Cadmium Resistance Conferred to Yeast by a Non-metallothionein-encoding Gene of the EarthwormEnchytraeus *

The earthworm Enchytraeus is able to survive in cadmium (Cd)-polluted environments. Upon Cd exposure, the worms express a gene encoding the putative non-metallothionein 25-kDacysteine-rich protein (CRP), which contains eight repeats with highly conserved cysteines in Cys-X-Cys and Cys-Cys arrangements exhibiting 36–53% identities to the 6–7-kDa metallothioneins of different organisms. Here, we demonstrate that the CRP protein confers a highly Cd-resistant phenotype to a Cd-hypersensitive yeast strain. Cd resistance increases with increasing numbers of expressed CRP repeats, but even one 3-kDa CRP repeat still mediates Cd resistance. Site-directed mutagenesis reveals that each single cysteine within a given repeat is important for Cd resistance, though to a different extent. However, replacement of other conserved amino acids such as Pro136 and Asp196 at the CRP repeat junctions does not affect Cd resistance. Our data indicate (i) that the non-metallothionein CRP protein is able to detoxify Cd and (ii) that this is dependent on the availability of sulfhydryl groups of the conserved cysteines.

control cellular zinc distribution, translocation, and availability (11), although, besides other functions, they may also act as general anti-stress factors (12).
Scarce information is available suggesting that, besides MTs, also larger non-MT proteins are involved in Cd detoxification (13). However, the investigation of these non-MTs has been oddly neglected to date. Especially some invertebrates have been reported to contain Cd-binding non-MTs but without any evidence for their role in Cd detoxification or molecular characterization (14 -17). We have identified a putative Cdbinding non-MT protein as a cDNA in the earthworm Enchytraeus using differential screening of cDNA library (18,19). These small oligochaete worms of high ecological relevance due to their function in soil formation and preservation are capable of surviving in acidic soils highly contaminated with Cd (20). The identified cDNA of Enchytraeus encodes a putative cysteine-rich non-MT 25-kDa protein, termed CRP, with eight tandemly arranged repeats exhibiting a characteristic conserved arrangement of cysteines (19). The crp gene is induced by Cd, and its transcript level positively correlates with Cd accumulation of worms (21). However, Cd inducibility of crp-mRNA does not necessarily mean that the CRP protein is directly involved in Cd detoxification. Indeed, a Cd-inducible mRNA encoding the non-MT CIP2 (cadmium-induced protein) has been recently detected in the fungus Candida sp. (22). However, the only 4 cysteines containing CIP2 is presumably not a Cd-binding protein, and it is not directly involved in Cd detoxification. Rather it is assumed to be involved in coping with the oxidative stress induced by Cd. Here, we provide experimental evidence for the actual role of the non-MT CRP protein in Cd detoxification. Transformation of crp-cDNA into Cd-hypersensitive yeast dramatically increases their Cd resistance, the extent of which is dependent on the number of CRP repeats and on the position of the different cysteines in a given CRP repeat.

EXPERIMENTAL PROCEDURES
Constructs of crp in Yeast Vector pRS425-Yeast expression vector pRS425 (23), kindly provided by Dr. G. Jansen (Institute of Microbiology, Heinrich-Heine University, Dü sseldorf, Germany), was modified by inserting a 6ϫ c-Myc tag as N-terminal fusion. Different arrangements of crp repeats were constructed by amplifying corresponding regions of the 1474-bp crp-cDNA (GenBank TM accession number X79344) and inserting the PCR products in the SalI restriction site of pRS425. Constructs of crp repeat 4 were cloned in pRS425 vector without a 6ϫ c-Myc tag.
Site-directed Mutagenesis-This was performed with the "Altered Sites II in Vitro Mutagenesis" system (Promega, Madison, WI). In brief, PCR-generated fragments of the complete coding region of the crp-cDNA (753 bp) and the crp repeat 4 (93 bp) were cloned in the SalI site of vector pALTER-1. Oligonucleotides used for mutation of the nine cysteines of repeat 4 and the conserved residues at the repeat junctions 4/5 and 6/7, respectively, were as follows: Cys 1 (5Ј-GAGTCGACAATG-AGCTCCTGTGGTT); Cys 3 (5Ј-ACAATGTGCTCCAGTGGTTCAGGA); * The costs of publication of this article were defrayed in part by the payment of page charges. This article must therefore be hereby marked "advertisement" in accordance with 18 U.S.C. Section 1734 solely to indicate this fact.
Western Analysis-Proteins extracted from yeast cells were separated by SDS-PAGE (31), electroblotted onto nitrocellulose, and processed as detailed previously (24). The filters were blocked with Rotiblock (Roth, Karlsruhe, Germany) and then incubated with a monoclonal mouse anti-c-Myc antibody (Santa Cruz Biotechnology, Santa Cruz, CA) and with peroxidase-conjugated rabbit anti-mouse IgG (1: 25,000) (Jackson ImmunoResearch Laboratory, West Grove, CT) as a secondary antibody before detection with ECL (Amersham Biosciences). Protein loading on filters was examined with a goat anti-actin antibody (Santa Cruz Biotechnology, Santa Cruz, CA).

RESULTS
The crp gene of Enchytraeus encodes the putative non-MT 25-kDa CRP protein consisting of eight tandemly arranged repeats (Fig. 1A). The repeats 1-7 contain 31 amino acids, whereas the repeat 8 is truncated at the C terminus (19). The repeats exhibit a characteristic arrangement of nine cysteines in Cys-X-Cys and Cys-Cys segments as it is shown for the consensus sequence of the CRP repeats (Fig. 1A). Data bank analysis does not reveal the existence of any other protein similar to CRP. Only the much smaller 6 -7-kDa MTs exhibit sequence similarities. The MTs from oligochaetes such as Lumbricus terrestris, Lumbricus rubellus, and Eisenia foetida display the highest identities of more than 50% to the same CRP region comprising repeats 2-4 in highly scored alignments.
The MT of the free living nematode Caenorhabditis elegans surprisingly aligns with only 37% identity to CRP repeats 6 -8.
Incidentally, the MT of C. elegans also exhibits lower identities in comparison with other invertebrate MTs (35). The MT of Drosophila melanogaster exhibits 46% identities to CRP repeat 3. The mouse MT-I and MT-II align best to CRP repeat 3 as those MTs of the other oligochaetes, although with lower identities. Like MT-I of C. elegans, mouse MT-IV is highly identical to CRP repeats 6 -8, whereas mouse MT-III fits best to CRP repeat 2. Human MT-I, -II, and -IV align predominantly to repeat CRP regions 4 -6, whereas the human and the mouse MT-III show highest identities of about 37% to CRP repeats 2 and 3. The multiple sequence alignment in Fig. 1B shows that there is a conspicuous conserved distribution of the cysteines in the CRP repeats and in the different MTs, besides some other amino acids such as glycine, lysine, and serine.
For expression in yeast, diverse crp constructs were generated by PCR ( Fig. 2A). All crp constructs, with the exception of crp-r4, were N-terminally tagged with a 8-kDa 6ϫ c-Myc by cloning in yeast expression vector pRS425 under the control of a MET25 promoter (23). The crp-r4 construct was cloned in pRS425 without a c-Myc tag to avoid possible steric hindrance by the c-Myc polypeptide. All constructs were transformed in Cd-hypersensitive Saccharomyces cerevisiae strain DTY167 (30), which harbors an inactivated gene for the yeast Cd factor (ycf1), normally required for Cd resistance (36). The crp constructs are expressed in approximately equal amounts in yeast (Fig. 2B). The expression rate is low, because the CRP polypeptides are not detectable in Coomassie-stained SDS gels but only by Western blot analysis using anti-c-Myc antibody. Confocal laser scanning microscopy reveals that the 25-kDa CRP is exclusively localized in the cytoplasm but not in the vacuole or nuclei of yeast cells (Fig. 2C).
In order to investigate the possible role of CRP in mediating Cd resistance, the DTY167 cells expressing the 25-kDa CRP (strain designated as DTY167-CRP), the isogenic wild-type DTY165 cells, and the Cd-hypersensitive DTY167-pRS425 cells transformed with empty vector were exposed for 72 h to Cd 2ϩ concentrations up to 500 M (Fig. 3A). The DTY167-CRP cells do not only restore the Cd resistance but even exhibit a dramatically increased Cd resistance in comparison with wild-type DTY165 cells. At 100 M Cd 2ϩ , DTY167-CRP reaches about 90% of growth that can be observed for control cultures without Cd. By contrast, wild-type DTY165 cells only reach a level of about 60% at 100 M Cd 2ϩ . There is no growth observed for DTY165 at 300 M Cd 2ϩ and higher concentrations. However, strain DTY167-CRP even tolerates 500 M Cd 2ϩ , because it still reaches 20% of growth level of the untreated control (Fig.  3A). The hypersensitive DTY167-pRS425 cells are unable to grow at the used Cd 2ϩ concentrations.
The levels of Cd resistance restored in DTY167 positively correlate with the increasing number of CRP repeats (Fig. 3B). All transformants expressing CRP repeats, except for DTY167-CRP-r4, show comparable growth at 100 M Cd 2ϩ after 72 h. Deletion of the N-terminal region and repeat 1 (DTY167-CRP⌬r1) has no significant influence on the Cd resistance compared with DTY167-CRP (Fig. 3B). The five repeats expressing transformants DTY167-CRP-r12345 and DTY167-CRP-r12678 have identical growth rates, but both display much lower resistance at 300 M Cd 2ϩ and higher Cd 2ϩ concentrations compared with DTY167-CRP. Growth of these strains stops at 400 M Cd 2ϩ . A further reduced Cd resistance is observed for those yeasts expressing only three repeats such as DTY167-CRP-r345 and DTY167-CRP-r678. They have equivalent levels of Cd resistance, although the three repeats expressed in the cells are different. Growth of both strains is completely inhibited at 300 M Cd 2ϩ . DTY167-CRP-r4 cells still exhibit a slight Cd resistance at 100 M Cd 2ϩ , because they reach 15% of the growth of the untreated control cells. Depending on the number of expressed repeats, generation times span between 3.17 Ϯ 0.23 h for DTY167-CRP and 4.60 Ϯ 0.64 h for DTY167-CRP-r4 at 40 M Cd 2ϩ , whereas hypersensitive DTY167-pRS425 cells double only after 13.65 Ϯ 1.73 h at this Cd concentration.
In order to investigate the importance of the individual Cys residues within a given CRP repeat for Cd resistance, Cys 3 Ser replacements were introduced in crp-r4 by oligonucleotidedirected site-specific mutagenesis. The nine mutated crp-r4 constructs cloned in vector pRS425 were transformed in hypersensitive strain DTY167 and exposed to 100 M Cd 2ϩ for 72 h. All Cys mutants reveal a dramatic decrease in Cd resistance compared with wild-type CRP-r4 expressing DTY167. Incidentally, the latter cells exhibit different growth depending on the used experimental conditions (cf. Figs. 3 and 4). The extent of the decreased Cd resistance depends solely on the mutated Cys position in crp-r4 (Fig. 4). Cys 3 , Cys 14 , Cys 22 , and Cys 23 are obviously of particular importance, because Cd resistance of the corresponding mutants decreases almost to the level of hypersensitive DTY167-pRS425 cells. Cys 9 and Cys 18 mutants confer a slightly better resistance since reaching about 10% of the growth of DTY167-CRP-r4. Mutations of Cys 1 , Cys 7 , or Cys 27 are less influential on Cd resistance than the other Cys positions, as the growth of these mutants is reduced by only 60% compared with DTY167-CRP-r4. Moreover, determination of generation times at a sub-lethal concentration of 30 M Cd 2ϩ confirms the relevance of Cys 3 . Mutation of that position at least doubles the generation time to 7.9 Ϯ 0.4 h, whereas the generation times of the other mutants span between 3.26 Ϯ 0.08 h (Cys 27 ) and 3.78 Ϯ 0.28 h (Cys 22 ). However, wild-type DTY167-CRP-r4 has a doubling time of 3.08 Ϯ 0.1 h at the same Cd concentration.
Besides the conserved Cys residues in the CRP repeats, there are also other conserved amino acids such as Gly, Pro, Val, and Asp at the repeat junctions (Fig. 5A). In order to investigate their role in functioning of CRP in Cd resistance, oligonucleotide-directed site-specific mutagenesis was used to generate amino acid replacements at two repeat junctions in the 25-kDa CRP (Fig. 5A). A Pro 136 3 Leu replacement was introduced in repeat 5 at the junction between repeat 4 3 5, and the mutated cDNA, cloned in pRS425, was expressed in strain DTY167 (designated as DTY167-CRPm4/5). Furthermore, a mutant CRP with an Asp 196 3 Asn replacement in repeat 6 at repeat junction 6 3 7 was constructed (DTY167-CRPm5/6). Comparative determination of the generation times at sub-lethal 40 M Cd 2ϩ and the growth in increasing Cd 2ϩ concentrations up to 750 M reveal no significant difference in Cd resistance among native DTY167-CRP, DTY167-CRPm4/5, and DTY167-CRPm6/7 (Fig. 5B).

DISCUSSION
Our data provide evidence that, besides MTs, also larger non-MT cysteine-rich proteins are able to detoxify Cd. Indeed, the non-MT 25-kDa CRP protein of the terrestric earthworm Enchytraeus mediates Cd resistance to the Cd-hypersensitive yeast strain DTY167 when transformed by crp. Remarkably, Cd resistance is not only restored but rather dramatically increased in comparison to isogenic wild-type yeast.
The CRP protein is unique to date, i.e. data bank analysis does not reveal any other similar non-MT protein in any other organism. The best fitting alignments can be obtained with MT of different sources. MT of other earthworms such as Lumbricus (37) or Eisenia (38) display the best identities of about 50 -53% to distinct regions of CRP, and even the more distant and diverse MT types of mouse and human exhibit identities of 37-44%. Besides the conserved cysteines, both MT and CRP contain still other conserved amino acids such as lysines, glycines, and serines. This structural similarity suggests a role of CRP in Cd detoxification similar to that of MT. Accordingly, mammalian MTs has been shown to mediate Cd resistance in yeast (39).
In yeast, Cd resistance is normally regulated by complex mechanisms. These involve GSH, the GSH-derived phytochelatins, and diverse membrane transporters (40). The hypersensitive strain we used in our study harbors an inactive ycf1 (yeast cadmium factor) gene (30). The ATP-binding cassette YCF1 protein is localized in the vacuolar membrane and serves as a pump, which transports Cd as a glutathione S-conjugate into the yeast vacuole (36,41). A similar pump mechanism is mediated in yeast by the transporter MRP1, the human multidrug-associated protein (42). CRP does not simply replace the lost function of YCF1 in Cd resistance of yeast, but rather CRP works by a different mechanism. This view is substantiated by our finding that CRP is uniformly distributed among the cytoplasm and is not detectable in the vacuolar membrane, although the CRP contains a putative transmembrane domain at the N terminus (19).
Currently, two mechanisms of MT in Cd detoxification are considered as follows: (i) chelation of Cd through coordinate covalent bonds to -SH groups of cysteines and/or (ii) scaveng-ing of free radicals originating during Cd-induced stress (8). CRP presumably functions by the same mechanisms. This view is supported by our finding that (i) Cd resistance of yeasts increases with increasing numbers of expressed CRP repeats and that (ii) mutations of distinct cysteines in a given CRP repeat result in a dramatic decrease or even loss of Cd resistance. However, our data also indicate differences in the mode of action of CRP and MT in Cd detoxification in yeast. First, each cysteine of a given CRP repeat is important for Cd resistance. Mutations of the cysteines in the CRP repeat at positions 1, 7, 9, and 27 result in a dramatic reduction of Cd resistance by at least 60%, whereas mutations in Cys 3 , Cys 14 , Cys 18 , Cys 22 , and Cys 23 even result in a complete loss of the capability of mediating Cd resistance. By contrast, mammalian MTs expressed in yeast contain at least some cysteines without any effect on Cd resistance at all (43). Second, S. cerevisiae is known to contain only one MT, i.e. the Cu-MT encoded by the CUP1 gene (44,45). This gene is specifically induced by copper but not by Cd, and its physiological role is the detoxification of copper and not that of Cd (44). However, when expressed under a constitutive promoter, Cu-MT is then also able to confer resistance to Cd (45). By contrast, the crp gene is not inducible by copper. However, if the gene is specifically induced by Cd, the copper toxicity is significantly decreased in Enchytraeus (21,46). Third, replacement of conserved amino acids other than the cysteines in MT results in a reduced Cd resistance of yeast hosts, whereas in CRP, substitution of conserved amino acids including the ␣-helix incompatible proline has no effect on CRP-mediated Cd resistance. The differences revealed between CRP and MT in yeast hosts support previous findings also revealing differences of CRP and MT at the RNA level. MT genes are constitutively expressed and respond to various stressors (47). However, the crp gene is not constitutively expressed; it is specifically induced by Cd but not by other stressors such as lead, mercury, copper, or H 2 O 2 (21). Collectively, our data indicate (i) that the non-MT CRP protein is able to detoxify Cd and (ii) that this capability is dependent on the availability of -SH groups similar to MTs. Although our results in yeast cannot yet be considered as applying to the situation in the earthworm Enchytraeus without further studies, the differences in the regulation of expression of the crp and MT genes reported to date suggest different actual physiological roles of both proteins.