Sulfur-containing amino acid residues function in antioxidative responses, which can be induced by the reactive oxygen species generated by excessive copper and hydrogen peroxide. In all Na+/K+, Ca2+, and H+ pumping P-type ATPases, a cysteine residue is present two residues upstream of the essential aspartate residue, which is obligatorily phosphorylated in each catalytic cycle. Despite its conservation, the function of this cysteine residue was hitherto unknown. In this study, we analyzed the function of the corresponding cysteine residue (Cys-327) in the autoinhibited plasma membrane H+-ATPase isoform 2 (AHA2) from Arabidopsis thaliana by mutagenesis and heterologous expression in a yeast host. Enzyme kinetics of alanine, serine, and leucine substitutions were identical with those of the wild-type pump but the sensitivity of the mutant pumps was increased towards copper and hydrogen peroxide. Peptide identification and sequencing by mass spectrometry demonstrated that Cys-327 was prone to oxidation. These data suggest that Cys-327 functions as a protective residue in the plasma membrane H+-ATPase, and possibly in other P-type ATPases as well.
All organisms have evolved complex antioxidant defenses that minimize oxidative damage. In proteins, the sulfur-containing amino acids cysteine and methionine are particularly sensitive to oxidation. These amino acids interact with heavy metals. Cysteine residues are involved in metal detoxification reactions  and coordination of metals during transport .
P-type ATPases constitute a superfamily of ion pumps that are phosphorylated on a conserved aspartate residue during each catalytic cycle. They are divided into five subfamilies, P1–P5, based on phylogeny and substrate specificity. Well-characterized members of this family are sarco/endoplasmic reticulum Ca2+-ATPase (SERCA) and the Na+/K+-ATPase, which both belong to the P2 ATPase subfamily, and plasma membrane H+-ATPase, which belongs to the P3 ATPase subfamily. Various P2 ATPases are sensitive to oxidative stress , and heavy metals have been reported to inhibit the mussel SERCA, the rat Na+/K+-ATPase, and the plant plasma membrane H+-ATPase [4–6]. The mechanism is unknown but it has been suggested that sulfhydryl groups are involved and that interference with Mg2+ binding in the phosphorylation (P) domain may play a role .
P-type ATPases all have three cytosolic domains: a nucleotide-binding domain (N-domain), a phosphorylation domain (P-domain), and an in-built de-phosphorylation domain called the actuator domain (A-domain). In the P-domain, two residues upstream of the phosphorylated aspartate residue and close to the Mg2+-binding site, is a cysteine residue that is conserved in all P2 and P3 ATPases. The high conservation of this cysteine residue and its close proximity to functionally important residues have led to several investigations of its role. So far, however, its function in the protein has remained elusive. Hydroxyl radicals inhibit SERCA activity in rabbit, and the corresponding conserved cysteine residue in SERCA (Cys-349) was suggested to be involved in the oxidative response . Cys-349 is also the most sensitive cysteine in SERCA to nitrosylation [8,9], and it showed a high reactivity toward amino acid peroxides . The role of these modifications is unknown. Mutagenesis studies of the yeast plasma membrane H+-ATPase Pma1p revealed that substituting alanine for the conserved cysteine did not alter expression level and led to a 7% decrease in ATPase activity. When seven out of the nine cysteines in the Pma1p protein were replaced with alanine, no significant decrease in ATP hydrolysis and proton pumping was observed . However, the substitution of the conserved cysteine with leucine was lethal for yeast. It was proposed that the conserved cysteine has an important structural role rather than a function related to hydrolytic activity . Similar observations were made for rabbit SERCA. When Cys-349 was replaced by alanine, there was no significant change in the rate of ATP-dependent Ca2+ transport . In Na+/K+ ATPase, the corresponding Cys-376-Ser mutant was wild type with respect to ouabain sensitivity . Substitutions at every cysteine residue in a sheep Na+/K+ ATPase caused no changes in ATP binding affinity . Thus, for a variety of P-type ATPases, the substitution of the conserved cysteine residue seems to have only minor effects on protein activity, and no specific functions for this residue have been postulated. This stands in contrast with the high conservation of this residue in P2 and P3 ATPases, and to descriptions in the literature of cysteine's high oxidative reactivity and ability to interact with heavy metals.
The role of the corresponding cysteine residue in plant plasma membrane H+-ATPases has so far not been investigated. In this work, a plasma membrane H+-ATPase (AHA2) from the model plant Arabidopsis thaliana was expressed in yeast. The conserved cysteine residue, Cys-327, was substituted with several amino acids and the resulting proteins were characterized. Our results point to a role for this residue in oxidative stress tolerance and protection against copper.
Materials and methods
Bacto agar and amino acid-free yeast nitrogen base were obtained from BD Biosciences. All other chemicals used were obtained from MERCK. PageRuler™ pre-stained ladder and Pierce™ 1-step transfer buffer were obtained from Thermo Fisher Scientific. Trypsin was obtained from Promega.
Strains and plasmid
The RS72 (MAT a, adel-100 his4–519 leu2–3,112) yeast strain, expressing native PM H+-ATPase PMA1 under the control by a galactose promotor, was used . Arabidopsis AHA2 was expressed in a plasmid (pmp1625) controlled by the native PMA1 promotor . PCR fragments for different substitutions were obtained using the primers listed in Supplementary Table S1. Single point mutations were introduced by homologous recombination. Therefore, plasmid and DNA fragment, containing the mutation of interest, were co-transformed in the yeast strain. Yeast cells were transformed with lithium acetate and polyethyleneglycol . All mutants were verified by sequencing.
Cultivation and growth
Yeast cells were transformed, grown, and harvested essentially as described by Villalba et al. . For microsome preparation, transformed yeast was cultured in 2% (w/v) d-galactose selection medium (0.7% (w/v) amino acid-free yeast nitrogen base including adenine (0.4 mg/ml) and L-histidine (0.3 mg/ml)) for 2 days at 30°C with shaking. Then, the cells were diluted 1 : 10 and cultured in 2% (w/v) d-glucose full medium (2% (w/v) bacto-peptone and 1% (w/v) yeast extract) for 20 h at 30°C with shaking. Yeast complementation assays were performed on 2% (w/v) d-glucose selection medium plates with the indicated pH. Then, 3 µl of OD600 0.1, 0.01, and 0.001 yeast suspension was added to the plates and cultured for 2 days at 30°C . Transformants were cultivated on 2% (w/v) d-galactose selection medium plates pH 5.5.
Microsomes were prepared in accordance with Pedersen et al. [20,21]. Briefly, yeast cells were harvested and homogenized at 4°C using glass beads. After centrifugation at 10 000g for 10 min at 4°C, microsomes were obtained from supernatant using ultra-centrifugation. The method from Pedersen et al. was altered as follows: the cells were homogenized in 100 mM MOPS (pH 7.5), 30% (v/v) glycerol, 10 mM EDTA, 40 mM KCl, and 1 mM PMSF. The obtained microsomes were washed and resuspended in 100 mM MOPS (pH 6.5), 20% (v/v) glycerol, 40 mM KCl, and 1 mM PMSF. The samples were frozen in liquid nitrogen and stored at −80°C .
ATP hydrolysis was detected according to Baginski et al. , following the modifications by Wielandt et al. . The assay was carried out at 30°C in 20 mM MOPS, 5 mM NaN3, 0.25 mM Na2MoO4, 25 mM KNO3, 5 mM ATP, and 10 mM MgSO4 at pH 6.5. Buffer and 0.5 µg protein were mixed and incubated for 45 min at 30°C. Then, equal volume of stop solution (93 mM ascorbic acid, 0.273 M HCl, 0.059% (w/v) sodium dodecylsulfate, and 5 mM (NH4)2MoO4) was added. 30 min after stop solution was added, two times volume of arsenite solution (154 mM NaAsO2, 68 mM trisodium citrate, and 350 mM acetic acid) was added . Absorbance was measured at 860 nm. To determine ATP hydrolysis at different pH values, buffers with pH 4.0, 4.33, 4.87, 5.32, 5.86, 6.26, 6.48, 6.70, 7.12, 7.44, 7.75, and 8.0 were used. To establish the ATP concentration dependence of ATPase activity, the hydrolytic activity was measured using 8 ATP concentrations ranging from 0–1.25 mM. All experiments were performed with at least three biological replicates ± SEM.
The effect of CuSO4, CdSO4, ZnSO4, HgCl2, Na3VO4, H2O2, and ONOO–on ATPase activity was tested by ATP hydrolysis. Proteins were mixed with inhibitors and incubated for 15 min at room temperature. Then, the assay was started by the addition of 3 mM ATP and 10 mM MgSO4 in 20 mM MOPS (pH 6.5) including 5 mM NaN3, 0.25 mM Na2MoO4, and 25 mM KNO3 (final concentrations) and incubated for 60 min at 30°C. Whereas ONOO− was dissolved and diluted in 0.01 M NaOH, the other compounds were dissolved and diluted in H2O.
Protein reconstitution and proton pumping
Microsomes were reconstituted in soybean (Glycine max) lipids [25,26] at a protein/lipid ratio of 1 : 22. The proton pumping activity was measured using 9-amino-6-chloro-2-methoxy-acridin (ACMA) in 50 mM MOPS (pH 6.5), 50 mM K2SO4, and 10% (v/v) glycerol (29, modified after 39). The reaction was started by adding 5 mM MgSO4 and the proton gradient was disturbed by adding 0.5 µM nigericin after 3.5 min.
S-alkylation of protein samples
For iodoacetamide (IAA) alkylation, microsomes were reduced with 10.5 mM DTT in 100 mM ammonium bicarbonate at pH 8.55 for 1 h at 55°C. Afterwards, 18 mM of IAA was added and the sample was incubated for 30 min at RT. After reduction with 10 mM DTT, the alkylation with N-ethylmaleimide (NEM) was carried out in 100 mM HEPES at pH 8, 1 mM EDTA and 0.1 mM neocuproin (HEN buffer) containing 10 mM N-ethylmaleimide and 0.1% (v/v) Triton X100 for 1 h. The samples were washed two times in HEN buffer, centrifuged at 100 000g for 1 h at 4°C and resuspended in HEN buffer. For oxidation detection, native proteins were run on a gel without former treatment.
In gel digestion of protein and sample preparation for mass spectrometry analysis
For IAA alkylated samples and oxidation detection the gels were prepared as described by Shevchenko et al. . In short, excised protein containing gel bands were washed in 50 mM ammonium bicarbonate and then 50% (v/v) acetonitrile. The gel pieces were shrunk in 100% (v/v) acetonitrile and air dried. The washing steps were repeated three times. Gel pieces were reconstituted in 0.025 µg/µl trypsin for 10 min, before the samples were digested in 50 mM ammonium bicarbonate overnight at 37°C. Supernatant was treated in accordance with Gobom et al. . After acidification with 0.25% (v/v) trifluoracetic acid (TFA) the samples were desalted on a custom-made micro column (plug from an Empore C18 disk (3 M) and 1–2 mm Poros R2 50 µm (Thermo Scientific) in a 10 µl pipet tip). The column was activated and equilibrated in 100% (v/v) acetonitrile and 0.1% (v/v) TFA, respectively. Following loading of the sample the column was washed twice in 0.1% (v/v) TFA. Peptides were sequentially eluted with 50% (v/v) acetonitrile, 0.1% (v/v) TFA and 70% (v/v) acetonitrile, 0.1% 8 (v/v) TFA, after which the pooled eluate was dried down.
The NEM alkylated samples were prepared as previously described by Bonn et al. . Briefly, gel lanes were fractionated into 10 gel pieces, cut into smaller blocks and transferred into low binding tubes. Samples were and washed until gel blocks were destained. After drying of gel pieces in a vacuum centrifuge, they were covered with trypsin solution. Digestion took place at 37°C overnight before peptides were eluted in water by ultrasonication. The peptide-containing supernatant was transferred into a fresh tube, desiccated in a vacuum centrifuge and peptides were resolubilized in 0.1% (v/v) acetic acid for mass spectrometric analysis.
Liquid chromatography mass spectrometry (LC–MS)
For IAA and oxidation detection, dried samples were dissolve in 0.1% (v/v) formic acid. The column setup was custom-made with a 3.5 cm, 100 µm ID precolumn of Reprosil-Pur 120 C18-AQ, 5 µm (Dr. Maisch) and an 18 cm, 75 µm ID analytical column of Reprosil-Pur 120 C18-AQ, 3 µm (Dr. Maisch). The samples were eluted by a gradient of 3%–45% (v/v) acetonitrile in 0.1% (v/v) formic acid for 45 min followed by a gradient from 45%–95% (v/v) acetonitrile in 0.1% (v/v) formic acid for 5 min. LC–MS/MS analyses was performed on an Q-Exactive HF orbitrap instrument (Thermo Fisher Scientific) using an EASY nLC chromatography system (Thermo Fisher Scientific). MS data was obtained using a top 12 approach, with 120 K mass resolution for MS and 30 K mass resolution for MS/MS, an MS range of m/z 350–1600, an isolation window of 1.2 m/z unit and dynamic exclusion for 20 s. Data was processed using Proteome Discoverer 2.4, with the search engine being Sequest, (Thermo Fisher Scientific), Xcalibur Qual browser (Thermo Fisher Scientific) and GPMAW (Lighthouse data).
For NEM detection, tryptic peptides were loaded on a self-packed analytical column (OD 360 µm, ID 100 µm, length 20 cm) filled with of Reprosil-Gold 300 C18, 5 µm material (Dr. Maisch) and eluted by a binary nonlinear gradient of 5%–99% (v/v) acetonitrile in 0.1% (v/v) acetic acid over 82 min with a flow rate of 300 nl/min. LC–MS/MS analyses were performed on an LTQ Velos Pro (Thermo Fisher Scientific) using an EASY-nLC II liquid chromatography system. For MS analysis, a full scan in the Orbitrap with a mass resolution of 30 000 was followed by higher-energy collisional dissociation (HCD) of the ten most abundant precursor ions. MS2 experiments were acquired in the linear ion trap. Data was processed using Sorcerer-SEQUEST 4 (Sage-N Research) and Scaffold V4.8.7 (Proteome Software). The MS data were deposited to the ProteomeXchange Consortium via the PRIDE partner repository with the dataset identifier PXD022964.
TCA precipitation and SDS–PAGE
Microsomes were precipitated in 10% (w/v) trichloroacetic acid (TCA) on ice. After centrifugation at 14 000 g for 15 min at 4°C, the sample was dissolved in 5× Laemmli buffer containing 0.3 M TRIS–HCl (pH 6.8), 3.8% (w/v) sodium dodecylsulfate, 50 mM DTT, 13% (w/v) sucrose, 0.25% (w/v) bromphenol blue and 10 mM EDTA. Sodium dodecylsulfate polyacrylamide gel electrophoresis (SDS–PAGE) was performed on a 10% mini-gel for 2 h at room temperature according to Villalba et al. .
Proteins separated by SDS–PAGE were electrophoretically transferred to a nitrocellulose membrane using Pierce™ 1-step transfer buffer and Pierce™ fast blotter at 25 V for 12 min. The membrane was blocked in 5% (w/v) skim milk powder in 20 mM TRIS-HCl (pH 8.0) and 150 mM NaCl. Primary antibody against the AHA2 C-terminal domain was added. Bands were detected using alkaline phosphatase conjugated to secondary antibody, as described in Villalba et al. .
The protein concentration of isolated microsomes was determined using Bradford reagent (0.01% (w/v) Coomassie Brilliant Blue G-250, 4.7% (v/v) ethanol, and 8.5% (v/v) phosphoric acid) and bovine serum albumin (2–20 µg protein) as standard . The samples were incubated for 15 min and measured at an absorbance of 595 nm.
Crystal structure of AHA2 (5KSD) was obtained from Protein Data Bank and based on studies from Focht et al. . The structure was visualized by PyMOL.
Alignment and statistical analysis
Protein sequences were obtained from UniProt (Supplementary Table S2) and aligned by MEGA-X. Statistical analysis was performed using GraphPad Prism® 6.0 and Sequence Manipulator Suite V2.
AHA2 Cys-327 is conserved in most Arabidopsis thaliana P-type ATPases
Many P-type ATPases have a cysteine residue that is located two residues upstream of the aspartate residue in the DKTGT phosphorylation site of the P-domain (Figure 1A). This residue corresponds to Cys-327 in AHA2, a P3A-type ATPase in the plant A. thaliana. A comparison of selected P-type ATPases from A. thaliana, Homo sapiens, and Saccharomyces cerevisiae representing all subgroups (P1B-, P2A-, P2B-, P2C-, P2D-, P3A-, P4-, and P5-type ATPases) showed widespread conservation of the cysteine residue across all three kingdoms. P1B- and P4-type ATPases were the only subfamilies in which the residue was not strictly conserved (Figure 1B). Alignment of all P-type ATPases from A. thaliana confirmed that the cysteine residue was conserved in all subfamilies except P1B- and P4-type ATPases (Supplementary Figure S1).
Localization and conservation of AHA2 Cys-327.
To investigate the role of this residue in P3A ATPases, we focused on AHA2. When yeast membranes expressing recombinant AHA2 were treated with the cysteine-modifying reagent NEM and subsequently digested with trypsin, three tryptic peptides, in which cysteine residues had been S-alkylated, could be recovered from AHA2 (Table 1). One of these was a peptide that included Cys-327. Additionally, it was possible to S-alkylate Cys-327 with iodoacetamide (IAA) (Figure 2). Taken together, this demonstrates the accessibility of this cysteine residue for external reagents. The tryptic peptide with Cys-327 included tree methionine residues that were detected to be partially oxidized both before and after reduction in protein samples with DTT in both wild type and a C327A mutant (Table 1 and Figure 2). Cys-327 of AHA2 was substituted with either leucine, alanine, aspartate, or serine; these mutant proteins had similar expression to wild-type AHA2 in yeast. However, it was not possible to express an aspartate-substituted protein (Supplementary Figure S2). To determine whether the mutant proteins were active plasma membrane H+-ATPases, we tested their ability to complement a yeast strain defective in expressing its endogenous proton pump PMA1 when grown on glucose medium. The leucine- and aspartate-substituted proteins did not support growth of the pma1 yeast strain, but the serine- and alanine-substituted proteins did, indicating that the latter were active proteins (Figure 3).
LC–MS/MS analysis of AHA2 wild type and serine mutant accessibility towards S-alkylation by iodoacetamide (IAA) and oxidation.
Yeast complementation assay.
To further investigate the effects of the substitutions, the mutated genes were expressed in yeast, total microsomes containing the proteins were isolated, and the biochemical properties of the mutant proteins were assayed. In comparison with the wild-type AHA2 pump, the alanine-, serine-, and leucine-substituted proteins had the same Vmax and KM for ATP. Likewise, they had maximum ATPase activity at similar pH (6.2–6.4) and no significant differences in their sensitivity towards the P-type ATPase inhibitor vanadate (Figure 4). Proteins substituted with aspartate or lysine showed low activity, and KM and Vmax could not be determined for them (Supplementary Table S3). Proton transport into lipid vesicles was evaluated for all mutant proteins. Reconstituted protein was added to the cuvette in an amount sufficient to provide the same hydrolytic ATPase activity as 10 µg wild-type protein. Compared with the wild type, the serine- and alanine-substituted mutants had a higher proton pumping rate, measured as a decrease in ACMA fluorescence (Figure 5), indicating that they were able to effectively pump protons.
Kinetic analysis of AHA2 Cys-327 substitution mutants.
Proton pumping activity of AHA2 cysteine substitution mutants.
Thus, the substitution of Cys-327 with the bulky and charged amino acid residues aspartate and lysine severely impacted the function of AHA2, whereas the same residue could be replaced by alanine, serine and leucine without significant loss of function. This indicates that Cys-327 is not essential for the general functionality of the ATPase but rather may have a specialized function.
Substitution of Cys-327 of AHA2 increases Cu2+ sensitivity
Cysteine residues interact with heavy metal ions. To test whether Cys-327 is susceptible to heavy metals, the activities of wild-type AHA2 and the alanine and serine substitution mutants were assayed in the presence of various heavy metal ions. Copper (Cu2+), zinc (Zn2+), and cadmium (Cd2+) ions completely inhibited wild-type AHA2 ATPase activity at millimolar concentrations, whereas mercury (Hg2+) ions inhibited this activity in the micromolar range. Hg2+ was the most potent inhibitor of wild-type ATPase activity followed by Cu2+, Zn2+, and Cd2+ (Figure 6). Among the tested heavy metals, Cu2+ had the most divergent effects on the activity of wild-type versus mutant proteins. Both mutants were significantly more sensitive than the wild type to Cu2+ inhibition. The IC50 for the wild type was 2.1 µM, but was 1.2 µM for the C327S mutant and 1.6 µM for the C327A mutant (Figure 5A and Table 2). Barely any differences were observed between the wild type and the substitution mutants in their sensitivities to Zn2+ and Cd2+ (Figure 6B,D). In contrast, Hg2+ inhibited the ATPase activity of both mutant proteins; however, these mutant proteins were more tolerant to Hg2+ than the wild-type protein, and the serine-substituted protein was the more tolerant of the two mutants (Figure 6C). Together, these results show that substitution of Cys-327 with other amino acids increased the sensitivity of the pump to Cu2+ but did not have a marked impact on its sensitivity to other heavy metals.
Heavy metal sensitivity of AHA2 substitution mutants.
|Heavy metal .||IC50 confidential interval (95%) .|
|WT .||C327S .||C327A .|
|Heavy metal .||IC50 confidential interval (95%) .|
|WT .||C327S .||C327A .|
AHA2 Cys-327 mutants show increased sensitivity to reactive oxygen species
Cu2+ promotes the formation of reactive oxygen species (ROS), which leads to oxidative stress. Cysteine residues are important redox regulators within cells. To determine the role of C327 in the ROS sensitivity of AHA2, the wild-type and mutant proteins were treated with different ROS. The ATPase activity of wild-type AHA2 was barely affected by hydrogen peroxide (H2O2), even at very high concentrations (10 mM) (Figure 7A). In contrast, the alanine and serine substitution mutants were both sensitive to H2O2. At 1 mM H2O2, the alanine substitution mutant was inhibited 28% and the serine substitution mutant 55% relative to the wild type. The wild-type protein and alanine-substituted protein showed an increase in ATP hydrolysis between 0.3 µM and 300 µM H2O2. A complete inhibition of ATP hydrolysis activity was not observed for any genotype.
ROS sensitivity of AHA2 C327S and C327A substitution mutants.
The highly reactive oxidant peroxynitrite (ONOO–) completely inhibited the ATPase activity of the wild-type pump (IC50 0.93 µM) (Figure 7B). Interestingly, both mutants showed a lower IC50 towards ONOO−. The serine substitution mutant was the most sensitive to ONOO− (IC50 0.29 µM) but the alanine substitution mutant also showed increased sensitivity (IC50 0.59 µM), indicating that the mutant proteins were more susceptible to oxidation. The differences in inhibition between the genotypes were smaller for ONOO− than they were for H2O2. Altogether, the substitution of Cys-327 results in a higher ROS sensitivity, especially towards H2O2. This idea is supported by the detected oxidations, including cysteine trioxidation, in non-reduced and non-S-alkylated wild-type AHA2 samples (Figure 2C).
A conserved cysteine is found close to the DKTGT phosphorylation site in many P-type ATPases. All but a few subfamilies of P-type ATPases have a conserved cysteine at this position, the exceptions being P1B-type heavy metal-transporting ATPases and P4-type phospholipid-flipping ATPases (Figure 1B). In this study, we demonstrated that Cys-327 in the plant P3A-type plasma membrane H+-ATPase AHA2 seems to acts as a protectant for the pump from Cu2+ and H2O2, both of which contribute to ROS generation in cells.
Plasma membrane H+-ATPases are the powerhouses of plant growth by energizing the cell membrane . These ATPases play an important role under stress conditions, regulate cytosolic pH, and coordinate cell growth. Recently, there has been increased focus on the importance of cysteine and its interplay with ROS sensing and stress responses in plants. External H2O2 is sensed by a leucine-rich repeat receptor kinase via modification of a cysteine residue . Inside cells, other reactive species, such as nitric oxide (NO) and peroxynitrite, can oxidize the sulfhydryl group of cysteine [3,34]. Recent studies indicate that plasma membrane H+-ATPase is involved in NO-mediated auxin action in cell growth [35,36]. Additionally, NO seems to promote H2-induced adventitious root formation in cucumber (Cucumis sativus) by regulating the expression and interaction of plasma membrane H+-ATPase and 14-3-3 protein . Despite these findings, little is known about the effect of redox-active agents on the plant plasma membrane H+-ATPase.
Cys-327 is not essential for AHA2 activity and can be substituted with different amino acids without loss of function. Substitution with residues similar to cysteine in size and polarity resulted in functional AHA2 proteins (Figure 4). This finding is in agreement with the results of previous studies in which the corresponding residue was mutagenized in P2-type Na+/K+ and Ca2+ transporting ATPases [6,11,12,14]. Non-conservative substitutions at this position might result in folding problems and retention of the protein in the endoplasmic reticulum. Such a phenotype has been observed in the S. cerevisiae Pma1p C376L mutant, in which the mutated protein accumulates in internal structures of the yeast cell . The ATP affinity and vanadate sensitivity of AHA2 were not altered in the tested substitution mutant (Figure 4C), which indicates that binding and hydrolysis of ATP remain unaffected.
Various heavy metals had distinct effects on the Cys-327 substitution mutants. Both substitution mutants tested had an increased sensitivity to Cu2+ inhibition, whereas they had decreased sensitivity to Hg2+ compared with the wild type. The interactions of proteins with Cu2+ and Hg2+ have been linked to sulfhydryl residues [5,39]. The wild-type and substitution mutant proteins did not vary significantly in their sensitivity to Cd2+ and Zn2+. The observed inhibitory concentrations of heavy metals are in accordance with the heavy metal sensitivities previously reported for plasma membrane H+-ATPases [39,40]. In heavy metal-pumping P1B-type ATPases, the heavy metal-binding motif CPX and metal-binding domain CXXC both rely on cysteine to coordinate heavy metals [2,41]. P1B-ATPases might not need a protective cysteine at the active site as the transport site and regulatory domains serve as competing sinks for heavy metals.
Cu2+ is an essential micronutrient for plants, particularly for photosynthesis [42,43]. Because Cu2+ can readily gain and lose an electron, it is a cofactor for many oxidases (e.g. amine oxidases, ammonia monoxidase, ceruloplasmin, and lysyl oxidase) and for enzymes reacting with superoxide radicals (e.g. superoxide dismutase and ascorbate oxidase). In planta however, excess Cu2+ is highly toxic to plants as it catalyzes Fenton-like reactions, which generate hydroxyl radicals, leading to ROS accumulation and oxidative stress [44–46]. In vitro, the autoxidation of heavy metal had been shown to have similar effect [47,48]. This increase in ROS leads to changes in the activity of many enzymes involved in antioxidative pathways . Additionally, Cu2+ ions from cuprous complexes with thiols, inhibiting protein activity . The plasma membrane H+-ATPase can now be added to the list of Cu2+-sensitive enzymes.
ROS interact with sulfhydryl residues. The serine and alanine substitution mutants of AHA2 Cys-327 were both inhibited at relatively low H2O2 concentrations compared with the wild type; even at 100 µM H2O2, wild-type ATP hydrolysis was not affected. Dremina et al.  showed that wild-type SERCA activity was also not inhibited by 100 µM H2O2. In AHA2, the substitution mutants were inhibited at lower concentrations of ONOO– than the wild type, which was completely inhibited at 4 mM ONOO–. The same concentration of ONOO− also completely inhibited SERCA activity . H2O2 preferentially reacts with cysteine residues, while ONOO− also favors reactions with tyrosine, tryptophan, phenylalanine, and methionine [50,51]. ONOO− may react with several amino acids and, therefore, may inhibit ATPase activity via a different mechanism than H2O2. In SERCA, ONOO− inhibits ATPase hydrolysis via tyrosine nitration and thiol oxidation .
In structural models of P-type ATPases, Cys-327 in AHA2 (PDB entry 5KSD) is buried in the P-domain and is not exposed on the surface of the protein (Figure 1A). However, Cys-327 is accessible to external reagents as it could be alkylated by NEM and IAA (Table 1 and Figure 2), in analogy to what has been shown for the corresponding cysteine in a P2A-type ATPase . In addition, Cys-327 has been shown to be prone to oxidation (-SO, -SO2, -SO3) even without treatment (Figure 2). Several studies have reported that in native P2A- and P2C-type ATPases the corresponding cysteine residue is modified following exposure to not only reactive oxygen but also nitrogen species [3,8–10] and fatty acids . Taken together, corresponding cysteine residues in P-type ATPases are reachable by several molecules and Cys-327 in AHA2 is most likely accessible to ROS. A cysteine residue directly exposed on the protein surface could be subjected to rapid thiolation and other unwanted reactions, whereas burying a cysteine residue may increase the likelihood of specific interactions . Further inspection of the location of Cys-327 within the structure revealed that its side chain protrudes into an internal cavity that would accommodate a bound Cu2+ ion coordinated by the side chains of Cys-327, Met-585, and His-581. While we were not able to observe actual binding of Cu2+ to this site, we note that disruption of the potential Cu2+-binding pocket by mutagenesis of Cys-329 resulted in increased sensitivity to Cu2+ as well as to other agents causing oxidative stress. This increased sensitivity suggests that the site acts to sequester Cu2+ and thereby confers protection against ROS. A similar role for a buried Cu2+-binding site has been reported for the heat shock protein αB-crystallin .
Endogenous thiol residues provide antioxidative protection in proteins. For example, cysteine and methionine residues account for 40%–80% of the antioxidant capacity from human serum albumin (HSA) . Mutagenesis studies revealed that Cys-34 was responsible for 68% of HSA antioxidative activity . Similar to buried methionine in glutamine synthase, the corresponding conserved cysteine in SERCA had been proposed to function as endogenous antioxidant [3,58]. If AHA2 Cys-327 functions as a protective residue, a substitution mutant should be more sensitive to oxidative inactivation. The results presented in this work are in accordance with this hypothesis; we provide experimental evidence that the cysteine at position 327 in AHA2 could act in a similar way. This protective property could be of importance under stress conditions that are accompanied by increased production of ROS. As a result, other residues around the phosphorylated reaction cycle intermediate, including the Mg2+-binding site of the protein, may be protected. The present study supports the previous proposition [58,59] that sulfur-containing amino acids may be key players in metal-chelating and redox-cycling activities that scavenge free radicals. The findings additionally provide the first evidence for redox regulation of plasma membrane H+-ATPases and suggest a similar function for the corresponding cysteine in other P-type ATPases.
The authors declare no conflict of interest.
This work was supported by the German Research Foundation (RTG 1947: grant no. 231396381) (M.W.). Work in the laboratory of M.P. is supported by the Danish National Research Foundation (PumpKin), the Innovation Fund Denmark (LESSISMORE), the Carlsberg Foundation (RaisingQuinoa; CF18-1113), and the Novo Nordisk Foundation (NovoCrops; 2019OC53580). Proteomics and mass spectrometry research at the laboratory of O.N.J. was supported by generous grants of VILLUM Centre for Bioanalytical Sciences (VILLUM Foundation grant no. 7292) and PRO-MS: Danish National Mass Spectrometry Platform for Functional Proteomics (grant no. 5072-00007B).
Marcel Welle: Conceptualization, Investigation, Writing — original draft. Jesper Torbøl Pedersen: Conceptualization, Supervision, Writing — original draft. Tina Ravnsborg: Methodology. Maki Hayashi: Supervision. Sandra Maaß: Methodology. Dörte Becher: Methodology. Ole Nørregaard Jensen: Methodology. Christine Stöhr: Writing — original draft. Michael Palmgren: Conceptualization, Supervision, Writing — original draft.
All supporting data are included within the main article and its supplementary files. MS data not included in main article are aviable at ProteomeXchange Consortium PXD022964.
This work was supported by the German Research Foundation (DFG) by the Research Training Group (RTG) 1947 (M.W.) and the Danish National Research Foundation project PUMPKIN (M.P.).We thank Sebastian Grund (University of Greifswald) for assistance in sample preparation and Claudia Hirschfeld (University of Greifswald) for her support with MS analysis. Proteomics and mass spectrometry research at SDU are supported by generous grants to the VILLUM Center for Bioanalytical Sciences [VILLUM Foundation grant no. 7292 to O.N.J.] and PRO-MS: Danish National Mass Spectrometry Platform for Functional Proteomics [grant no. 5072-00007B to O.N.J].