mammalian cells (Beguinot etal., ’84; Miller et al., ’86; Felder et al., ’90;Honnegger et al., ’90; Sorkin et al., ’91; Sorkin,’96; Wang et al., ’96), the observed changes inEGFR distribution following ligand binding weretemperature dependent and related to the timeof exposure to EGF.

In digestive cells exposed to EGF for up to 30min at 4°C, no significant EGFR redistributionwas observed in response to ligand addition. Onthe other hand, when cells were kept at 18°C, EGF

Fig. 5. Histograms showing the labelling densities on cy-toplasm, cell membrane, and lysosomes. Results are the meanof 20 observations ± SE. In each bar, values identified bycommon symbols (x, *, #) are not significantly different fromone another.

IMMUNOELECTRON MICROSCOPY OF EGFR IN MYTILUS CELLS 697

addition resulted in significant changes in the la-belling pattern that were related to the time ofexposure. After 5 min, EGF induced receptor re-cruitment to the plasma membrane from cytoplas-mic structures; moreover, exposure to EGF for 20min lead to a net increase in labelling within smalllysosomal vacuoles, indicating that ligand-inducedreceptor internalization results in increased EGFRlysosomal targeting for degradation. Althoughslight differences in time-dependent changes wereobserved among individual cells, the effect of EGFwas clearly observed in all digestive cells, as dem-onstrated by the quantitative evaluation of theEGFR labelling density. These results thereforeindicate that, in the cells of marine invertebrateslike in mammalian cells, the binding of EGF in-duces a negative feedback response by stimulat-ing receptor degradation.

Finally, it must be mentioned that, althoughEGF exposure induced lysosomal degradation ofthe EGFR, in some mammalian cell types the re-ceptor distribution could subsequently return tostarting levels, indicating that receptor distribu-tion could subsequently return to starting levels,indicating receptor recycling also in the presenceof EGF (Lai et al., ’89; French et al., ’94). More-over, receptor recycling was inversely related tothe EGF concentration and did not occur in thepresence of excess EGF. The results of the presentwork, indicating that after 30 min from EGF ad-dition the EGFR labelling pattern was similar tothat of control cells, indicate a similar situationin mussel digestive cells. In fact, these resultswere obtained in cells exposed to a concentrationof EGF lower than that eliciting maximal re-sponses in isolated mussel digestive gland cells(Canesi et al., ’97). Further studies will clarify thepossibility of EGF dependence of receptor recy-cling in mussel cells.

The apparent variability in EGFR expressionin both control and EGF-exposed cells deservessome explanation. The variability in the amountof labelling observed among different digestivecells seemed to be correlated to the cell size andthe number of intracellular vacuoles. In musseldigestive glands, the heterogeneity of the diges-tive cell population is related to different func-tional stages in the digestive process (Langston,’75; Morton, ’83). Depending on feeding and di-gestion, digestive tubules go through cytologicalchanges: food particles are engulfed through en-docytosis by digestive cells and are degraded bylysosomal enzymes. During digestion, many sec-ondary lysosomes are formed, and the apical zone

undergoes a disintegration phase. The cyclicbreakdown and replacement of digestive cells re-sults in variations in digestive epithelium volume.

The fact that the more actively digesting cellsapparently showed a greater receptor expressionsuggests that EGF-related peptides may play arole in the regulation of digestive cell function.EGF has been shown to regulate protein turnoverin mammalian kidney cells (Singh et al., ’95), notonly by stimulating protein synthesis but also byinducing the expression of lysosomal proteases.Similarly, EGF might play a role in the regula-tion of the activity of lysosomal proteases also inmussel digestive cells. Interestingly, in these cells,the activity of lysosomal enzymes is related notonly to protein digestion but also to the regula-tion of free amino acid concentration, which is in-volved in the adaptation to osmotic changes in theenvironment (Livingstone et al., ’79). Therefore,the possible role of hormonal stimuli in proteinmetabolism in mussel digestive cells requires fur-ther investigation.

ACKNOWLEDGMENTSWe are particularly grateful to Dr. F. Mannello

for the statistical evaluation of the morphometricaldata. We also thank Dr. Marcheggiani and Dr. Bettifor their excellent technical assistance.

LITERATURE CITEDBeguinot L, Lyall RM, Wilingham MC, Pastan I. 1984. Down-

regulation of the epidermal growth factor receptor in KBcells is due to receptor internalization and subsequent deg-radation in lysosomes. Proc Natl Acad Sci 81:2384–2388.

Berkers JAM, van Bergen en Henegouwen PMP, Boonstra J.1991. Three classes of epidermal growth factor receptorson HeLa cells. J Biol Chem 266:922–927.

Canesi L, Ciacci C, Orunesu M, Gallo G. 1997. Effects of epi-dermal growth factor on isolated digestive gland cells ofmussels (Mytilus galloprovincialis Lam.). Gen CompEndocrinol 107:221–228.

Carpenter G. 1987. Receptors for epidermal growth factor andother related polypeptide mitogens. Annu Rev Biochem56:881–914.

Carpenter G, Cohen S. 1990. The epidermal growth factor. JBiol Chem 14:7709–7712.

Carpentier JL, Rees AR, Gregoriou M, Kris R, SchlessingerJ, Orci L. 1986. Subcellular distribution of the external andinternal domains of the EGF receptors in A-431 cells. ExpCell Res 166:312–326.

Felder S, Miller K, Moeheren G, Ullrich A, Schlessinger J,Hopkins CR. 1990. Kinase activity controls the sorting ofthe epidermal growth factor receptor within the multivesi-cular body. Cell 61:623–634.

French AR, Sudlow GP, Wiley HS, Lauffenburger DA. 1994.Postendocytic trafficking of epidermal growth factor com-plexes is mediated through saturable and specific endosomalinteractions. J Biol Chem 269:15749–15755.

698 L. CANESI ET AL.

Greenwald I. 1985. lin-12, a nematode homeotic gene, is ho-mologous to a set of mammalian proteins that includes epi-dermal growth factor. Cell 43:583–590.

Hesketh TR, Moore JP, Morris JDH, Taylor MV, Rogers J,Smith GA, Metcalfer JC. 1985. A common sequence of cal-cium and pH signals in the mitogenic stimulation ofeukariotic cells. Nature (London) 313:481–484.

Honnegger AM, Schmidt A, Ullrich A, Schlessinger J. 1990.Separate endocytic pathways of kinase-defective and -ac-tive EGF receptor mutants expressed in the same cells. JCell Biol 110:1541–1548.

Laemli UK. 1970. Cleavage of structural proteins during theassembly of the head bacteriophage T4. Nature 227:680–685.

Lai WH, Cameron PH, Wada I, Doherty JJ II, Kay DG, PosnerBI, Bergeron JJ. 1989. Ligand-mediated internalization, re-cycling, and down-regulation of the epidermal growth fac-tor receptor in vivo. J Cell Biol 109:2741–2749.

Lamaze C, Schmidt SL. 1995. Recruitment of epidermalgrowth factor receptors into coated pits requires their acti-vated tyrosine kinase. J Cell Biol 129:47–54.

Langton RW. 1975. Synchrony in the digestive diverticulataof Mytilus edulis L. J Mar Biol Ass UK 55:221–229.

Livingstone DR, Widdows J, Fieth P. 1979. Aspects of nitro-gen metabolism of the common mussel Mytilus edulis: ad-aptation to abrupt and fluctuating changes in salinity. MarBiol 53:41–55.

Livneh E, Glazer L, Segal D, Schlessinger J, Ben-Zion Shilo.1985. The Drosophila EGF receptor gene homolog: conser-vation of both hormone binding and kinase domains. Cell40:599–607.

Miller K, Beardmore J, Kanety H, Schlessinger J, HopkinsCR. 1986. Localization of the epidermal growth factor (EGF)receptor within the endosome of EGF-stimulated epidermoidcarcinoma (A431) cells. J Cell Biol 102:500–509.

Moolenar WH, Aerts RG, Tertoolen LGJ, de Laat SW. 1986.The epidermal growth factor-induced calcium signal in A431cells. J Biol Chem 261:279–284.

Morton B. 1983. Feeding and digestion in Mollusca. In:Saleuddin ASM, Wilbur KM, editors. vol. 5, Physiology, part2. London: Academic Press. p 65–147.

Odinstova NA, Nesterov AM, Korchagina DA. 1993. A growth

factor from tissues of the mussel Mytilus edulis. CompBiochem Physiol 105A:667–671.

Opresko LK, Chang CP, Will BH, Burke PM, Gill GN, WileyHS. 1995. Endocytosis and lysosomal targeting of epider-mal growth factor receptors are mediated by distinct se-quences independent of the tyrosine kinase domain. J BiolChem 270:4325–4333.

Robledo Y, Cajaraville MP. 1997. Isolation and morphofunc-tional characterization of mussel digestive gland cells invitro. Eur J Cell Biol 72:362–369.

Singh AK, Song RH, Frankfater AF, Leehey DJ. 1995. Physi-ological and biochemical response of glomerular epithelialcells to exogenous epidermal growth factor. Exp Nephrol3:355–361.

Sorkin A. 1996. Receptor-mediated endocytosis of growth fac-tors. In: Heldin CH, Purton M, editors. London: Chapmanand Hall. p 109–126.

Sorkin A, Krolenko S, Kudrjiaceva N, Lazebnik J, TeslenkoL, Soderquist AM, Nikolsky N. 1991. Recycling of epider-mal growth factor-receptor complexes in A431 cells: identi-fication of dual pathways. J Cell Biol 112:55–63.

Sorokin A, Lemmon M, Ullrich A, Schlessinger J. 1994. Sta-bilization of an active dimeric form of the epidermal growthfactor receptor by introduction of an inter-receptor disul-fide bond. J Biol Chem 269:9752–9759.

Towbin H, Stahelin T, Gordon J. 1979. Electrophoretic trans-fer of proteins from polyacrylamide gels to nitrocellulosesheets: procedure and some applications. Proc Natl AcadSci USA 76:4350–4354.

Wang Z, Tung PS, Moran MF. 1996. Association of P120 RasGap with endocytic components and colocalization with epi-dermal growth factor (EGF) receptor in response to EGFstimulation. Cell growth and differentiation 7:123–133.

Wells A, Welsh JB, Lazar CS, Wiley HS, Gill GN, RosenfeldMG. 1990. Ligand-induced transformation by a non-inter-nalizing EGF receptor. Science 247:962–964.

Wiley HS, Herbst JJ, Walsh BJ, Lauffenburger DA, Rosen-feld MG, Gill GN. 1991. The role of tyrosine kinase activ-ity in endocytosis, compartmentation, and down-regulationof the epidermal growth factor receptor. J Biol Chem 266:11083–11094.