The Arabidopsis thaliana CUTA gene encodes a 182-amino-acid-long putative precursor of a chloroplast protein with high sequence similarity to evolutionarily conserved prokaryotic proteins implicated in copper tolerance. Northern analysis indicates that AtCUTA mRNA is expressed in all major tissue types. Analysis of cDNA clones and RT-PCR with total mRNA revealed alternative splicing of AtCUTA by retention of an intron. The intron-containing mRNA encodes a truncated 156-amino-acid protein as a result of stop codons in the included intron. The sequence of AtCutAp encoded by the fully spliced transcript suggests that the precursor consists of three domains: an N-terminal chloroplast transit sequence of 70 residues, followed by a domain with prokaryotic signal-sequence-like characteristics and finally the most conserved C-terminal domain. The N-terminal chloroplast transit sequence was functional to route a passenger protein into isolated pea chloroplasts with possible sorting to the envelope. Chloroplast localization was confirmed by Western blot analysis of isolated and fractionated chloroplasts. Recombinant AtCutA protein was expressed in Escherichia coli without the N-terminal 70-amino-acid chloroplast transit sequence. This recombinant AtCutAp was routed to the bacterial periplasm of E. coli. Purified recombinant AtCutAp is tetrameric and selectively binds Cu(II) ions with an affinity comparable to that reported for mammalian prion proteins.
The transition metal copper is an essential element for all organisms. Despite being necessary for key functions, intracellular levels of transition metals such as copper must be precisely regulated to prevent damage caused by reactivity with sulfhydryl groups and formation of free radical species (Nelson, 1999). Studies in the model organism, yeast, indicate that cytosolic free copper levels are indeed kept below one atom per cell (Rae et al., 1999).
Copper functions within plant cells in a variety of processes, including electron transfer in photosynthesis and respiration, and lignification of cell walls (Fox and Guerinot, 1998). Copper homeostasis is also important for antioxidative responses and senescence (Mira et al., 2002). Key components of the photosynthetic machinery require specific metal cofactors including copper, as do other enzymes within chloroplasts. Plastocyanin is thought to contain more than 50% of the copper in chloroplasts (Marschner, 1995).
The plastid genome does not contain all genes necessary for chloroplast function. A large number of proteins must be synthesized in the cytosol from nuclear transcripts, imported through the chloroplast envelope, and targeted within the organelle based on information in an N-terminal transit peptide and secondary sorting information (Bruce, 2000; Keegstra and Cline, 1999; Schnell, 1998). Much of the post-translational assembly of photosynthetic metalloproteins occurs within chloroplasts after import of the precursors (reviewed by Merchant and Dreyfuss, 1998). We are interested in the molecular mechanisms that underlie the regulation of metal ion uptake, storage, and mobilization by plant chloroplasts. As a first step, we need to identify specific gene products involved in chloroplast metal homeostasis or metal binding.
Copper homeostasis has been well characterized in yeast and mammals, and a number of genes involved in copper trafficking have been identified in plants including RAN1, a homolog of P-type ATPase Ccc2 (Hirayama et al., 1999), and CCH, a homolog of the copper chaperone Atx1 (Himelblau et al., 1998; Mira et al., 2001). Much of what is known about copper trafficking in Arabidopsis thaliana has been characterized by complementation studies in yeast mutants (for an extensive review of copper traffic in plants, see Guerinot, 2000).
One component that may be involved in divalent metal ion homeostasis in prokaryotes is CUTA. The CUTA region in Escherichia coli was characterized by evidence of increased copper uptake and sensitivity in mutant cells (Crooke and Cole, 1995; Fong et al., 1995). The E. coli CUTA gene (also called CUTA1 or Orf 112) encodes a soluble protein of 112 amino acids that may have a role in divalent metal tolerance (Fong et al., 1995). However, thus far, the exact biochemical role of the CUTA protein in E. coli is not clear. The Arabidopsis Genome Initiative (AGI, http://www.arabidopsis.org/home.html) revealed the presence of a CUTA homolog, AtCUTA, in A. thaliana. Here, we present evidence that AtCUTA is expressed in root and shoot tissues, that the encoded protein has specific copper-binding properties, and that the protein is localized to chloroplasts. In addition, we find that AtCUTA mRNA is alternatively spliced. Results are discussed in the light of a possible function of AtCUTA in plastid metal ion homeostasis.
To identify chloroplast proteins involved in divalent metal homeostasis, we screened databases for genes encoding possible homologs of prokaryotic proteins involved in divalent metal homeostasis that are predicted to have plastid targeting sequences. AtCUTA was one of the genes identified as a homolog of E. coli CUTA (Fong et al., 1995) using this approach. AtCUTA (Figure 1) is predicted to encode a precursor protein of 182 amino acids with three predicted domains. Amino acids 1–70 make up a chloroplast targeting sequence, as predicted by the neural network-based programs ChloroP and TargetP (http://www.cbs.dtu.dk/services/ for literature reference, see Emanuelsson et al., 1999). Amino acids 71–106 include a number of hydrophobic residues and comprise a domain with signal-sequence-like characteristics that, when combined with the 70 N-terminal amino acids, potentially functions as a bipartite targeting peptide (Schnell, 1998; Smeekens et al., 1986). The C-terminal domain of AtCuTAp (amino acids 107–182) is the most conserved region (Figure 1). Cleavage of the AtCutA precursor at amino acid 70 results in a 112-amino-acid protein with 58% sequence similarity and 39% sequence identity to the 112-amino-acid E. coli CutAp. Homologs of CUTA can be identified in numerous species by blast (Basic Local Alignment Search Tool; http://www.ncbi.nlm.nih.gov/BLAST/). An alignment of AtCutAp with several homologs is shown in Figure 1. Notable is a high degree of sequence conservation among phylogenetically divergent organisms including bacteria, Archaea, and Eukarya. No other homologs of CutAp were detected by blast search in the completely sequenced genome of A. thaliana.
To begin a functional analysis of AtCUTA, two clones of ESTs for AtCUTA were obtained. The insert in one of the clones (92B37T) was very short (<200 base pairs, as determined by PCR and restriction analysis), and was thus determined to be incomplete. Restriction digest and PCR analysis indicated that the second clone (213A16T7) was complete. Analysis of the genomic sequence predicts six exons and five introns to give about a 700-base transcript. Surprisingly, sequence analysis revealed that this clone (213A16T7) retained the fifth intron and would predict a transcript of about 850 bases. Although the predicted protein translated from a fully spliced mRNA is 182 amino acids in length, translation of the EST clone would result in a protein lacking 26 C-terminal amino acids because of stop codons in all reading frames of the retained intron. The results suggest that AtCUTA is alternatively spliced. Indeed, RT-PCR revealed the presence of mRNA species corresponding to both fully spliced and intron-containing transcripts (Figure 2).
The 700-base pair fragment shown in lane 4 of Figure 2 was cloned, and its sequence was compared to that of the intron-containing clone. Sequence data confirmed that the shorter RT-PCR product was identical to the intron-containing clone, except for the included intron, and they were from transcripts of the same gene.
Northern blot analysis of poly(A)+ RNA from different tissues was carried out to determine the expression levels of AtCUTA mRNA. All major tissue types contained a single transcript of about 1.35 kb that hybridized with the AtCUTA probe (Figure 3). The presence of this single transcript indicates that the intron-containing mRNA (which is less abundant by RT-PCR) is not expressed at high levels. These results, together with the RNA blot analysis of ubiquitin (Figure 3), suggest that AtCUTA is expressed in all major plant organs and the mRNA level is comparable in all tested tissues.
We identified two independent T-DNA insertional knockout lines for AtCUTA. The T-DNA insertions are located in the third intron and in the second exon, respectively. The first line contained three additional T-DNA insertions, while the second line contained two additional T-DNA insertions. RT-PCR confirmed that AtCUTA mRNA was absent in the homozygous plants of both lines. Neither homozygous knockout line showed an obvious phenotype when grown on soil, in tissue culture, and in tissue culture with 30 µm copper when compared to wild-type plants. The second line containing only three T-DNA insertions was used for preliminary tissue culture experiments. We found no difference between 14-day-old knockout and wild-type plants grown on Murashige and Skoog (MS) agar media (Murashige and Skoog, 1962) with either no additional copper or 30 µm CuSO4 added. Mean fresh weights were as follows: wild-type control, 31.4 mg; knockout control, 27.56 mg; wild-type copper-treated, 20.8 mg; and knockout copper-treated, 19.0 mg. A second trial was performed with similar results. All analysis of variance P-values were greater than 0.10. Obvious differences in growth on soil were also not observed between wild-type and knockout lines. We tentatively conclude that AtCUTA is not essential either under normal conditions or under high-copper conditions.
In order to investigate biochemical and metal-binding properties, recombinant proteins lacking the putative N-terminal chloroplast targeting sequence of 70 amino acids were expressed in BL21-DE3 E. coli. Expression of amino acid residues 71–182, encoded by fully spliced mRNA (see Figure 1), resulted in the accumulation of a 12.4-kDa protein (Figure 4, lane 2). This size suggests that the putative signal-sequence-like domain, residues 71–106, is not cleaved in E. coli. Recombinant full-length AtCutAp was recovered from BL21-DE3 culture by osmotic shock and was separated from the bulk of the cellular proteins, suggesting a periplasmic location. Recombinant full-length AtCutAp was purified from the periplasmic fraction by ion-exchange chromatography on a DEAE Sepharose column (Figure 4, lane 3). The sequence of the N-terminal six amino acids of the purified protein, as determined by Edman degradation, was MEESSK. This result identified the purified protein as AtCutAp and confirmed that the putative 36-amino-acid signal-sequence-like domain (boxed in Figure 1) was not processed by E. coli. Purified recombinant AtCutAp was analyzed by gel filtration. The protein eluted as a single peak from a S200 Hi-prep 16/60 column (Pharmacia, Piscataway, NJ, USA) with an apparent molecular weight of 43 kDa, suggesting that it is purified from E. coli as a tetramer. To obtain the truncated protein encoded by the intron-containing clone, the sequence corresponding to residues 71–156 in Figure 1 was expressed in BL21-DE3. The truncated protein also accumulated in the periplasmic fraction following osmotic shock. The truncated protein was purified with the same techniques as the fully spliced protein (Figure 4, lane 6). SDS–PAGE indicated a molecular weight corresponding to the predicted 9.8 kDa for the truncated protein. Gel filtration analysis indicated that truncated AtCutAp is predominantly tetrameric (results not shown).
To test the hypothesis that AtCUTA may be involved in metal homeostasis, we analyzed the recombinant protein for interactions with metal ions by immobilized metal ion affinity chromatography (IMAC). Purified protein was incubated with metal-saturated resin. The unbound fraction was removed, and the resin was washed with the binding buffer. The resin was washed with a pH 5.5 buffer and was then finally eluted with EDTA to remove all metals and associated protein. Results are shown in Figure 5. No recombinant full-length AtCutAp was recovered in the unbound fraction for copper-saturated resin, nor AtCutAp released in the pH 5.5 wash. Full-length AtCutAp was quantitatively recovered when the copper-saturated resin was washed with EDTA (Figure 5, panel a). In contrast, recombinant AtCutAp was quantitatively recovered in the unbound fraction of all other metal-saturated resins and was not present in the EDTA washes. For the truncated AtCutAp, some, but not all, protein was recovered in the unbound fraction of copper and nickel-saturated resin (Figure 5, panel b). Truncated AtCutAp was recovered in the pH 5.5 wash of nickel- and zinc-saturated resin. Most of the loaded truncated AtCutAp was recovered in the EDTA wash of copper-saturated resin. Some truncated AtCutAp was also recovered in the EDTA wash for nickel as well as zinc. These results indicate that fully spliced AtCutAp interacts specifically with copper, while truncated AtCutAp has less specificity in metal binding.
Copper affinity and Cu(I) versus Cu(II) specificity were determined for purified fully spliced AtCutAp by incubation with copper ions in solution followed by unbound copper ion removal by ultrafiltration. The results indicate that purified AtCutAp preferentially bound Cu(II) ions in solution at an approximately 1 : 1 ratio. AtCutAp had low affinity for Cu(I) in solution by this test, while Cu(I) binding was below the detection limits of the assay when EDTA was added to the buffer solution to remove divalent copper ions (Figure 7; Rae et al., 1999). Therefore, the observed binding to Cu(I) may be caused by partial oxidation to Cu(II). Half-maximal binding under experimental conditions was at about 50 µm Cu(II) (Figure 6, inset). Aggregation of protein was visible as a precipitate at copper concentrations above 100 µm. The positive control, BSA (Linder, 1991), bound Cu(II) at approximately 3.5 : 1 ratio, while lysozyme, the negative control, did not show Cu(II) binding within the detection limits of the test (not shown). Truncated AtCUTAp had a similar half-maximal binding specificity for Cu(II) as fully spliced AtCUTAp under the same experimental conditions (data not shown). Unfortunately, the truncated protein showed an increased tendency to aggregate at high Cu(II) concentrations, which precluded the determination of a binding constant.
Precursor-AtCutAp contains characteristics of a precursor protein with a targeting peptide for plastid localization. In vitro import studies were performed using isolated pea chloroplasts to analyze the presence of chloroplast targeting and possible suborganellar routing information in the AtCutA precursor. A fusion clone with plastocyanin mature sequence as a passenger was constructed to provide for efficient labeling with 35S-methionine. The results, shown in panel (a) of Figure 7, indicate that the plastocyanin control protein was imported to the thylakoid fraction and was processed to its mature size (Figure 7, panel a, top). Upon incubation of the AtCutA-plastocyanin fusion precursor with isolated chloroplasts, labeled proteins with smaller size than the precursor but larger than mature plastocyanin, were detected in the chloroplasts (Figure 7, panel a, bottom). This intermediate-sized protein was fully protected against thermolysin treatment of intact chloroplasts and was thus fully translocated across the chloroplast outer membrane. Quantification indicated that the import efficiency of the fusion was comparable to that of the authentic plastocyanin precursor. After hypotonic lysis of chloroplasts and centrifugation, the fusion protein was recovered in the soluble fraction and not in the membrane pellet. The lysis procedure separates thylakoids from soluble stromal and envelope fractions, indicating that the fusion protein was either in the stroma or in the envelope intermembrane space.
Protease treatments were used to distinguish between stromal and envelope-localized proteins in additional import experiments. Trypsin can pass through the chloroplast outer membrane but not the inner membrane, and will digest proteins exposed to the intermembrane space of the chloroplast envelope; thermolysin cannot pass through the outer envelope (Lübeck et al., 1997). Chloroplasts were isolated and incubated with AtCutA-plastocyanin fusion as above, followed by fractionation and protease treatments designed to differentiate between stromal and chloroplast envelope localization. The results are shown in panel (b) of Figure 7. The AtCutA-plastocyanin fusion protein was imported and protected from digestion by thermolysin. However, the protein was sensitive to trypsin even after import, indicating a possible chloroplast envelope location. The soluble nature of AtCutAp suggests that it is located in the intermembrane space rather than as a membrane protein in the inner or outer envelope. Degradation of the AtCutAp-plastocyanin protein in the stromal fractions by both thermolysin and trypsin confirmed susceptibility to both proteases.
To directly analyze chloroplast localization of AtCutAp in Arabidopsis, we raised antibodies against the purified protein. The antiserum was specific to AtCutAp and detected as little as 20 ng of purified AtCutAp. In wild-type plants grown under standard conditions, detection of AtCutAp required overloading of gels for Western blots. AtCutA protein was detected in isolated chloroplasts with this approach (result not shown). Based on comparisons of the intensity of staining on Western blots of purified protein and plant samples, we calculated approximately 50 000 copies of AtCutAp to be present per plastid. To obtain more optimal resolution, we used AtCutAp-overexpressing line SPL-6B to detect the protein in subcellular fractions. Panel (c) of Figure 7 shows AtCutAp in total plant homogenate, intact chloroplasts, and a stromal/soluble envelope fraction, thus confirming chloroplast localization of AtCutAp. Positive control blots on the same fractions were performed with proteins for chloroplast thylakoids (plastocyanin; Smeekens et al., 1986) and chloroplast stroma (AtNifS; Pilon-Smits et al., 2002). Negative controls for chloroplast isolation and immunodetection from the overexpressing plants were for mitochondria (PMO35; Tom Elthon, personal communication), endoplasmic reticulum (AtSEC12; Bar-Peled and Raikhel, 1997), and cytosol (p28; Sebastian Bednarek, personal communication). The negative controls were present in the total homogenate but not in the chloroplast fractions, as indicated in panel (d) of Figure 7.
The A. thaliana gene AtCUTA was characterized in this study. Comparison of AtCUTA with sequences from other organisms showed that CutA protein sequences are evolutionarily conserved, which may reflect a critical function of the protein. AtCUTA mRNA is expressed in all major tissue types, and the mRNA is alternatively spliced. Alternative splicing may have a role in controlling the functional characteristics of AtCUTA: altering the physiochemical properties of AtCutAp by inclusion or exclusion of an intron and translation of a truncated or full-length protein. Although numerous animal genes have been identified that have tissue- or developmental stage-specific alternative splicing (Green, 1991), the role of alternative splicing in gene expression in plants is a relatively new area of research, and the functional significance is not yet clear in the case of AtCUTA. Among the characterized examples of alternative splicing in plants, functions range from control of flowering time (Macknight et al., 1997) to controlling subcellular localization, as influenced by light regulation (Mano et al., 1999), and, in one case, a single mitochondrial targeting sequence carrying either of two mature proteins (Kubo et al., 1999).
An important clue to the function of AtCutAp is obtained from its subcellular localization. Targeting sequences that route cytosolically synthesized proteins to plastids and signal sequences for suborganellar sorting have been documented in the literature (Keegstra and Cline, 1999; Schnell, 1998; Smeekens et al., 1986). The domain structure of the AtCutA precursor is reminiscent of precursors with bipartite targeting peptides, such as preplastocyanin, that serve to localize the mature protein to the thylakoid lumen (Smeekens et al., 1986). However, AtCutAp does not contain the AXA↑X motif described by Schnell (1998) as a consensus thylakoid cleavage site. In vitro uptake experiments with isolated chloroplasts indicated that AtCutAp is routed to chloroplasts and sorted to either the stroma or the envelope. It is possible for an envelope protein to be sorted as a soluble stromal intermediate and exported to the envelope, as is the case of IEP110 (Lübeck et al., 1997), although we suspect that AtCutAp will be in the intermembrane space rather than in the inner envelope membrane. The observation that recombinant AtCutAp expressed in E. coli accumulated in the periplasm, a location that can only be reached by undergoing export from the cytosol, supports the assertion that the putative secondary signal sequence functions in export. Assuming that the signal-sequence-like domain of AtCutAp was responsible for the periplasmic location in bacteria, a similar mechanism may be used to export AtCutAp to the chloroplast envelope after import into the stroma and cleavage of the N-terminal 70-amino-acid transit sequence. Mammalian CutAp (mCutAp) contains an N-terminal membrane anchor (Navaratnam et al., 2000). This could also be the case with the similar putative secondary signal sequence in AtCutAp. Thus, it is feasible that the mature AtCutAp is localized to the intermembrane space and retained by the hydrophobic N-terminal region inserted in the inner membrane.
Purified recombinant AtCutAp showed a specific interaction with Cu(II) ions, both with copper ions bound to IDA-agarose resin and with copper ions in solution. The IDA-agarose assay has been used to characterize other proteins that bind transition metals, but the target proteins usually interact with several metals in this assay (Dykema et al., 1999). Therefore, the absolute specificity for Cu that is observed with AtCutAp stands out. Truncated AtCutAp showed more promiscuity in binding to metal-saturated IDA-agarose than did fully spliced AtCutAp, suggesting that the C-terminal region of the protein might help in determining metal specificity.
Analysis of AtCUTA knockout lines indicated that AtCUTA is not essential for copper tolerance or accumulation. Nevertheless, AtCUTA may be involved in chloroplast copper homeostasis. The specific interaction with Cu(II) ions suggests that AtCutAp occupies a role other than a typical copper chaperone, as copper chaperones have been characterized as Cu(I)-binding proteins (Harrison et al., 2000). Thus, if AtCutAp does serve as a copper chaperone, it would be a novel Cu(II)-binding chaperone. AtCutAp does not contain a consensus metal-binding motif (such as CXXC) commonly found in metal-binding proteins, such as copper chaperones or transporters (Fox and Guerinot, 1998). Furthermore, the single cysteine residue in the mature protein is insufficient to have sequence characteristics of a metallothionein (Goldsbrough, 1999; Zhou and Goldsbrough, 1995). Rather than having a role in metal sequestration, the observed interaction of AtCutAp and Cu(II) may suggest a regulatory or signaling role for AtCutAp, as has been suggested for the mammalian cellular prion proteins, which bind Cu(II) with a comparable affinity (Stöckel et al., 1998). CutA proteins lack the octarepeat peptides found in prion proteins, which have been reported to be involved in Cu(II) binding (Stöckel et al., 1998). Therefore, AtCutAp may bind to copper by a novel copper-binding motif. A regulatory or signaling role may also be possible for bacterial CutAp because E. coli mutants in the operon containing CUTA are more sensitive to increased copper levels than the wild type and accumulate increased amounts of divalent cations (Fong et al., 1995).
In conclusion, AtCUTA is expressed in all major tissues; it is alternatively spliced and codes for a chloroplast-localized copper-binding protein. AtCutAp is a member of a class of copper-binding proteins whose function has not been determined yet. AtCutAp is found in the chloroplast, an organelle containing peptides and enzymes that require copper for proper function. Interestingly, there is a human CUTA homolog localized to neural synapses in the brain (Navaratnam et al., 2000). Thus, the human CUTA, like AtCutAp, has a post-translational routing outside the cytosol. Further characterization of AtCUTA will include detailed analysis of phenotypes in plants with altered expression levels and may provide insight into chloroplast copper homeostasis in plants as well as in other types of organisms.
Vectors and clones
GenBank accession numbers for fully spliced and intron-containing AtCUTA are AF327524 and AF327525, respectively. AtCUTA EST clones were obtained from the Ohio State University Arabidopsis Biological Resource Center as SalI/NotI inserts in pSPORT vectors (Life Technologies, Rockville, MD, USA). pBluescriptKS is a Stratagene (San Diego, CA, USA) product; pET11-d was obtained from Novagen (Madison, WI, USA). Vector pSP64-PC was described by Smeekens et al. (1986). pMOG 18 and pMOG23 were described by Sijmons et al. (1990).
Murashige and Skoog (MS) salts were obtained from Sigma (St Louis, MO, USA). The RNeasy RNA isolation kit and Qiaquick gel extraction kit were obtained from Qiagen (Valencia, CA, USA). The high-fidelity PCR kit (HF-PCR) and Titan RT-PCR kits were obtained from Roche (Palo Alto, CA, USA). The Ampliscribe in vitro transcription kit was purchased from Epicentre (Madison, WI, USA). M-MlV reverse transcriptase and wheat germ extract in vitro translation reagents were obtained from Promega (Madison, WI, USA). DEAE Sepharose and 35S-labeled methionine (1000 Ci mmol−1) were obtained from Amersham–Pharmacia (Uppsala, Sweden). Imino-diacetic acid agarose was obtained from Novagen (Madison, WI, USA). All other reagents were of the highest quality commercially available. A. thaliana ecotype Columbia seed was a gift from Dr June Medford (Colorado State University). Antibodies to AtSEC12 protein and p28 protein were gifts from Dr Sebastian Bednarek (University of Wisconsin). Antibodies to PMO35 were a gift from Dr Tom Elthon (University of Nebraska). The Salk Institute Genomic Analysis Laboratory provided one sequence-indexed Arabidopsis T-DNA insertion mutant with an insertion in the second exon of AtCUTA, line # SALK 029578 (http://signal.salk.edu/index.html); the second T-DNA insertion line had an insert in the third intron and was found in the Wisconsin KO collection in seed pool CSH169-94 (http://www.biotech.wisc.edu/Arabidopsis).
Two cDNA clones were obtained from the Ohio State University Arabidopsis Biological Resource Center. The clones were analyzed by restriction digest and PCR in which general sequencing primers M13(-20) and M13rev amplified the inserts. The clone 92B3T7 insert was very short (<200 base pairs) and was determined to be incomplete. It was not used for further analysis. DNA was sequenced on both strands (M13(-20) and M13rev primers and two internal primers) by the dideoxy method. To obtain a fully spliced cDNA clone, cDNA from 14-day-old seedlings grown on 0.5× MS medium was generated from isolated RNA using RT-PCR. The following optimized PCR conditions were used: hot start at 94°C for 2 min; melting at 94°C for 30 sec; annealing at 55°C for 45 sec; extension at 68°C with cycles 1–10, 2 min 30 sec; cycles 11–20, 4 min; cycles 21–30, 5 min; and cycles 31–35, 6 min. Primers A (5′-CACAGTGATGGCTTCGTCTCT-3′), B (5′-TCATTTGACGAACTGGTCCAGT-3′), and C (5′-CGCATTACTCATGTTTTTATATG-3′) were designed based on sequence data. The PCR product for a fully spliced AtCUTA transcript (700 base pairs) was extracted from a 1% agarose gel and was modified to introduce NcoI and BamHI sites (HF-PCR). Primer Cl1 (5′-GGGGTACCATGGCTTCGTCTCTCACCACTAG-3′) was used to introduce KpnI and NcoI sites upstream, and primer Cl2 (5′-GGGGTACCGGATCCTGGTCCAGTATTACACTTCACA-3′) introduced KpnI and BamHI sites downstream of the coding sequence. The PCR products were cloned into the KpnI site of pBluescriptKS. This clone was identified as pBS-spl (AtCutA fully spliced) and was sequenced to confirm fidelity. The insert was then subcloned from pBS-spl into the pET11d expression vector and pMOG18 as a BamHI/NcoI restriction fragment to generate pET-spl and pMOG18-spl, respectively. The Cl1 and Cl2 primers were also used to introduce restriction sites into the EST clone 213A16T7. This cDNA was then subcloned to pBSKS to generate pBS-int and then to pET11d to generate pET-int.
HF-PCR was employed to remove the putative chloroplast targeting sequence from both pBS-spl and pBS-int. The Mod1 (5′-GGGGTACCATGGAGGAGAGCAGCAAAACTG-3′) upstream primer was used to introduce KpnI and NcoI restriction sites, while the downstream primer was the Cl2 primer used in cloning. These products were cloned as KpnI fragments to pBluescriptKS vector to generate pBS-M5 (fully spliced) and pBS-M10 (intron containing). pBS-M5 and pBS-M10 were subcloned to pET11-d as BamHI/NcoI restriction fragments to generate pET-M5 and pET-M10. The targeting sequence of AtCUTA was amplified from the pBS-spl clone using a primer for the upstream SP6 promoter and the Fus1 (5′-GCGCAAGCTTGCAACAGCGATGGCTTCGTCTCTCAC-3′) downstream primer to introduce an NcoI restriction site. This restriction fragment was inserted as a PstI/NcoI fragment in the PstI/NcoI-digested pSP64 vector, which contained the mature plastocyanin sequence, to generate pCAPC.
Overexpression of AtCutAp in Arabidopsis
The fully spliced AtCUTA cDNA was subcloned from pMOG18, along with the Cauliflower Mosaic Virus 35S promoter and enhancer sequence to pMOG23, to generate pMOG23-spl. The pMOG23-spl binary vector was transformed to Agrobacterium tumafaciens strain C58C1 (An et al., 1988). A. thaliana (ecotype Columbia) plants were transformed with this strain by the floral dip method (Clough and Bent, 1998). Transgenic lines were selected for kanamycin resistance. Twelve independent lines were confirmed by PCR to include the AtCUTA construct. None of these lines showed an obvious phenotype. Western blotting was used to identify lines with increased expression of AtCutAp. Of four homozygous lines with significant increased expression, one line was chosen for the experiment shown in Figure 7(c).
Copper tolerance in AtCUTA-knockout
To examine the tolerance of AtCUTA-knockout seedlings to excess copper, WT and knockout seeds were surface-sterilized essentially as described by Pilon-Smits et al. (1999). For each treatment, approximately 25 seeds of each line were sown in a grid pattern in a single Magenta box (Sigma, St Louis, USA) on medium containing 0.5× MS salts and vitamins (Murashige and Skoog, 1962), 10 g l−1 sucrose, and 4 g l−1 agargel (Sigma), with or without 30 µm CuSO4. After 14 days at 25°C and 16-h light/8-h dark photoperiod, individual seedlings were harvested and weighed.
Expression in E. coli and protein isolation
The pET-M5 and pET-M10 plasmids were transformed to E. coli BL21-DE3 (Studier et al., 1990). Cultures were grown in 500 ml batches of LB medium at 37°C with vigorous shaking to an OD600 of 0.3 and were induced by adding IPTG to 0.3 mm. Incubation continued to an OD600 of 1.7 when cells were collected by centrifugation at 4000 g in a fixed angle rotor at 4°C. Cells were subjected to osmotic shock to isolate periplasmic proteins, essentially as described by Neu and Heppel (1965). Proteins were further purified at 4°C on a 1.5 cm × 14 cm DEAE Sepharose column equilibrated with 25 mm Tris–HCl (pH 7.5) + 10% (w/v) glycerol. After allowing the protein to bind, the column was washed with 10 ml of the same buffer. Elution began with 10 ml of 200 mm NaCl in 25 mm Tris–HCl (pH 7.5) + 10% (w/v) glycerol and continued in a continuous gradient from 250 to 500 mm NaCl in the same buffer for 500 ml. The recombinant AtCUTA protein expressed from pET-M5 eluted at 425–475 mm NaCl and was free from contaminants as observed by SDS–PAGE with Coomassie Brilliant Blue staining. Peak fractions were pooled for a yield of approximately 2.5 mg protein/500 ml of original culture. The recombinant AtCUTA protein expressed from pET-M10 eluted from the column at 275–300 mm NaCl. Peak fractions were pooled for a yield of approximately 2.0 mg protein/500 ml of original culture. Gel filtration used a Sephacryl S-200 column, 60 cm × 1.6 cm (Amersham–Pharmacia) in 10 mm NaPO4 (pH 7.5), 150 mm NaCl. The column was calibrated with the following standards: IgY (180 kDa), bovine serum albumin (67 kDa), ovalbumin (43 kDa), chymotrypsin (25 kDa), and RNAseA (14 kDa).
Immobilized metal binding chromatography
Purified full-length AtCutAp and truncated AtCutAp were analyzed using a metal-saturated IDA-agarose (Novagen, Madison, WI, USA). One hundred microliter aliquots (settled volume) of resin were stripped by incubation in 1 ml of 100 mm EDTA, 500 mm NaCl, 20 mm Tris (pH 7.5), washed three times with sterile distilled water, and saturated with one of the following 400 mm solutions in water: NaCl, CuSO4, NiSO4, ZnSO4, MnSO4, or CdSO4. Resins were washed first three times with 1 ml of distilled water and then three times with 1 ml of the binding buffer (50 mm potassium phosphate (pH 7.5), 250 mm NaCl). Fifty microliters of each resin was nitric-acid-digested and analyzed by ICP-AES (Zarcinas et al., 1987) to confirm metal saturation. The remaining 50 µl of resin portions was then incubated with 20 µg of the purified AtCutAp. Lysozyme (Sigma–Aldrich, St Louis, MO, USA) and bovine serum albumin (Sigma–Aldrich) were used as a controls) in a total of 300 µl of buffer for 10 min on ice. The resin was collected by centrifugation (2500 g for 45 sec in a bench top microfuge), and the supernatant was collected as the unbound fraction. Three additional washes with the binding buffer were followed by a low-pH wash (50 mm KPi (pH 5.5), 250 mm NaCl) whereby the supernatant was collected as the pH 5.5 wash. The final wash was with 100 mm EDTA (50 mm KPi (pH 6.0), 250 mm NaCl) to release metal ions and associated protein from the resin. Protein in each fraction was concentrated with a final concentration of 0.0015% w/v deoxycholic acid and 15% w/v trichloroacetic acid. Precipitations were incubated on ice for 10 min and pelleted by centrifugation at 13 000 g and 4°C for 10 min. Pellets were washed with ice-cold 100% acetone and dried in a vacuum desiccator before dissolution in SDS–PAGE sample buffer (125 mm Tris–HCl (pH 6.8), 10% w/v glycerol, 2% SDS, 10 mm dithiothreitol). Each fraction was analyzed by SDS–PAGE followed by staining with Coomassie Brilliant Blue.
Copper affinity assay
To determine Cu(II) binding in solution, 800 µg of full-length or truncated AtCutAp was incubated for 5 min at room temperature in 10 mm Tris (pH 7.5), 100 mm NaCl, 0–1.5 mm CuSO4 in a volume of 500 µl. The solution was loaded into a Nanosep centrifugal device (Pall Gelman Laboratory, Ann Arbor, MI, USA) with a molecular weight cut-off of 10 kDa. The device was centrifuged in a bench top microcentrifuge for 10 min at 4000 g after which less than 50 µl remained. The retentate was then diluted back to 500 µl in 10 mm Tris (pH 7.5), 100 mm NaCl and was centrifuged as before. This was repeated two more times, and the final retentate was brought back to 500 µl in the same buffer. Ten microliters was used for protein determination by Bradford assay. TCA (100% w/v) was added to the remaining retentate to a final concentration of 9.2%, and the reaction was placed on ice for 10 min. The tube was then centrifuged at 14 000 g for 10 min at 4°C to collect the protein. The supernatant containing released copper ions was then neutralized with 80 µl of 6 m NaOH and 100 µl of 1 m untitrated Tris buffer. To this, 10% (w/v) ascorbic acid was added to 0.6% final concentration to reduce Cu(II) ions to Cu(I) ions. Bathocuproindisulfonic acid (BCS; Sigma) was added to a final concentration of 0.65 mm to determine Cu(I) concentration by OD485 (Zak, 1958) with a standard curve from 0 to 200 µm Cu(I). Cu(II) binding by lysozyme as a negative control and BSA as a positive copper-binding control was performed simultaneously under the same conditions. Cu(I) binding was obtained by reducing 1 mm CuSO4 in 1% (w/v) ascorbic acid and by using this as a stock to repeat the copper-binding test as described. A second Cu(I)-binding test was performed with the addition of 1 mm EDTA to remove any Cu(II) ions that might be present (Rae et al., 1999). Various buffer conditions were tested to determine suitable conditions for Cu(II) affinity tests. Cu(II) affinity was determined for AtCutAp by using the described procedure, with CuCl2 concentrations of 0, 10, 20, 30, 40, 60, and 80 µm in 50 mm MES (pH 6.0), 150 mm NaCl (Stöckel et al., 1998). Standard curves for spectrophotometric determination of Cu content with BCS were made by adding TCA, NaOH, Tris, ascorbic acid, and BCS, as described for experimental tests.
To analyze the expression of AtCUTA, total RNA was isolated from roots, stems, leaves, and flowers by the TRIzol reagent method (Life Technologies, Carlsbad, CA, USA). Poly(A)+ RNA was isolated using the oligotex mRNA kit (Qiagen, Valencia, CA, USA) according to the manufacturer's instructions. Twelve micrograms of poly(A)+ RNA was electrophoresed on a 1% agarose gel containing 4% formaldehyde, transferred to a nylon membrane, and was probed with a 32P-labeled 700 bp AtCUTA cDNA. Pre-hybridization and hybridization were performed at 65°C in a solution containing 0.5 m sodium phosphate and 0.7% SDS (w/v). Following hybridization, the membrane was washed with 0.1× SSC and 0.1% SDS at 65°C, wrapped in a cling film, and placed in a phosphorimaging cassette (Molecular Dynamics). The screen was scanned in a STORM 840 phosphorimager (Molecular Dynamics, Sunnyvale, CA, USA).
The hybridized probe was removed by washing the membrane three times at 65°C, 5 min each, with the stripping buffer (0.1× SSC, 1% SDS, 40 mm Tris (pH 7.5)) containing 50% (v/v) formamide and once with the stripping buffer without formamide, as described by Ausubel et al. (1998). The membrane was re-probed with a 32P-labeled 2 kb cDNA fragment that encodes for Arabidopsis ubiquitin. Pre-hybridization, hybridization, and washing were performed as before.
In vitro transcription and translation
Plasmids pSP64-PC (preplastocyanin; Smeekens et al., 1986) and pCAPC (this study) were linearized by EcoRI and PvuII, respectively, and were transcribed in vitro using the Ampliscribe SP6 transcription kit with Cap analog (Epicentre, Madison, WI, USA) according to the manufacturer's instructions. In vitro transcribed messenger RNA (0.2 μg) from each transcription was used in a wheat germ extract kit (Promega) with 35S-methionine (25 µCi/50 µl reaction). Translations were analyzed with SDS–PAGE and were visualized using a STORM phosphorimaging system (Amersham Pharmacia, Uppsala, Sweden).
Chloroplast import and fractionation
Pea chloroplasts (from cultivar Little Marvel) were isolated using procedures detailed by Pilon et al. (1992). Five hundred microliters of import reactions contained 30 µl of translation mixture (see above) incubated with the isolated chloroplasts as described by Pilon et al. (1992). Procedures to obtain crude stromal and thylakoid fractions are detailed by Smeekens et al. (1986).
For selective protease treatment experiments, incubated chloroplasts (150 µg chlorophyll) were re-isolated by centrifugation through a 40% (w/v) Percoll cushion and were re-suspended in 600 µl of the import buffer. Half of these were kept on ice and subjected to hypotonic lysis to obtain a stromal fraction. The remaining chloroplast portion was subdivided into three 100 µl fractions, which were treated with either thermolysin or trypsin (Lübeck et al., 1997), or were left untreated. Thermolysin treatment (125 µg ml−1) for chloroplasts was at 4°C for 15 min. The thermolysin-treated chloroplast reaction was stopped by the addition of EDTA in import buffer to a concentration of 100 mm. Trypsin treatment (235 µg ml−1) for chloroplasts was for 30 min at 20°C and was stopped by the addition of trypsin inhibitor to 920 µg ml−1. Intact and protease-treated chloroplasts were collected by centrifugation (3 min, 1200 g) and dissolved in SDS–PAGE sample buffer. Similarly, the stromal fraction was divided into three fractions for thermolysin, trypsin, or control treatment at the same protease concentrations and conditions. Stromal protease treatment was stopped by TCA precipitation (0.0015% w/v deoxycholic acid and 15% w/v trichloroacetic acid) on ice. Fractions in TCA were centrifuged at 13 000 g for 30 min, and the supernatants were removed and washed with 100% acetone. Pellets were air-dried and re-suspended in 40 µl of sample buffer. Precursor and import fractions were separated by SDS–PAGE, and the dried gel was analyzed with a STORM 840 phosphorimaging system.
Arabidopsis chloroplast isolation
To determine localization of AtCutAp in Arabidopsis, chloroplasts were isolated essentially as described by Rensink et al. (1998). Chloroplasts were isolated from rosette leaves (3–4 weeks old) of either wild-type or AtCutAp-overexpressing line SPL-6B plants. Chloroplasts were precipitated by centrifugation and were re-suspended in 330 mm sorbitol, 50 mm Hepes-KOH (pH 8). One-half of each fraction was again precipitated and lysed by re-suspension in 300 µl of 10 mm Tris–HCl (pH 8). After 2 min on ice, an equal volume of 660 mm sorbitol, 100 mm Hepes-KOH (pH 8) was added. Thylakoid membranes were precipitated by centrifugation and were re-suspended in 330 mm sorbitol, 50 mm Hepes-KOH (pH 8). Total plant homogenate, chloroplasts, stroma (including envelopes), and thylakoid fractions were precipitated in 80% acetone, incubated on ice for 30 min, and collected by centrifugation at 12 000 g for 20 min. Pellets were air-dried and were suspended in SDS–PAGE sample buffer. Proteins were separated by 15% SDS–PAGE, with each sample containing protein equivalent to 3 µg chlorophyll, as determined by OD652 (Bruinsma, 1961). Proteins were transferred to a nitrocellulose membrane by Western blot and were visualized by immunodetection using AtCutAp-specific antiserum.
Protein concentrations were determined by a protein assay (Bradford, 1976) that was corrected to the absorption at 280 nm because of the presence of aromatic amino acids (Perkins, 1986).
Statistical analyses for tolerance and accumulation were performed using the statistical software program jmp-in from the SAS Institute (Cary, NC, USA).
For AtCutAp antibody production, the purified protein was dialyzed to 25 mm Na-phosphate (pH 7.0), 150 mm NaCl. Polyclonal antibodies were raised in chickens at a commercial facility (Aves Laboratories, Tigard, OR, USA).
This research was supported by grant number MCB-0091163 from the US National Science Foundation to M.P. and grant number G8A11586 to E.A.H. P.-S. from the US Environmental Protection Agency.