Cadmium-binding proteins in the mussel, Mytilus edulis.

Inducible cadmium-binding proteins (Cd-BP) in the mussel, Mytilus edulis, were resolved into two molecular weight components, designated Cd-BP10 and Cd-BP20, by gel-permeation chromatography on Sephadex G-75. Each of these two molecular weight components were further resolved into four subcomponents by DEAE-ion-exchange chromatography. All eight subcomponents bound cadmium and exhibited significant UV absorption at 254 nm and little absorption at 280 nm. Each subcomponent was purified and subjected to amino acid composition analysis. Two classes were identified, one having higher cysteine (23.9-26.6 mole-%) and lower glutamic acid contents compared to the other class (11.6-18.2 mole-% cysteine). All subcomponents have a relatively high glycine content (approximately 15 mole-%) relative to mammalian metallothioneins (approximately 8 mole-%). Although the Cd-BP20 have apparent molecular weights almost twice the Cd-BP10, the exact molecular relationship between these binding proteins is not known.


Introduction
In recent years several investigators have reported the occurrence of metal binding proteins (BP) in the mussel Mytilus edulis as well as Myutilus galloprovincialis.
Much of this research was stimulated by the original paper of Noel-Lambot which appeared in 1976 (1). This pioneering work on mussels clearly defined some of the unique properties of mussel BPs. First, when induced in the laboratory by exposure to cadmium in sea water, there are two BP components resolvable by gel-permeation chromatography on Sephadex G-75 columns. One component has an apparent molecular weight similar to that of mammalian metallothioneins (MT) while the second has a significantly larger apparent molecular weight. Secondly, the Cd-BP obtained had a high percentage of cysteine and an absorption spectrum typical of mammalian MT. Third, the proteins induced in the laboratory contained very low levels of zinc and copper. Soon after the Noel-Lambot report, Talbot and Magee (2) demonstrated that similar proteins were present in mussels exposed to contaminating metals in the natural environment. More recent studies have verified the presence of BPs in Mytilus edulis following exposure to cadmium (3)(4)(5) and mercury (6). In a related species, Mytilus galloprovincialis, BPs have also been demonstrated following exposure to both cadmium (7) and copper (8)(9)(10).
In spite of the interest in the mussel metal BPs, relatively little research has been directed toward a careful characterization of the proteins themselves. George et al. (3) isolated three subcomponents of the higher molecular weight BP (apparent MW = 25,000 + 5000) on DEAE-*The Johns Hopkins University, 615 North Wolfe Street, Baltimore, MD 21205. cellulose anion-exchange chromatography. Amino acid composition for these proteins (Table 1) indicated a significantly lower value for the cysteine content than mammalian MT. The investigators suggest that the BPs of higher molecular weight may be dimers of the lower molecular weight component, although no characterization of the smaller BPs is given. Frankenne et al. (11) also attempted to characterize the mussel BPs. Unfortunately, they did not resolve the two molecular weight BP components on Ultrogel AcA 54, and thus the material used for amino acid composition studies was a mixture of various components as indicated by their PAGE results. In spite of this, they obtained a very high cysteine content (25.5 mole-%) ( Table 1). The only other study characterizing mussel BPs is that of Roesijadi and Hall (6) following induction with mercury These investigators were able to obtain a protein(s) which exhibited multiple bands on PAGE and had an amino acid composition with relatively low cysteine content (7.8%) and a large molecular weight (14,400).
The data reported here are the results of an attempt to further characterize the nature of the cadmium BPs in the mussel, Mytilus edulis, following induction by cadmium under laboratory conditions. In this study both molecular weight classes of induced Cd-BPs are explored.

Materials and Methods
Adult mussels, Mytilus edulis, were obtained from either the North Esk or Ythan estuaries (Scotland) and maintained in sea water aquaria at the Institute of Marine Biochemistry, Aberdeen, Scotland. Mussels were exposed to 100 ,ug Cd2+/L at 100C for 3 to 4 months. Water  (3,6,11).
was changed three times per week and cadmium dosing extracted with 100 mL of 0.5 M Tris, pH 7.5, for 2 hr. was repeated at each water change. Following exposure, The mixture was refiltered, the DEAE washed with an mussels were shucked in batches of 30 to 50 organisms additional 50 mL of extraction buffer and the filtrate and and the soft tissues were dissected to exclude adductor wash were pooled for the next step. The pH was readmuscle and foot. Tissues (50-100 g) were homogenized in justed to 8.5 with NaOH. a volume of distilled water equal to the tissue weight. The DEAE extract (approximately 130 mL) was ap-The homogenate was centrifuged at 79,000g for 120 min plied to a Sephadex G-75 column (5 x 90 cm) and eluted and the resulting supernatant filtered and collected for with 2 L of 0.02 M Tris, pH 8.5, at a flow rate of 120 further processing. mL/hr. Fractions of 10 mL volume were collected and Cadmium binding proteins were purified by the follow-analyzed for cadmium by atomic absorption spectrophoing procedure. Mussel cytosol was diluted 1:20 with distometry without sample dilution. The major cadmium tilled water and the pH adjusted to 8.5 with 0.1 N NaOH. binding component was resolved into two overlapping DEAE-cellulose (Whatman DE52) was added (100 mL peaks. The fractions constituting the leading portion of of swollen gel equilibrated with 0.02 M Tris, pH 8.5). the higher molecular weight peak and the trailing portion The mixture was stirred at low speed for 2 hr with N2 of the lower molecular weight peak were each collected constantly bubbled through the solution. The DEAE-and designated Cd-BP10 and Cd-BP20, respectively (the cellulose was recovered by filtration and washed with 0.02 subscripts refer to crude estimates of molecular weights). M Tris, pH 8.5, on the filter, transferred to a beaker and The lower molecular weight component, Cd-BP10, was   Cd-BP20(I)-Cd-BP20(IV).
Each partially purified subcomponent was further purified. Only the central fractions of each peak were pooled and the pooled sample rechromatographed on DEAE under the conditions as described above. In all cases the rechromatography profiles now gave only one symmetrical peak with little evidence of contamination by other subcomponents. Fractions with cadmium concentration greater than 0.5 p,g/mL were pooled. The individual subcomponents were desalted on a Sephadex G-50 column, frozen and lyophilized.
Amino acid composition of each subfraction was determined after performic acid oxidation on a Durrum D-500 Analyzer in Professor J. Kagi's laboratory (University of Zurich). Values for threonine and serine were corrected for 5 and 12.5% loss during hydrolysis, respectively.

Results
In a typical preparation the cadmium, zinc and copper concentrations in fresh mussel tissues used for the extraction of Cd-BPs were 31.3, 12.8, and 1.4 ,ug/g wet weight, respectively. Almost 60% of the cytosolic cadmium is associated with Cd-BP20, the larger Cd-BP, and only 25% with Cd-BP10. The remaining cytosolic cadmium was associated with high molecular weight macromolecules (MW> 50,000). Cadmium was the major metal associated with both BPs. Smaller amounts of zinc were bound to the BPs and only traces of copper were present.
Typical chromatographic profiles for Cd-BP10 and Cd-BP20 on DEAE are given in Figures 1 and 2, respectively. Four cadmium peaks are observed in each profile. Each peak exhibits significant UV absorption at 254 nm and relatively low absorption at 280 nm. The amino acid compositions of each of these subcomponents are given in Table 2. Subcomponents 10I and 10III have high cysteine and low glutamic acid residue contents compared to 10II and 10IV All subcomponents have a relatively high glycine content (approximately 15 residue-%) compared to mammalian metallothioneins (8-10 residue-%). A similar pattern is observed for the higher molecular weight proteins (Cd-BP20). Subcomponents 20I and 201V have high cysteine and lower glutamic acid residue contents than 20II and 20III. All subcomponents have high glycine content. Subcomponents 10II and 20II both have higher serine content than other subcomponents and seem to be quite similar in composition.
An attempt to reduce Cd-BP20 (III) by the procedure described by Suzuki and Yamamura (12) to split metallothionein dimers gave inconclusive results. If reduction occurred, only a very small percentage of the higher molecular weight component was reduced to a smaller component.

Discussion
Careful anion exchange chromatography of mussel Cd-BPs reveal eight subcomponents, four Cd-BP10 and four Cd-BP20. Subclasses of these proteins appear to exist based on cysteine content. The high cysteine content (23.9-26.6 residue -%) subcomponents (10I, 10III, 201 and 20IV) have generally similar amino acid compositions. The low cysteine content (11.6-18.2 residue -%) subcomponents are further subdivided by their serine contents into low cysteine-low serine subcomponents (1OIV and 20III) and low cysteine-high serine subcomponents (10II and 20II). The functional role of the multiple molecular weight species and charge forms of the BP is not known.
The amino acid composition of one of the high cysteine subcomponents [Cd-BP1o (I)] is compared to several other cadmium/zinc binding proteins in Table 3. The oyster protein reported in this table was obtained in collaboration with Dr. Marius Brouwer (Duke University Marine Laboratory). The very close correspondence in amino acid composition is apparent and may indicate similar proteins in these two mollusk species. However, it should be noted that DEAE-anion-exchange chromatography indicated there was only one oyster Cd-BP component.
The relationship between the Cd-BP10 and Cd-BP20 is an important question to be resolved. If the Cd-BP20 are dimers of the Cd-BP10, then why is dimer formation favored (under the experimental conditions identical to those employed in this study rat MTs never form dimers). If on the other hand the Cd-BP20 are in fact single polypeptide chains, then the origin and function of these proteins must be investigated.
In conclusion, it appears that there are Cd-BPs present in mollusks which differ significantly from mammalian MTs. The relationship between the structure of these proteins and their function in these invertebrates will no doubt further our understanding of the biological role of metallothioneins.
The author would like to thank Drs. Thomas Coombs, Steven George and Julian Overnell of the Institute of Marine Biochemistry, Aberdeen, Scotland, for their collaboration in this project. In addition, the assistance of Mr. Peter Young in the preparation of purified mussel subcomponents and Mrs. Marletta Regner in the preparation of this manuscript is greatfully recognized.