AtHMA1 Is a Thapsigargin-sensitive Ca2+/Heavy Metal Pump*

The Arabidopsis thaliana AtHMA1 protein is a member of the PIB-ATPase family, which is implicated in heavy metal transport. However, sequence analysis reveals that AtHMA1 possesses a predicted stalk segment present in SERCA (sarcoplasmic/endoplasmic reticulum Ca2+ ATPase)-type pumps that is involved in inhibition by thapsigargin. To analyze the ion specificity of AtHMA1, we performed functional complementation assays using mutant yeast strains defective in Ca2+ homeostasis or heavy metal transport. The heterologous expression of AtHMA1 complemented the phenotype of both types of mutants and, interestingly, increased heavy metal tolerance of wild-type yeast. Biochemical analyses were performed to describe the activity of AtHMA1 in microsomal fractions isolated from complemented yeast. Zinc, copper, cadmium, and cobalt activate the ATPase activity of AtHMA1, which corroborates the results of metal tolerance assays. The outcome establishes the role of AtHMA1 in Cd2+ detoxification in yeast and suggests that this pump is able to transport other heavy metals ions. Further analyses were performed to typify the active Ca2+ transport mediated by AtHMA1. Ca2+ transport displayed high affinity with an apparent Km of 370 nm and a Vmax of 1.53 nmol mg–1 min–1. This activity was strongly inhibited by thapsigargin (IC50 = 16.74 nm), demonstrating the functionality of its SERCA-like stalk segment. In summary, these results demonstrate that AtHMA1 functions as a Ca2+/heavy metal pump. This protein is the first described plant P-type pump specifically inhibited by thapsigargin.

Plant Ca 2ϩ and heavy metal-ATPases belong to the superfamily of P-ATPases (1,2). Their common characteristic is the presence of a phosphorylated intermediary in the catalytic cycle. In Arabidopsis, the heavy metal-ATPases belong to the P IB subfamily and normally have eight predicted transmembrane domains, whereas the Ca 2ϩ -ATPases are part of the P IIA and P IIB subfamilies and have ten predicted transmembrane domains (2). A common feature among the P IB -ATPases is the presence of a CPX motif, which is though to play a role in metal translocation, as well as putative metal-binding domains located at the amino or carboxyl terminus (3). On the other hand, P IIB ATPases have a calmodulin-binding domain that regulate their activity; however, the P IIA do not have this domain (4). The most important difference between the calcium and heavy metal-ATPases is their substrate specificity (1). However, in vitro metal transport studies performed with membrane fractions isolated from seedlings suggest that one or more members of the superfamily of P-ATPases are capable of transporting calcium and heavy metals. The studies showed competitive inhibition between active transport of heavy metals (such as copper and cadmium) and calcium. 2 Interestingly, both transport activities were inhibited by the sesquiterpene lactone thapsigargin, a potent and specific inhibitor of SERCA 3 -type pumps (5)(6)(7). To date no one has described plant ion pumps that transport calcium and heavy metals in biochemical terms, nor have scientists described genes encoding for ion pumps inhibited by thapsigargin or plant mutants with thapsigargin-sensitive/tolerant phenotypes. Based on the results of biochemical assays suggesting the existence of thapsigargin-sensitive Ca 2ϩ /heavy metal-ATPase, we searched for potential candidate proteins in the Arabidopsis genome. This was made possible by the highly conserved "stalk segment" or S3 sequence adjacent to the third transmembrane segment of the SERCA pumps (8,9) composed of amino acids DEF-GEQLSK (5)(6)(7). This sequence was almost complete and was annotated as a stalk segment (using the topology prediction software ARAMEMNON) in the Arabidopsis heavy metal pump AtHMA1 (At4g37270). This pump belongs to the subclass of zinc/cobalt/cadmium/lead-ATPases and is the most divergent metal pump of the Arabidopsis P IB -ATPases (1, 2, 10, 11) (see Fig. 1). It lacks an amino-terminal heavy metal-binding domain, such as GMXCXXC or GICC(T/S)SE, which is often found in other members of the group. It has an intramembranous SPC instead of the CP(C/H/S) motif located at the putative metal transporting site of P IB -ATPases (11,12). The pump possesses other structural characteristics related to heavy metal binding and transport, such as a poly-H motif commonly found in zinc-binding proteins, several C and CC pairs, and a HP dipeptide (11,12). Initial studies performed by Higuchi and Sonoike (13) show that an Arabidopsis disruption mutant of AtHMA1 is sensitive to high concentrations of zinc. Recently, it was demonstrated that AtHMA1 is localized in the chloroplast envelope, and Arabidopsis insertional mutants exhibit a lower chloroplast copper content and a diminution of the total chloroplast superoxide dismutase, activity suggesting a role for AtHMA1 in copper homeostasis (14). At this point, two key questions about the functions and ion specificities of AtHMA1 must be addressed. (i) Can AtHMA1 transport Ca 2ϩ ? (ii) In which heavy metal homeostatic mechanisms is AtHMA1 involved? We expressed AtHMA1 in yeast (Saccharomyces cerevisiae) in order to perform functional complementation assays. We demonstrated that AtHMA1 complements the mutant phenotype displayed by two Ca 2ϩ transport-deficient yeast strains. In vitro calcium transport assays showed that AtHMA1 promotes thapsigargin-sensitive Ca 2ϩ transport. AtHMA1 also complements a Cd 2ϩ -hypersensitive yeast mutant and confers high Cd 2ϩ tolerance onto wild-type yeast. Cadmium-stimulated ATPase activity of AtHMA1 confirms that this pump plays a role in cadmium transport. Expression of AtHMA1 also confers high zinc, copper, and cobalt tolerance onto yeast, and ATPase activity is also stimulated by the addition of these metals. Taken together, our results strongly suggest that AtHMA1 is a metal pump capable of transporting a wide range of ions including calcium. As such, AtHMA1 may represent a key element in the mechanism used to overcome ion deficiency or toxicity in Arabidopsis thaliana.
Expression of AtHMA1 in Yeast-AtHMA1 expression in transformed yeast was verified using RT-PCR. Transformed yeast was grown on a selective medium until an A 600 of 0.6. Cells were then collected by centrifugation and used to isolate total RNA following the Chomczynski phenol-chloroform extraction method (21). The first strand of cDNA was prepared using a RevertAid TM first strand cDNA synthesis kit (Ferrmentas) with oligo(dT) and 1 g of total RNA. After reverse transcription of mRNA to cDNA, a 565-bp fragment of the AtHMA1 cDNA was amplified by PCR using the following primers: 5Ј-ATGATGTTAACTGGGGACC-3Ј (sense) and 5Ј-CTAAT-GTGCAGAGCTTAAACTGTTGCTGCTGCTACT-3Ј (antisense). Amplification of ACT1 cDNA was used as an internal control in all reactions (22). The PCR products were separated by agarose gel electrophoresis.
Functional Complementation Assays-Functional complementation of Ca 2ϩ transport-defective mutants K616 and ⌬pmr1 was carried out as described previously (19). Mutant and wild-type yeast strains transformed with pGPD426 or pGPD426-AtHMA1 were streaked onto growth plates lacking uracil and supplemented with 10 mM EGTA. Ca 2ϩ -depleted plates were then incubated for up to 3 days at 30°C. For functional complementation of the cadmium-hypersensitive yeast mutant ⌬ycf1, the complementation was carried out by streaking transformed yeast onto growth plates lacking uracil and supplemented with 70 M CdCl 2 . High Cd 2ϩ plates were incubated for up to 5 days at 30°C.
Metal Toxicity Tests-To test sensitivity to high cadmium concentrations in ⌬ycf1 yeast transformed with AtHMA1, cells grown to log phase were diluted to an A 600 of 0.1 and incubated in the presence of 70 -200 M CdCl 2 . Growth at 30°C was monitored by the change in A 600 every 3 h for 21 h, and the rate of exponential growth of yeast cultures was expressed as generation time in hours, i.e. the doubling time of each yeast population. Sensitivity to other transition metals was tested by growing transformed ⌬ycf1 on solid media supplemented with 6 mM CoCl 2 , 4 mM CuSO 4 , or 28 mM ZnCl 2 . Plates were incubated for 5 days at 30°C. Additionally, the capacity of AtHMA1 to confer cadmium tolerance to wild-type yeast (W303) was tested by growing transformed cells in liquid medium with 70 -200 M CdCl 2 , and the A 600 of cultures was followed for 5 days.
Carboxypeptidase Y (CPY) Detection Assay-Transformed K616 and wild-type cells were grown in 3 ml of selective medium for 2 days at 30°C. To obtain yeast lysate samples, cells were resuspended in 200 l of SUMEB buffer (1% SDS, 8 M urea, 10 mM MOPS, pH 6.8, 10 mM EDTA, and 0.01% bromphenol blue) plus 100 l of 500-m glass beads (Sigma) and then vortexed three times for 1 min. The mixture was incubated for 10 min at 65°C and centrifuged at 2,000 ϫ g. The supernatant was collected. Two l of supernatant (or yeast lysate) and 600 l of growth media were transferred simultaneously onto a nitrocellulose membrane using a dot-blot device (Bio-Rad). The membrane was then washed briefly with water and blocked with 5% milk in TBS (10 mM Tris-HCl, pH 7.5, 150 mM NaCl, 0.1% Tween 20). Secreted CPY was detected by monoclonal anti-CPY antibody (1:1,000) (Molecular Probes). Anti-mouse IgG antibody-horseradish peroxidase conjugate (1: 10,000; Molecular Probes) was used as secondary antibody.
Isolation of Yeast Membranes-Membranes were prepared from yeast cells as described previously (23). 2-5 ml of stationary phase cultures of transformed yeast was diluted into 300 ml of appropriate selective medium, grown for 24 h at 30°C to an A 600 of ϳ0.8, and collected by centrifugation. After being washed with distilled water, the cells were resuspended in 1 volume of a buffer solution containing 10% sucrose, 25 mM Hepes-bis-Tris propane, pH 7.5, 2 mM MgCl 2 , 2 mM dithiothreitol, 1 mM EGTA, and a mixture of protease inhibitors (1 mg/ml aprotinin, 1 mg/ml leupeptin, 1 mg/ml pepstatin, and 1 mM phenylmethylsulfonyl fluoride). One volume of 500-m glass beads was added to the mixture, and yeast cells were disrupted by vortexing at maximum speed for 2 min. The homogenate was centrifuged at 5,000 ϫ g for 5 min at 4°C. The supernatant was collected and centrifuged at 110,000 ϫ g for 1 h. The resulting pellet was resuspended in 200 l of a buffer solution containing 10% sucrose, 20 mM Hepes-bis-Tris propane, pH 7.0, 1 mM dithiothreitol, and the mixture of protease inhibitors described above. The membrane preparations were aliquoted and stored at Ϫ80°C until needed. Protein content was measured using the bicinchoninic acid method (Pierce Chemical) with bovine serum albumin as standard.
ATPase Activity Assays-These assays were carried out as described previously (24). Yeast vesicles (50 g protein/ml) were mixed with a reaction buffer containing 50 mM Tris, pH 7.5, 3 mM MgCl 2 , 3 mM ATP, 20 mM Cys, 1 mM dithiothreitol, and each metal (at 1 M) listed in Fig. 6A. The reaction mixture was incubated for 15 min at 30°C. Released inorganic phosphate was determined colorimetrically (25). Background activity measured in the absence of metals was less than 10% of the highest metal-stimulated activity and was subtracted from the activity determined in the presence of metals. The results are the average of four independent experiments with measurements performed in triplicate. The relative activity was standardized by setting the highest value obtained from the measurements (48.15 nmol g Ϫ1 ) at 100% relative activity. All subsequent values were then adjusted accordingly and the S.E. was calculated. 45 Ca 2ϩ Uptake Assays-Ca 2ϩ uptake by membranes of transformed yeast was assayed measuring 45 Ca 2ϩ accumulation using the filtration method (26). Membrane samples were suspended (0.3-0.6 mg protein/ml final concentration) in a reaction buffer (3 ml) containing 250 mM sucrose, 25 mM Hepes-bis-Tris propane (pH 7.0), 10 mM KCl, and 0.4 mM NaN 3 . Free Ca 2ϩ concentration in the reaction buffer was set at desired values using variable amounts of 45 CaCl 2 (containing 2.4 Ci of 45 Ca 2ϩ mmol Ϫ1 CaCl 2 , PerkinElmer Life Sciences) and EGTA. Free Ca 2ϩ concentrations were calculated using the WinMaxC 2.05 computer program. Vesicles were incubated in the buffer at 25°C for 1 min, and transport was initiated (time zero) by adding a mixture of ATP and MgCl 2 to final concentrations of 2 and 3 mM, respectively. Samples of 200 l were taken at different times and filtered through 0.45-m nitrocellulose membranes (Millipore, Bedford, MA). The filters were rinsed immediately with 5 ml of ice-cold STOP buffer contain-ing 250 mM sucrose, 2.5 mM Hepes-bis-Tris propane, pH 7.0, and 0.2 mM CaCl 2 . The 45 Ca 2ϩ radioactivity associated with the filters was determined by liquid scintillation counting. Ca 2ϩ uptake was expressed as nmol Ca 2ϩ /mg protein in the presence and absence of ATP-Mg. To determine the K m for Ca 2ϩ uptake activity, the reaction mixture contained 500 M EGTA and various amounts of Ca 2ϩ To provide the desired range of free Ca 2ϩ concentration (from pCa 5.5-7.2). ATP-dependent Ca 2ϩ uptake (nmol mg Ϫ1 min Ϫ1 ) was defined as the difference between the 45 Ca 2ϩ retained in the filters following incubation in the presence and absence of MgATP. The effect of thapsigargin (TG) was tested by adding the compound to a final concentration in the reaction mixture of 0.1 M at 15 min before the uptake assay was started. 45 Ca 2ϩ uptake and ATPase activity assays were performed at least four times in triplicate.

RESULTS
Growth Inhibition of ⌬pmr1 and K616 on Ca 2ϩ -deficient Media Is Overcome by the Expression of AtHMA1- Fig. 1 shows a comparison between AtHMA1 and SERCA1. Like all P-type ATPases, both enzymes are transmembrane proteins. SERCA1 possesses ten transmembrane helices, like the vast majority of characterized Ca 2ϩ -ATPases. AtHMA1, instead, has seven predicted transmembrane helices (according to the topology prediction software ARAMEMNON). This differs from the characteristic eight predicted transmembrane helices presented in the rest of plant metal-ATPases recruited in the P IB -ATPase subfamily (2). On the other hand, AtHMA1 and SERCA1 share the highly conserved Asp in the signature sequence DKTGT that is phosphorylated during the catalytic cycle of P-type ATPases (1,27,28). Other common features of P-type ATPases are slightly modified in AtHMA1. For example, the transduction domain TGES (29) in AtHMA1 is TGEX. The GDGXNDXP and TGD motifs involved in ATP binding (27, 28, 30 -33) in AtHMA1 are GEGINDAP and TGD, respectively. AtHMA1 also presents motifs involved in heavy metal sensing, binding, and transport. These include several C and CC pairs (34 -36), a poly-H domain located at the amino terminus, an HP dipeptide (11,12,35,37), the site SPC essential for metal transport, and a HEGG motif (11,12,14). It is noteworthy that AtHMA1, as a heavy metal pump, possesses the signature sequence DEFGENYSK (black circles), which is very similar to the DEFGEQLSK sequence involved in the inhibitory effect of thapsigargin in SERCA-type Ca 2ϩ pumps (6,38).
To determine whether AtHMA1 plays a role in Ca 2ϩ homeostasis, we performed functional complementation assays using mutant yeast deficient in Ca 2ϩ transport. Heterologous expression of the complete AtHMA1 cDNA was performed under the control of the glyceraldehyde-3-phosphate dehydrogenase gene (GPD) promoter (39) and was verified using RT-PCR ( Fig. 2A). Fig. 2B shows that parental wild-type yeast strains YR98 and W303 grew normally on media containing 10 mM EGTA. As demonstrated previously (26), ⌬pmr1, a strain in which the secretory pathway Ca 2ϩ -ATPase PMR1 is disrupted (40), was unable to grow in media depleted of Ca 2ϩ in cells transformed with the empty vector. When this mutant was transformed with pGPD426-AtHMA1, growth on Ca 2ϩ -deficient medium was restored. AtHMA1 expression also complemented the Ca 2ϩ depletion sensitivity of the triple mutant K616, which lacks both endogenous Ca 2ϩ pumps (PMR1 and PMC1) and calcineurin (CNB1) function (19,41,42). Similar to ⌬pmr1, K616 transformed with the empty vector did not grow on a medium containing very low Ca 2ϩ (10 mM EGTA). Nevertheless, the triple mutant transformed with pGPD426-AtHMA1 became tolerant of this condition (Fig. 2B), supporting the hypothesis that AtHMA1 encodes a functional Ca 2ϩ pump.
AtHMA1 Reverts the Missorting Phenotype Displayed by K616-Given that yeast mutants lacking the secretory pathway Ca 2ϩ -ATPase PMR1 partially secrete carboxypeptidase Y, an enzyme normally destined to the vacuole (40), we evaluated the effect of AtHMA1 expression on this characteristic phenotype. CPY secreted due to missorting was detected using a monoclonal anti-CPY antibody. CPY was detected in a culture media sample recovered from K616 cells transformed with the empty vector as shown in Fig. 2C. However, CPY was not detected in the culture media of K616 cells transformed with pGPD426-AtHMA1. In both cases, intracellular accumulation of CPY was verified in cell lysates. The wild-type strain was used as a negative control, as CPY is not secreted under normal conditions (26).

Active Ca 2ϩ Transport in K616 Membrane Vesicles Is Increased upon Expression of AtHMA1 and Is Inhibited by TG-
Our results indicate that mutant yeast strains K616 and ⌬pmr1 are functionally complemented by AtHMA1. To determine whether this is due to a direct increase in Ca 2ϩ transport, we assayed in vitro ATP-dependent 45 Ca 2ϩ uptake into intracellular membrane vesicles isolated from K616 cells transformed with the empty vector or a vector carrying AtHMA1 cDNA. As expected, K616 membranes exhibited no significant increase in total 45 Ca 2ϩ accumulation after ATP was added to the reaction medium (Fig. 3A). We subsequently analyzed Ca 2ϩ uptake into membranes harvested from K616 cells expressing AtHMA1. 45 Ca 2ϩ accumulation in these vesicles was stimulated by ATP and was ϳ2-fold higher than the 45 Ca 2ϩ retained in membranes of the K616 mutant (Fig. 3A). The kinetic parameters of Ca 2ϩ uptake activity displayed by AtHMA1 were determined by measuring the initial rates of 45 Ca 2ϩ uptake at different free  3 and 6). ACT1 cDNA was used as an internal control (lanes 2 and 4). B, lethality of Ca 2ϩ transport-deficient yeast strains in Ca 2ϩ depleted media was reverted by the expression of AtHMA1. Yeast mutants K616 and ⌬pmr1 transformed with AtHMA1 and pGPD426 were streaked onto selective growth plates containing 10 mM EGTA. Wild-type strains W303 and YR98 transformed with pGPD426 were used as controls. C, AtHMA1 abolished the characteristic missorting of CPY displayed by K616 mutant. A dot-blot assay undertaken with an anti-CPY antibody was performed with samples of culture media and yeast lysate obtained from K616 cells transformed with pGPD426 or with AtHMA1. Wild-type cells (W303) were used as control. Ca 2ϩ concentrations (Fig. 3B). The calculated apparent K m for Ca 2ϩ was 370 nM (pCa 6.43), and the maximum rate of Ca 2ϩ uptake was 1.53 nmol mg Ϫ1 min Ϫ1 . This demonstrates that AtHMA1 functions as a high affinity Ca 2ϩ -ATPase in yeast. As described above, AtHMA1 has an S5 segment very similar to the S3 segment involved in determining sensitivity to TG in SERCA type Ca 2ϩ pumps (6,38). The ATP-dependent Ca 2ϩ uptake exhibited by the complemented mutant was reduced by 84% in the presence of 100 nM TG (Fig. 4A). In contrast, TG did   APRIL 11, 2008 • VOLUME 283 • NUMBER 15 not affect calcium uptake in empty vector-transformed yeast.

AtHMA1, a Plant Ca 2؉ /Heavy Metal ATPase
To further characterize the effects of this inhibitor on AtHMA1, we studied the concentration dependence of transport inhibition by TG. The results showed that the apparent inhibitor concentration causing 50% inhibition (IC 50 ) for TG was 16.74 nM (Fig. 4B).

AtHMA1 Expression Reverts the Cadmium-hypersensitive Phenotype Displayed by the Mutant Yeast ⌬ycf1 and Confers High Cadmium Tolerance onto Wild-type Yeast-Although
AtHMA1 belongs to the subclass of Zn 2ϩ /Co 2ϩ /Cd 2ϩ /Pb 2ϩ P IB -ATPases (2), there is no substantial evidence that AtHMA1 plays a role in cadmium transport or detoxification. To elucidate the role of AtHMA1 in cadmium transport, we performed functional complementation assays of the cadmium-hypersensitive yeast ⌬ycf1 and tested the effect of AtHMA1 expression on cadmium tolerance of wild-type yeast. Yeast cells were grown on solid media supplemented with 70 M Cd 2ϩ in an effort to evaluate the cadmium sensitivity displayed by ⌬ycf1 transformed with AtHMA1 or empty vector. AtHMA1-expressing cells grew in a manner similar to wild-type cells (Fig.  5). As expected, ⌬ycf1-pGPD426 cells did not survive in this medium (43). The high Cd 2ϩ tolerance of ⌬ycf1 yeast expressing AtHMA1 was confirmed by growing transformed cells in liquid growth medium supplemented with increasing concentrations of this heavy metal. The first remarkable effect of AtHMA1 expression in ⌬ycf1 cells was an increase in the generation time (G t ) (the doubling time of each yeast population) under control conditions compared with the growth exhibited by ⌬ycf1 and the wild-type strain (DTY165) ( Table 1). Nevertheless, at higher concentrations of cadmium (from 100 to 200 M Cd 2ϩ ), ⌬ycf1 cells expressing AtHMA1 grew faster than the control strains. This feature is reflected in a decrease of G t compared with control strains (Table 1). To determine whether AtHMA1 could also enhance the intrinsic cadmium tolerance displayed by wild-type yeast, we compared growth on high Cd 2ϩ over 5 days in empty vector-transformed WT (W303) cells (WT-pGPD426) and AtHMA1-expressing WT cells (WT-AtHMA1). WT-pGPD426 was able to grow in all Cd 2ϩ concentrations assayed (from 70 to 200 M) (Fig. 6A), but the G t increased as the heavy metal concentration increased (Table 2). Under control conditions (i.e. no cadmium added), WT-AtHMA1 grew slower than WT-pGPD426 (Fig. 6B, Table 2). This effect is very similar to that observed in ⌬ycf1 cells expressing AtHMA1 (Table 1). Interestingly, the enhancing effect of AtHMA1 expression over the cadmium tolerance of WT yeast became more evident when Cd 2ϩ reached toxic levels for empty vector-transformed WT yeast (up to 100 M). WT-AtHMA1 cells grew faster than WT-pGPD426 in all of the Cd 2ϩ concentrations assayed. Table 2 shows that the generation time for WT-AtHMA1 was similar for all treatments.
Cadmium, Cobalt, Copper, and Zinc Activate AtHMA1 ATPase, and Its Expression Enhances Tolerance to These Metals in Yeast-Our results indicate that AtHMA1 participates in cadmium detoxification in yeast. To confirm that AtHMA1 is a Cd 2ϩ -ATPase, we performed a biochemical characterization of its heavy metal specificity after expression in ⌬ycf1 yeast. Microsomal membranes were analyzed for heavy metal-activated ATPase activity. The ATPase measurements indicate that AtHMA1-transformed yeast is activated more by Cd 2ϩ (6-fold) than yeast transformed with empty vector (Fig. 7A). This coincides with the results obtained in the complementation assay of ⌬ycf1. Zn 2ϩ (15-fold), Cu ϩ (13-fold), and to a lesser extent Co 2ϩ (3-fold) also stimulated ATPase activity.
Given that AtHMA1 has been classified in the group of Zn 2ϩ /Co 2ϩ /Cd 2ϩ /Pb 2ϩ ATPases (2), our evidence indicates  that AtHMA1 transports Ca 2ϩ with high affinity but also participates in Cd 2ϩ detoxification. Our research also suggests that its ATPase activity could also be activated by Cd 2ϩ , Zn 2ϩ , Cu ϩ , and Co 2ϩ . To confirm that AtHMA1 plays a role in zinc, copper, and cobalt homeostasis, ⌬ycf1 cells expressing AtHMA1 were used to test its ability to confer tolerance onto toxic levels of these heavy metals. Fig. 7B shows that ⌬ycf1 cells transformed with empty vector (⌬ycf1-pGPD426) exhibited poor growth on solid media containing 6 mM CoCl 2 , 4 mM CuSO 4 , or 28 mM ZnCl 2 . However, AtHMA1expressing cells (⌬ycf1-AtHMA1) grew normally under these conditions.

DISCUSSION
AtHMA1 Functions as a Thapsigargin-sensitive Ca 2ϩ Pump in Yeast-Because TG is a plant-derived compound and there are no molecularly characterized TG-sensitive Ca 2ϩ pumps, it has been proposed that plants may have developed insensitivity to this compound (23). Neither the members of the plant Ca 2ϩ -ATPase family in A. thaliana nor calcium pumps of other plant species possess the characteristic conserved motif DEF-GEXXSK associated with the binding and highly specific inhibition of TG. AtHMA1 is the only plant P-type ATPase among the subclass of the heavy metal pumps that possesses this particular motif in its sequence.
Interestingly, AtHMA1 is able to transport calcium with an affinity similar to other Ca 2ϩ -ATPases described in Arabidopsis (23). Moreover, the IC 50 measured was 16.74 nM, which is close to the range of sensitivity of SERCA-ATPases (44). In Arabidopsis, calcium pumps that exhibit homology to SERCA-ATPases are classified in the P IIA subfamily of P-type ATPases (2). The four members of this subfamily transport calcium through the membrane of organelles or the plasma membrane (1,2). In addition to AtHMA1, some members of the P IIA subfamily could also transport transition metals such as manganese and zinc (45). This demonstrates the importance of functional characterization for classification purposes in addition to grouping P-type ATPases based on structural similarities.
AtHMA1 decreases sensitivity to a medium with low calcium for two calcium transport-deficient yeast mutants (⌬pmr1 and K616). In addition, expression of AtHMA1 fully restores the CPY missorting phenotype of K616 (26). On the other hand, it is likely that AtHMA1 expressed in ⌬pmr1 restores the transport of calcium into the lumen of the Golgi apparatus, the organelle where calcium ions are needed for normal protein processing and sorting in yeast (26,40).
AtHMA1 Participates in Heavy Metal Detoxification in Yeast-Cadmium is a nonessential element and in fact is toxic for cells. Functional expression of AtHMA1 phenotypically complements a cadmium-hypersensitive yeast strain and, interestingly, also confers an extraordinary cadmium tolerance onto wild-type yeast. In addition, cadmium significantly stimulates the ATPase activity of AtHMA1. It is possible that expression of AtHMA1 in S. cerevisiae may help to concentrate cadmium in organelles or mediate its extrusion to the extracellular medium, overcoming the stress produced by elevated concentrations of cadmium.
A role for AtHMA1 in copper transport in plants has been proposed by Seigneurin-Berny et al. (14). They demonstrated through yeast expression that AtHMA1 is involved in both zinc and copper homeostasis. Furthermore, they have provided evidence that AtHMA1 is located in the chloroplast envelope, and measurements of ATPase activity in purified chloroplast enve-ATPase activity (%) 20 FIGURE 7. ATPase activity of AtHMA1 is activated by cadmium, copper, zinc, and cobalt, and its expression enhances metal tolerance in yeast. A, metal-activated ATPase activity of AtHMA1 was determined in membranes from ⌬ycf1-pGPD426 cells (white bars) and ⌬ycf1-AtHMA1 cells (gray bars). B, growth of ⌬ycf1-pGPD426 and ⌬ycf1-AtHMA1 cells on solid selective medium in the presence of 6 mM CoCl 2 , 4 mM CuSO 4 , or 28 mM ZnCl 2 . Plates were incubated for 5 days at 30°C. lope membranes demonstrated that AtHMA1 activity is stimulated by copper but not by zinc, cobalt, iron, manganese, or silver (14). Our results show some differences. One explanation for this discrepancy is that Seigneurin-Berny et al. (14) used a truncated version of AtHMA1 (where the transit peptide of the protein was deleted) for the heterologous expression and functional studies in yeast. In contrast, we were able to detect protein activity using full-length cDNA expression in yeast. Moreover, we found that yeast cells expressing AtHMA1 are able to grow faster than those transformed with the empty vector in a growth medium supplemented with high copper. These data demonstrated that full-length AtHMA1 increases metal tolerance in yeast. We also provided evidence of participation of AtHMA1 in cadmium, zinc, and cobalt detoxification. It also has been suggested that AtHMA1 may be involved in zinc homeostasis (13). The evidence supporting this hypothesis is based on studies showing that an Arabidopsis disruption mutant of AtHMA1 is sensitive to high concentrations of zinc. This is consistent with the presence of a long His stretch in the sequence of AtHMA1, which is reminiscent of the poly(His) sequence that has proved to be involved in zinc transport in other species (36,46,47). Despite this characteristic sequence, no direct evidence of AtHMA1 participation in zinc transport had been obtained. We demonstrated that growth in a medium with high concentrations of zinc is faster in cells transformed with AtHMA1 than in cells transformed with the empty vector. This suggests that AtHMA1 could function as a zinc transporter. Furthermore, ATPase activity of AtHMA1 was highly stimulated by zinc, which provides additional support for the theory that this protein does transport this heavy metal.
There are structural differences between AtHMA1 and the other members of the P IB family. One unique characteristic that has been used to infer its function is its homology with CoaT. CoaT is a P-type ATPase from Synechocystis PCC6803, which appears to confer Co 2ϩ tolerance onto this organism (48). Based on this homology, it has been suggested that AtHMA1 may play a significant role in cobalt homeostasis in Arabidopsis (24). Here we confirm that AtHMA1 participates in cobalt detoxification in yeast. Yeast cells transformed with AtHMA1 grow faster than empty vectortransformed cells when they are grown in a medium supplemented with high cobalt. This result suggests that AtHMA1 could participate in the transport of this trace element. Cobalt (as well as cadmium, copper, and zinc) also stimulated the ATPase activity of AtHMA1, providing further evidence that this pump is able to transport Co 2ϩ .
Functional Expression of AtHMA1 in Yeast-Recent data suggest that AtHMA1 is an intrinsic protein of the chloroplast membrane envelope in Arabidopsis. This raises questions regarding the way in which AtHMA1 manages to complement defective yeast strains that lack ion transporters in other compartments such as the vacuole and the Golgi apparatus. The yeast cells provide a convenient system for uptake studies using plant transporters and represent a useful tool for studying their transport mechanisms and specificity. However, the activity of these transporters in plants may be modulated by interactions with proteins that are absent in yeast, and the expression of the genes might be developmentally or environmentally regulated, which may contribute to their specific functions (48). This is not the first example of transporters that appears to be localized differently in yeast and plants (49 -51). Triose phosphate translocators in the chloroplast membrane are found in endoplasmic reticulum membranes upon expression in yeast (52). This suggests that AtHMA1 could be translated into the endoplasmic reticulum and be present in organelles of the yeast secretory pathway such as the endoplasmic reticulum, Golgi, and vacuole. Thus functional AtHMA1 would be compensating for a failure in a specific organelle, explaining why AtHMA1 was able to complement the ⌬pmr1, K616, and ⌬ycf1 yeast mutants. Based on these results, we suggest that AtHMA1 has diverse specificity that might allow it to have various in planta roles related to metal and Ca 2ϩ homeostasis. The role of AtHMA1 as a multimetal transporter is beginning to be unraveled, and further studies are needed to elucidate its relevance for ionic homeostasis in A. thaliana.