Environmental Health Perspectives Vol. 65, pp. 77-85, 1986 Characterization of a Highly Negative and Labile Binding Protein Induced in Euglena gracilis by Cadmium by David J. Gingrich,* Daniel N. Weber,t C. Frank Shaw,*? Justine S. Garvey,* and David H. Petering* The physiochemical properties and physiological significance of the cadmium-binding protein (CdBP) of the algae Euglena gracilis have been studied. Following in vivo exposure of cells to 0.4 or 1.3 ,ug/mL of Cd2", all the cytosolic Cd is bound to high molecular weight species. At 4.7 ,ug/mL, appreciable CdBP has formed in cells grown under illumination or in the dark. An analogous ZnBP could not be isolated from control or Zn-exposed (20 ,ug/mL) cells, but zinc and a trace of copper were bound to the CdBP when 2-mercaptoethanol (2-ME) is added to the homogenates of Cd-treated cells and the buffers used during isolation. The large pool of very low molecular weight zinc species previously reported is increased when cells are exposed to high cadmium levels. Two distinct species, BP-1 and BP-2 are resolved by ion-exchange chromatography on DEAE-Sephadex. Unusually high conductivities (25 and 40 mSiemen) are required to displace them, indicating that they are very negatively charged proteins at pH 8.6. The pH for half-titration of bound Cd2' is between 5 and 6. EDTA (0.4 M) and the CdBP isolated by gel-exclusion chromatography react biphasically with pseudo-first-order rate constants of 4 ? 3 x 10-4 sec' and 7 ? 2 x 10' sec'. Neither form of the CdBP cross-reacts with antibodies to rat liver metallothionein (MT) antibodies. The structural, chemical, and functional differences between the Euglena CdBPs and mammalian MTs are discussed. When cells are exposed to high levels of Cu (5 or 10 ,ug/mL), a CuBP is induced, and the very low molecular weight zinc band is depleted. Their copper-rich cytosol has a green-black color which is lost upon oxidation. Introduction such as fish (6), crustacea (7), and various microorga- The reactions of metals with cells are important in understanding the essential metal metabolism, environ- mental metal intoxication, and the mechanism of action of metallodrugs. Zinc and cadmium metabolism in Eu- glena gracilis have been studied extensively as a model for human and mammalian cells (1-4). These cells show characteristic zinc and cadmium antagonism which is common to most species. However, a recent examination of the biochemistry of zinc and cadmium in Euglena gra- cilis found significant differences from that of mammalian cells such as the Ehrlich ascites tumor cell (5). Most striking was the finding that up to 80% of the cytosolic zinc was present as a very low molecular weight species. In mammalian cells, metallothionein-bound zinc is the physiologically active form, whereas in Euglena gracilis, the very low molecular weight form of zinc seems to be related to the zinc status of the cells (5). Metal-inducible binding proteins in lower organisms *Department of Chemistry, University of Wisconsin-Milwaukee, Mil- waukee, WI 53201. tDepartment of Biology, Syracuse University, Syracuse, NY 13210. tDepartment of Pharmacology and Toxicology, Medical College of Wisconsin, Milwaukee, WI 53226. ?Author to whom correspondence should be addressed. nisms (8,9) vary greatly in chemical and physical prop- erties among themselves and in comparison to mammal- ian metallothioneins (MTs). In contrast to the multiplicity of invertebrate binding proteins (BPs), MTs from numerous mammalian species have a highly con- served amino acid sequence and similar properties (10). Examination of the structure and chemical reactivity of BPs from lower organisms will elucidate their functions and mechanisms of reaction in metal metabolism, and by contrast will help to clarify the structure-function re- lationships of MTs. The Cd-inducible BP of Euglena gra- cilis was found to have a smaller apparent molecular weight and more negative charge than MTs, and to re- semble MTs in having a high cysteine content and low aromatic amino acid content. In this paper we report further investigations of the chemical properties and physiological significance of the CdBP of Euglena gracilis and some results on exposure of these cells to copper. Methods and Materials Media All cultures were maintained by using the modified medium of Hutner et al., with MnCl2 substituted for GINGRICH ET AL. MnSO4 and glutamic acid for ammonium glutamate as previously described (5). The media were sterilized by autoclaving after the pH was adjusted to 3.5 using 6 M KOH. Zinc-sufficient (Zn +) and zinc-deficient (Zn -) me- dia were identical except for the withholding of zinc from the (Zn -) medium. (Zn +) medium contained 1 jig Zn/ mL and (Zn-) medium less than 0.01 jig Zn/mL as measured by atomic absorption. Cadmium, zinc, or cop- per levels above normal media levels wre achieved by the addition of aqueous solutions of CdC12, ZnC12, or CuC12 either before or after autoclaving. Cultures Euglena gracilis (Klebs), "z" strain, #753 was pur- chased from the Culture Collection of Algae at the Uni- versity of Texas at Austin. Unless otherwise noted, individual experiments were begun using inoculations of stock culture cells in the log phase of growth. In experiments requiring multiple cul- tures, one large culture was made and then divided into 250 mL cultures in cotton-plugged 500 mL Erlenmeyer flasks and placed in a shaker bath maintained at 26 to 27?C under a 20-W GE F20T12-CW fluorescent tube ap- proximately 40 cm overhead. Cultures grown in the ab- sence of light ("dark-grown") were set up identically and wrapped in aluminum foil to exclude light. Cell counts were obtained by using an American Op- tical Series 10 Microstar microscope at 450 x magnifi- cation with an American Optical Spencer Bright Line Phase Hemacytometer #1475 with Neubauer ruling. Cells Exposed to Elevated Copper Levels The effects of elevated medium copper levels upon cy- tosolic copper were investigated by the addition of copper as aqueous CuCl2. The normal copper level in the medium is 0.064 ,ug/mL (1 x 10-6 M). Inoculations to 1 x 104 cells/mL were made from stock cultures and copper was then added at 10 ,ig/mL (1.6 x 10-4 M) or 5 pLg/mL (7.8 x 10-5 M). The cells were harvested after 4 or 10 days of growth and the cytosol chromatographed over a Seph- adex G-75 column. Cytosol Preparation Cells were typically harvested by centrifugation at 3000g for 5 min. This was followed by two or three wash- ings which consisted of resuspending the pelleted cells in a 10 mL of 0.15 M KCI per gram of cells and recen- trifuging at 3000g. The cells were then resuspended in 0.15 M KC1, 0.05 M Tris buffer, pH 8.6, at room tem- perature, and sonicated in a 7 mL tube in an ice-water bath and a Sonifier Disruptor 200 for 2 min at 60% duty and 75 W of power. A 5-mL portion of homogenization medium was used per gram of cells in a 7-mL tube. The preparation was kept between 0 and 4?C after sonication and, unless otherwise noted, 5 to 10 mM 2-mercaptoeth- anol (2-ME) was added prior to sonication to retard ox- idation of the metal-binding protein and subsequent re- distribution of metals, as has been observed in mammalian systems (11). The sonicate was centrifuged at 40,000g for 20 min, resulting in a compacted pellet, a supernatant and a small, top layer consisting of fluffy material. In cells grown on high cadmium levels, the top layer has a rubberlike consistency The middle, super- natant layer was then ultracentrifuged at 100,000g for 1 to 1.5 hr, and a similar set of three layers resulted. The middle layer was carefully removed and was designated the cytosol fraction. In some preparations, Cd + was added to the media before fractionation, in order to dis- place Zn2+ from the BP fractions. Metal Analyses Measurements of metal concentrations were made us- ing either a Perkin-Elmer 360 or an Instrumentation Lab- oratories (AA/AE) 357 spectrophotometer with D2 back- ground correction applied for cadmium and zinc. Calibration standards were made using aqueous dilutions of 1000 pLg/mL atomic absorption standards obtained from Fisher. Whenever appropriate, dilutions were made using glass distilled water. Electron Paramagnetic Resonance Electron paramagnetic resonance (EPR) spectra of cy- tosol prepared from cells grown in medium containing 10 ,ug/mL copper were obtained by using a Varian E-3 EPR spectrometer. The field was set at 3017 G, and 4-min scans of frozen cytosol were recorded. Diphenylpicrylhydrazyl (DPPH) was used as a marker. The frozen samples were prepared by sealing the end of a glass tube with parafilm, filling the tube with cytosol, and immersing the tube into liquid nitrogen. The solid pellet formed was then (after slight warming) transferred to an EPR tube, and the spectrum was recorded. The sample was removed from the EPR tube, allowed to thaw, refrozen, and additional spectra obtained at selected times. Gel Chromatography The fractionation of cytosolic species was generally ac- complished by applying the sample to a 2.5 x 60 cm Sephadex G-75 column at 4 to 7?C, and eluting with 0.05 M Tris buffer, pH 8.6 (room temperature). The resulting fractions were analyzed for metal content by atomic ab- sorption via direct aspiration into the flame. The columns were calibrated by using blue dextran (MW 2,000,000), myoglobin (MW 17,000), horse kidney or rabbit liver me- tallothionein (MW 6,000; apparent MW 10,000 due to its elliptical shape), K3Fe(CN)6 and Na+. Elution positions are reported as Kd, the distribution coefficient. Kd = (Ve - Vo)/V8, where VO is the void volume (considered here as the elution volume of blue dextran), Ve is the elution volume of the species of interest, and V. is the volume of solvent inside the gel and available to very small molecules (V8 = VO - VNa'). 78 CADMIUM-BINDING PROTEIN FROM ALGAE Ion-Exchange Chromatography DEAE-Sephadex-A-25 ion-exchange chromatography was performed on the pooled, cadmium-containing, 10,000-dalton peak obtained from Sephadex columns. Early trials demonstrated that no cadmium-containing species eluted from the column until the conductivity reached at least 15 mSeiman. Consequently, the linear gradients used thereafter generally consisted of either 0.1 M to 0.4 M KCI, or 0.15 to 0.6 M KCI, in 0.05 M Tris buffer, pH 8.6, or alternately from 0.05 M Tris containing 0.2 M KCI to 0.5 M Tris containing 0.2 M KCI. Some columns were run with 10 mM ascorbic acid or 1 mM 2- ME added to the gradient buffers in efforts to stabilize the protein during this procedure. The integrity of the gradients was monitored using a Radiometer-Copen- hagen CDM3 conductivity meter equipped with a type CDC 314 cell. Reaction of EDTA with CdBP Sephadex G-50 CdBP fractions were concentrated us- ing ultrafiltration, reapplied to a Sephadex G-75 column, and eluted with 0.05 M Tris, pH 8.6, buffer. A series of samples was prepared that contained 0.2 mL 2 M KCl and 3.77 mL of CdBP having a Cd concentration of 4.0 ,ug/mL. These samples were kept at -20?C (approxi- mately 1 week) until the reactions were performed. The reactions were initiated by the addition of sufficient aqueous EDTA solution to 1.985 mL of the CdBP sample, yielding 0.4 mM EDTA in the final reaction mixture. The blank consisted of 0.1 mL 2 M KCl, 1.885 mL of Tris buffer and 0.015 mL of the EDTA solution. The reaction mixture was then transferred to a 1-mL cuvet within 30 sec and the course of the reaction was monitored at 254 nm over time. Due to their long duration, the reactions were monitored continuously for roughly 5 hr and then periodically at later time points. Four repetitions using two independent CdBP samples were performed. The data were analyzed by plotting the log (Ao. - A,) versus time. Radioimmunoassay Samples of CdBP from Sephadex G-50 fractions and CdBP-1 and CdBP-2 from DEAE-Sephadex A-25 frac- tions were prepared by transferring aliquots appropri- ately diluted with glass-distilled water into small vials such that each vial contained 1.2 to 2.8 x 10-4 mole of protein. The estimations of protein were made by meas- uring the Zn, Cu, and Cd content and assuming the ac- cepted mammalian values of 7 g-atoms of Zn or Cd and 10 g-atoms Cu bound per mole of metallothionein. The samples were lyophilized and sent to Syracuse University They were redissolved in doubly distilled H20, and ap- propriate serial dilutions were made and assayed in a double-antibody radioimmunoassay (RIA) following a protocol as previously described (12,13). Both the '5I- labeled antigen and a reference antigen were prepared from the same rat liver Cd,Zn-thionein. This isoform is known to be cross-reactive with various mammalian me- tallothioneins. Inverse variance weighted standard curves were developed from the data following established protocols (14). Changes in the sequence or in conforma- tion affecting the principal determinants of metallothi- onein (15), and thus the affinity of the altered determi- nant for the primary antibody, may be evaluated in terms of either changes in slopes of the standard curves with respect to that of the reference isoform or in terms of changes in concentration of the competing antigen re- quired to reduce the fraction bound of labeled antigen to 0.5. Results Isolation of CdBP CdBP for physiochemical studies was isolated from Euglena gracilis grown in culture media containing 5 ,ug/mL of Cd=+ and following procedures previously de- scribed (5). Fractionation of the cytosol (100,000g su- pernatant fraction) over Sephadex G-75 yielded a Cd-rich band, which also contained zinc and a trace of copper under some conditions. After experimenting with a va- riety of isolation procedures (aerobic, N2 atmosphere, added ascorbate or 2-mercaptoethanol), it was found that maintaining a 1 mM concentration of 2-mercaptoethanol (2-ME) in the buffers throughout the preparation yielded the most reproducible results. Rapid handling during all steps subseqeunt to the cell sonication was necessary to obtain good yields of the crude BP after gel exclusion chromatography. Even with the most stringent precautions, loss of the zinc from the BP was observed in subseqeunt steps. Therefore, in later preparations of BP for physiochemical studies, Cd2" was added to the cytosol to displace Zn2+ from the BP and yield a more stable all-cadmium BP preparation. The presence of zinc in the BP when 2-ME is used, but not in its absence, demonstrates that the enormous very low molecular weight pool of zinc in Eu- glena gracilis is not an artifact due to extraction by 2- ME of Zn from the BP or high molecular weight zinc pools (Table 1). Elution of the BP from DEAE-Sephadex ion-exchange resin (Fig. 1) required high salt concentrations in the eluting gradient. Typically, 0.1 to 0.4 or 0.6 M KCI in 50 mM Tris-HCl was used. The presence of 1 mM 2-ME was necessary to prevent erratic elution behavior. Under these conditions some cadmium eluted in the column wash (s 30%). The remainder eluted as two CdBP bands at 22 to 25 mSiemen (BP-I) and 40 to 44 mSiemen (BP-II). The relative amounts of cadmium in the two BPs was sufficiently variable that in some cases BP-I was the major species and sometimes the minor. Temperatures between - 35 and - 80?C prevented degradation of these BP preparations during storage. Relationship of BP Cytosolic to Medium Cd Levels To assess the function of the BP, one must determine (1) whether it is constitutive or inducible or both; (2) at 79 GINGRICH ET AL. Table 1. Effects of added 2-ME and Cd2" on cytosolic metal distributions.a Added reagents Cytosolic Cdlugb Cytosolic Zn/,g 2-ME/mM Cd+2/4lg HMW BP VLMW HMW BP VLMW 0 0 29 82 17 >200 1 0 38 63 20 30 152 1 130 79 169 49 17 7.3 121 a Cells were grown for 6 days in media containing 5 ,g Cd/L. Results are normalized for a 250 mL culture. 2-ME was added to the homogenate; Cd2" to the cytosol. bHMW = high molecular weight; VLMW = very low molecular weight. 1.5 E -2 z 1.0 0 z lii 0 z o 0.5 0 cl 0 z w 50 iw E 40 > 30 F D 0 20 z 0 1 0 z 0 z z 0 -J Lii 10 0 10 20 30 40 50 60 70 FRACTION NUMBER FIGURE 1. DEAE-Sephadex A-25 chromatography of CdBP from Euglewa gracilis. The applied sample (333 Rg of Cd2" in 9.8 mL was eluted with a 150 to 600 mL KC1 gradient buffered at pH 8.6 with 50 mM Tris-HCl and containing 1 mM 2-ME: (0) Cd; (A) conductivity what levels of Cd2" does induction occur; and (3) whether the presence of the BP is toxic or protective to the cell. The ability to bind cadmium is the only known function and the most unique and easily determined property of the BP Therefore, the presence of BP in cells grown in Cd-free media was sought by addition of Cd2+ to the cytosol before fractionation. When the added Cd2+ was 40% of the total cellular zinc, a substantial amount of Cd bound to high molecular weight species (0.097 [Lg/108 cells). A distinct Cd band at Kd = 0.67, close to the normal elution profile of CdBP, was present (Fig. 2a and Table 2) and contained most of the cadmium. When the Cd/Zn ratio was 1.6, the high molecular weight Cd spe- cies remained constant, while the band at 0.67 became a shoulder on a large very low molecular weight band of Cd, Kd = 0.88 (Fig. 2b). Interestingly, the amount of high molecular weight cadmium is comparable to the amount found after in vivo exposure to high levels of Cd2+ in the media. The formation of the band at Kd = 0.67 suggests the presence of apo-BP in the cell cytosol. When cells were trown in media containing 0.4, 1.3, and 4.7 ,ug/mL of Cd +, the cytosol of the cells from the two lower concentrations contained no measurable amounts of CdBP (Fig. 3). Only at the highest level did the Cd Appear in the BP fractions. At the lowest levels of Cd2 , the cytosolic Cd appears exclusively in the high molecular weight bands, which have Cd/Zn mole ratios of 0.1 and FRACTION NUMBER _,. . _5 .5 (Zn+) CYTOSOL + Cd Na+ E 16Cd l 1-1 IZn z 10 2. -0 30 405n0 0 8 z Lii 0 z 0 0o.5 -i 0 I10 2 0 3 0 4 0 5 0 60 70 8 0 b FRACTION NUMBER FIGURE 2. Sephadex G-75 chromatography of cytosol after in vitro addition of (a) 0.4 Cd/Zn and (b) 1.6 Cd/Zn, based on total cytosolic zinc content. 50 mM Tris-HCl buffer, pH 8.6; (a) 2.0 mL sample and 4.0 mL fractions; (b) 5.0 mL sample and 3.8 mL fractions. 0.4. There is clearly no CdBP present in the cytosol. For the higher concentration, which yields the CdBP, the mole ratio is 0.9. (Note that Fig. 3 is plotted as ,ug/mL and not in molar units.) Possible explanations of these ob- servations are (1) the BP may only be inducible by high levels of Cd2" in the cell or external media or (2) the BP may have a weak affinity for Cd+ and may bind it only 80 CADMIUM-BINDING PROTEIN FROM ALGAE Table 2. Distribution of cytosolic zinc and cadmium (limole/108 cells) of Euglena gracilis grown in media with controlled cadmium levels and after in vitro addition to homogenates. Cadmium added Cadmium, pmole/108 cells Zinc, ,mole/108 cells Homogenate, Media, ,ug/mL Cd/Zn HMW BP VLMW HMW BP VLMW 0.4 0 0.01 a a 0.8 a 0.21 1.3 0 0.03 a a 0.07 a 0.27 4.7 0 0.11 0.33 a 0.12 a 0 0.46 0.097 0.10 a 0.023 a 0.28 0 1.6 0.11 0.31 a 0.031 a 0.19 a Below the limits of detection. 1.0 I . 0 5 z IL z 0 < 05- L z C) z ; i 1o r 05- 0 SEPHADEX G-75 13 ,ug/ml Cd - 1- I 1 -I I 4.7 ug/mI Cd 10 20 30 40 50 60 70 80 FRACTION NUMBER FIGURE 3. Sephadex G-75 chromatographic profiles of cytosol of cells exposed for 6 days to 0.4 (top), 1.3 (middle), and 4.7 (bottom) ,ug Cd/mL added to the media. For each sample equal amounts of cy- tosolic protein were applied. Fraction volumes 4.4 (top), 3.9 (mid- dle), and 4.1 mL (bottom). Conditions as in Fig. 2. after stronger binding sites in the hmw fractions are saturated. Reaction with EDTA Recent work on mammalian MTs has established that these proteins are dynamic and reactive entities (11,16- 18) in contrast to the earlier view that they are inert metal-sequestering agents (10). A variety of ligands, competing metals, and electrophiles are able to effect displacement of metals from metallothioneins. Since the Euglena gracilis BP appeared to be more labile than MT and to differ in characteristic properties such as size and charge, EDTA was used to probe its reactivity. Rate constants for the CdMT reaction with EDTA are known (16). Samples of the Sephadex G-75 BP preparation were reacted with a 0.4 mM EDTA solution. Since the EDTA is present in large excess, pseudo-first-order kinetic treatment was used to determine the reaction order with respect to cadmium bound to the protein. The semilog- arithmic plots of absorbance versus time yielded two steps in pseudo-first-order rate constants of 4 + 3 x 10-4 sec ' and 7 ? 2 x 10-5 sec -'. Approximately one- third of the cadmium reacts in the fast step and two- thirds in the slow step. Thus both steps have a depen- dence on the CdBP: ratef = kf[CdBP] and rate, = k,[CdBP] These data establish that EDTA reacts more rapidly with the CdBP than the mammalian MT, for which the pseudo-first-order rate constant under comparable con- ditions (1 mM EDTA) shows a single step with a rate constant of 2-4 x 10-6 sec- 1. The slow step observed here is an order of magnitude faster than for mammalian MTs and the fast step is two orders of magnitude faster. These data provide striking quantitative confirmation of the qualitative observation of the labilty of Cd in this protein. pH Titrations The thiol group of cysteine residues, which are ex- pected to play a major and possibly exclusive role in metal binding, lose their protons upon binding cadmium. Thus, pH titrations can be used to assess the strength of metal binding: Kb CdBP +nH+ = H BP+n2 + Cd2+ For mammalian MTs, the half-titration point (i.e., the pH at which half the bound metal is displaced by protons) is 4.5 for Zn2+ and 3.2 for Cd2+ (19). Aliquots of the BP were subjected to progressively smaller pH values and quickly fractionated at that pH (Fig. 4). By pH 5-6, the half-titration point was reached. At pH 3, which is close to the half-titration point of MT, the cadmium is com- pletely displaced from the BP 81 I GINGRICH ET AL. E o 0.6- 0- 0 10 20 30 40 FRACTION NUMBER FIGURE 4. Effect of pH on CdBP Aliquots (1.0 mL) of CdBP were adjusted to the pH of the column, incubated 10 min, then chro- matographed over Sephadex G-25 using 100 mM phosphate buffer. The stoichiometry of protons bound per Cd2" displaced is not known, and therefore, the binding constant cannot be calculated quantitatively, but the qualitative result, the much higher pH of the half-titration point, indicates a much weaker thermodynamic interaction between the BP and cadmium. Thus, the Euglena gracilis BP has an affinity for Cd that is much lower than the value of Kb = 7 x 1014 estimated for CdMt (19). Radioimmunoassay Isolated samples of CdBP-I and CdBP-II, which were shown to give only single bands by polyacrylamide gel- electrophresis, were subjected to competitive radioim- munoassay according to published methods (12,13). In this technique, a radiolabeled 125I-rat liver MT and the increasing amounts of the competing ligand are assayed with antibodies raised against rat liver MT. The results of assaying Euglena CdBP-I and CdBP-II and, as a con- trol, a mixture of rat liver MT-I and MT-II is shown in Table 3 and reported as the fraction Y of bound 125I1 labeled MT for various concentrations of competing li- gand. The amount of the rat liver MT required to achieve Y = 0.50 was 0.94 pmole as determined by regression analysis. In contrast, using 100 pmole of CdBP-I or CdBP-II yielded Y = 0.89 and Y = 0.98, respectively. Thus, the CDBPs do not effectively compete with the radiolabeled antigen, and there is negligible cross-reac- tivity of the Euglena BPs with antibodies to mammalian MTs. CdBP Synthesis in Dark-Grown Cells Euglena are heterotrophic organisms which can grow under illumination or in the dark. Because much of the previous work on zinc deficiency in this organism was done in the dark, we also examined their ability to gen- erate CdBP when grown under these conditions. Figure 5 shows the CdBP band which results from 6 days of growth with Cd2+ (5 ,g/mL) in the media and without illumination. The zinc and cadmium distribution of con- trol and cadmium-exposed cells are compared in Table 4. The cadmium in the high molecular weight band (180 nmole/108 cells) was marginally more than that observed in light-grown cells, and 85% of the cadmium (960 nmole/ 108 cells) was in the CdBP band. No cadmium was de- tected in the low molecular weight region. The distri- bution of zinc was altered, compared to the control cells, by a dramatic increase in the very low molecular weight zinc pool, 105 versus 1080 nmole/108 cells. Figure 4 shows that a similar but not so dramatic increase occurred with illuminated cells grown at comparable Cd levels. Cytosolic Copper Experiments were undertaken to determine whether high levels of copper in the media would act like cadmium, inducing a BP, or like zinc, appearing as a very low mo- lecular weight species. Cells were grown for 10 days in media containing 10 pug/mL of added Cu2+ and attained a density of 5 x 106 cells, comparable to the density of cells in normal media after 5 to 6 days (5). Three bands of copper are evident after fractionating the cytosol over Sephadex G-75 (Fig. 6). The copper bound to the high molecular weight fractions greatly exceeded the zinc con- tent: 180 versus 15 nmole/108 cells. The biggest pool of copper (300 nmole/108 cells), however, was in the region characteristic of metal-binding proteins and accounted for 43% of the cytosolic copper. A third major band was present in the very low molecular weight region and ac- counted for 32% of the copper. No zinc was bound to the CuBP and only a very small amount was present in the very low molecular weight region. In additional experiments where cells were grown for 4 days on media containing 5 to 10 ,ug/mL of added copper, the concentration of copper in the high molecular weight region again greatly exceeded the zinc content, and the majority of the zinc was in the BP region. There was, however, no very low molecular weight copper band, and the very low molecular weight zinc was depleted in the cells on 5 ,ug/mL of copper and absent from those on 10 ,ug/mL. Thus, copper does apparently induce and bind to some species of cytosolic protein similar in size and elution to the CdBPs. However, unlike cadmium-exposed cells, in which the very low molecular weight peak of zinc is com- 82 CADMIUM-BINDING PROTEIN FROM ALGAE Table 3. Radioimmunoassays: fraction of .2.I-labeled rat MT-I angiten bound (Y) as a function of competing ligand.a Y at competing antigen amount, pmole Competing agent 0.1 1 10 100 1000 10000 CdBP-I 0.98 0.89 0.88 CdBP-II 0.98 0.99 CdBP (G-50) 1.03 1.03 50/50 Rat MT-I and MT-II 0.87 0.49 0.41 aValues of responses with 50/50 mix of Rat MT-I and MT-II as competitor derived from standard curve [linear-log regression relating to response Y to Q (log pmole competing antigen)]: Y = 0.490-0.377 Q. SEM of Y values greater than 0.90 is ? 2%. E 14 DARK-GROWN CELLS . . 5 ,ug/ml Cd, 6 DAYS 10 20 30 40 50 60 70 80 FRACTION NUMBER FIGURE 5. Sephadex G-75 chromatographic proffle of (0) cytosolic Zn and (0) Cd from cells grown without mlumination for 6 days in media containing 5 ,ug Cd/mL. Applied sample: 2.5 mL. Fractions 3.4 mL. 50 mM Tris-HCl buffer, pH 8.6. parable to or larger than that for control cells, the very low molecular weight zinc band is diminished by copper exposure. An unexpected observation in the course of this study was that the cytosol of the copper-exposed cells was grey- ish-black in color. This contrasts with the light green color of the cells grown in normal media or with 5 ,ug/mL of Cd. On exposure of the cytosol to air, the color fades, and eventually the green color of normal cells is obtained. The EPR spectra of frozen samples of the grey-black cytosol are almost featureless, suggesting that Cu(I) is present as the sample is oxidized by air exposure, with concomitant loss of the grey-black color, a characteristic signal for Cu(II) is apparent after 1 hr and very intense after 5 hr. The loss of the black color is apparently due to oxidation of some unidentified Cu(I) species to copper (II). Discussion As metal-binding proteins in lower organisms are in- vestigated in greater detail, it becomes increasingly clear that they are not simply homologues of mammalian MTs, but a diverse and complex array of small proteins and peptides, varying in structure, chemical reactivity and physical properties. For example, in rainbow trout (Salmo gairdneri), four binding proteins have been iden- tified: two resemble MT and are present only after in- jecting cadmium into the fish and two are present after exposure to environmental cadmium but do not resemble MTs (6). The blue-green algae Synechoccus (sp) and the green alga Chlorella ellipsoidea each had only a single form of BP (8,20), in contrast to the inevitable two (and possibly more) isoproteins of mammalian MTs. Oligo- meric cadmium-containing aggregates of cysteine-rich penta- and heptapeptides (cadystin A and B, respec- tively) have been isolated from the fission yeast Schizo- saccharomyces pombe after cadmium exposure (9). Thus, the results presented here demonstrating signifi- cant differences in physiological behavior, structure, and chemical reactivity between the Euglena BP and mam- malian MTs are not unexpected. Turning first to its physiological significance, the BP could be isolated with an appreciable zinc content only after exposure to high levels of cadmium (> 4.7 ,g/mL) and with the use of 2-ME in the preparation. Following this finding, we repeated our earlier efforts to isolate a ZnBP (5) by exposing the cells to 20 ,ug/mL of zinc in the media and using 2-ME to stabilize any resulting ZnBP during the preparation. The results, however, were negative. While the possibility of losing zinc from the BP during isolation attempts cannot be unambiguously elim- inated, the variety of conditions employed, including an anaerobic atmosphere of N2 previously reported (5), are sufficiently varied and stringent to support the conclusion that in Euglena the very low molecular weight zinc pool is a real phenomenon. Under zinc deficient conditions the very low molecular weight Zn is analogous in behavior to zinc MT of Ehrlich ascites cells (5). At present, there is no evidence to support a role for the BP in zinc metab- olism. The formation of CdBP when media cadmium lev- Table 4. Comparison of cytosolic metal distribution for control and Cd-exposed cells grown in the dark. Zinc, nmole/108 cells Cadmium, nmole/108 cells Media Cd, ,g/mL HMW BP VLMW HMW BP VLMW oa 66 c 105 5b 120 c 1080 170 960 c a Harvested on day 6. b Harvested on day 7. 'Below the limits of detection. 83 GINGRICH ET AL. 0) FRACTION NUM R Na H H z LU z 0 C- .3 -j 10 20 30 40 50 60 70 80 90 100 110 FRACTION NUMBE-R FIGURE 6. Sephadex G-75 chromatographic profile of (C) Cu and (0) Zn in the cytosol of Euglena gracilis grown for 10 days in media containing 10 ,g Cu/mL. 50 mM Tris-HCI buffer, pH 8.6, was used to elute the column. els were high, but its absence at 1 ,ug/mL or less of Cd2" are indicative of a role during acute cadmium exposure, but not during low level, chronic exposure. This result could reflect induction of Cd2" only at very high media and intracellular Cd2" levels. Alternatively, it may result from a low affinity of the BP for Cd2 . When intracellular concentrations are low, cadmium is bound almost exclu- sively to high molecular weight species. It only appears in the BP fractions after a level of ca. 11 ,umole Cd/108 cells has been achieved. In support of the proposal that the BP has a low affinity for Cd2+ is the finding that in vitro addition of Cd2+ leads to a saturation of high mo- lecular weight binding sites at levels of Cd2+ (10-11 nmole/108 cells) similar to those achieved in vivo, and to the presence of Cd in the BP fraction. These data suggest that the BP is constitutive and may have other biological functions or at least be involved in other equilibria which render it less available to bind intracellular Cd2 . The thermodynamic, kinetic and structural properties observed here substantiate and extend our previous find- ings (5) that the Euglena BP is smaller and more nega- tively charged than metallothioneins. Table 5 compares and contrasts the properties of the BP and mammalian MTs. Clearly, although the BP is a cysteine-rich metal- binding protein, it has a unique set of chemical and phys- ical properties and is not simply a homologue of the MTs. Whereas half the cadmium of MT is displaced at ap- proximately pH 3, that point is reached by pH 5 to 6 for the BP (Fig. 4), indicating that much lower concentra- tions are necessary to compete for the Cd-binding sites. Thus, the pH titration provides direct chemical evidence that the equilibrium constant, Kb for the reaction Kb Cd2+ + H"BP = CdBP + nH+ is lower than the value 7 x 1014 which is characteristic of mammalian MTs (19). Similarly, the EDTA reaction provides evidence of in- creased kinetic lability for the CdBP Extraction of Cd from CdMT or Cd,ZnMT by 1 mM EDTA results in a pseudo-first-order rate constant of 2-4 x 10-6 sec-1, which accounts for over 90% of the total reaction (16). The slow and fast reactions of the Euglena BP, comprise two-thirds and one-third of the cadmium removed and are 10 and 100 times faster, respectively. These differ- ences in the rates of reaction must arise from substantial structural changes between the MT and the CdBPs. The two steps for the Euglena BP may be due either to the two forms of BP, one more labile than the other, or to two classes of binding sites present in both BP-I and BP-II. Table 5. Comparison of Euglena BP and mammalian metallothioneins. Property Euglena CdBPs Metallothioneins Apparent molecular weight -8000 10,000 daltons Molecular weight ? 6000 DEAE elution (conductivity mS) 20-40 mS 6-10 mS Cysteine content High 33-34% 1/2 titration pH 5-6 3-4 EDTA pseudo-first-order 4 x 10' sec-1 (-33%) Very fast (10%) rate constants (% reacting) 7 x 10-' sec'1 (-67%) 2-4 x 10-6 sec-' (90%) [EDTA] [0.4 mM EDTA] [1 mM EDTA] Inducibility High (Cd2+) Low (Cd2+) Forms CdBp-I and CdBP-II MT-I and MT-II Cross-reactivity to rat liver MT antibody No Yes 84 CADMIUM-BINDING PROTEIN FROM ALGAE 85 Further studies to resolve this question are in progress. The radioimmunoassay of BP-I and BP-II against rat liver metallothionein antibodies indicated essentially no cross-reactivity. The antigenic sites of metallothionein are sequential. The immunodominant site involves resi- dues 20-26; a secondary site involves residues 1-5 at the (acetylated) amino terminal. The antigenicity of various metallothioneins is not significantly altered by dramatic structural changes such as replacing zinc or cadmium with gold (15) or by removal of metals and oxidation of the thiols to cysteic acid groups (18). Thus, the absence of cross-reactivity of the Euglena BP-I and BP-II to the antibodies cannot be attributed to conformational differ- ences and most likely indicates that the antigenic site is not present. Interestingly, the immunodominant deter- minate sequence includes two to three lysines, and the absence of these and perhaps other lysines from the BP sequence may contribute to the increased negative charge found by ion-exchange chromatography. The high conductivity required for DEAE elution, the lower apparent molecular weight, and the absence of cross-reactivity with MT antibodies all point to sub- stantial differences in amino acid sequence for the BPs. One would expect on this basis that the sites of metal binding would differ from those of the three- and four- metal cluster (21) of MT. Alterations in the metal-binding structures should give rise to changes in the equilibrium binding constants for Cd and to the rates of reactions with competing metal ligands, as have been reported here. Finally, changes in the structure and reactivity may indicate a physiological role of the CdBP in Euglena that is very different from that of MT in mammalian kidney and liver. The failure of Cd2+ to bind to the BP upon low level (< 1 [xg Cd/mL) exposure supports the hypothesis of altered function. Also, the failure of zinc to bind, except in the presence of bound cadmium, may also reflect an- other function. Clearly, continuing efforts to unravel the structure, chemical reactivity and function of the BP are warranted not only by their intrinsic interest, but also because of the light they can shed on MT structure and function in relation to human metal metabolism and health. The authors (CFS, DHP DJG and DNW) wish to acknowledge sup- port from the NIEHS Aquatic Biomedical Research Center, (ES-01985) and the UWM Graduate School. The radioimmunoassays (JSG) were performed with support from NIH (NIEHS 01629). REFERENCES 1. Crossley, L. G., Falchuck, K. H., and Vallee, B. L. Messenger ribonucleic acid function and protein synthesis in zinc-deficient Eu- glena gracilis. Biochemistry 21: 5339-5363 (1982). 2. Bonaly, J., Bariaud, A., Delcourt, A., and Mestre, J.-C. Les effets des ions Cd"2 sur les cellules d'Euglena gracilis. La cytotoxicit6 et aquisition d'une resistance. C. R. Acad. Sci. Paris 287: 463-466 (1978). 3. Falchuck, K. H., Fawcett, D. W, and Vallee, B. L. Role of zinc in cell division in Euglena gracilis. J. Cell Sci. 17: 57-78 (1975). 4. Price, C. A., and Vallee, B. L. Euglena gracilis, a test organism for study of zinc. Plant Physiol. 37: 428-433 (1962). 5. Gingrich, D. J., Petering, D. H., and Shaw, C. F., III. Zinc and cadmium metabolism in Euglena gracilis: metal distribution in normal and zinc-deficient cells. Marine Environ. Res. 14: 89-102 (1984). 6. Dafydd, G. T., Solbe, J. F., de L. G., Kay, J., and Cryer, A. Environmental cadmium is not sequestered by metallothionein in rainbow trout. Biochem. Biophys. Res. Commun. 110: 584-592 (1983). 7. Lerch, K., Amer, D., and Olafson, R. W Crab metallothionein: primary structures of metallothioneins 1 and 2. J. Biol. Chem. 257: 2420-2426 (1982). 8. Olafson, R. W, Abel, K., and Sim, R. G. Prokarytoic metallothi- onein: preliminary characterization of a blue-green alga heavy metal-binding protein. Biochem. Biophys. Res. Commun. 89: 36- 42 (1979). 9. Kondon, N., Imai, K., Isobe, M., Goto, T., Murasugi, A., Wada- Nakagawa, C., and Hayashi, Y Cadystin A and B, major unit peptides comprising cadmium binding peptides induced in a fission yeast. Tetrahedron Letters 25: 3869-3872 (1984). 10. Kagi, J. H. R., and Nordberg, M. Metallothionein. Birkhauser Verlag, Basel, 1979. 11. Minkel, D. T., Poulson, K., Wielgus, S., Shaw, C. F, III, and Petering, D. H. On the sensitivity of metallothioneins to oxidation. Biochem. J. 191: 475-485 (1980). 12. Vander Mallie, R. J., and Garvey J. S. Production and study of antibody produced against rat cadmium metallothionein. Immu- nochemistry 15: 857-868 (1978). 13. Garvey, J. S., Vander Mallie, R. J., and Chang, C. C. Radioim- munoassay of metallothioneins. Methods Enzymol. 84: 121-138 (1982). 14. Robard, D., Lenox, R. H., Wray, H. L., and Ramseth, D. Statis- tical characterization in the random errors of the radioimmunoassay dose-response variable. Clin. Chem. 22: 350-358 (1956). 15. Winge, D. R., and Garvey, J., S. Antigenicity of metallothionein. Proc. Natl. Acad. Sci. (U.S.) 80: 2472-2476 (1983). 16. Li, T.-Y, Kraker, A. J., Shaw, C. F, III, Petering, D. H. Ligand substitution reactions of metallothioneins with EDTA and apocar- bonic anhydrase. Proc. Natl. Acad. Sci. (U.S.) 77: 6334-6338 (1980). 17. Li, T.-Y, Minkel, D. T., Shaw, C. F, III, and Petering, D. H. On the reactivity of metallothioneins with DTNB. Biochem. J. 183: 441-446 (1981). 18. Laib, J., Shaw, C. F, III, Petering, D. H., Eidsness, M. K., Elder, R. C., and Garvey, J. S. The formation and characterization of aurothioneins: Au,Zn,Cd-thionein, Au,Cd-thionein and (thioma- lato-Au),-thionein. Biochemistry 24: 1977-1986 (1985). 19. Bachowski, G., Shaw, C. F, III, and Petering, D. H. Apparent stability constants for zinc and cadmium binding to apometallo- thionein. Unpublished data. 20. Nagano, T., Miwa, M., Suketa, Y, and Okada, S. Isolation and physiochemical properties and amino acid composition of a cadmium binding protein from cadmium-treated Chkorella ellipsoidea. J. Inorg. Biochem. 21: 61-71 (1984) 21. Otvos, J. D., and Armitage, I. M. Structures of metal clusters in rabbit liver metallothionein. Proc. Natl. Acad. Sci. (U.S.) 77: 7094- 7098 (1980).