EPR studies of the mitochondrial alternative oxidase. Evidence for a diiron carboxylate center

The alternative oxidase (AOX) is a ubiquinol oxidase found in the mitochondrial respiratory chain of plants as well as some fungi and protists. It has been predicted to contain a coupled diiron center on the basis of a conserved sequence motif consisting of the proposed iron ligands, four glutamate and two histidine residues. However, this prediction has not been experimentally verified. Here we report the high level expression of the Arabidopsis thaliana alternative oxidase AOX1a as a maltose-binding protein fusion in Escherichia coli. Reduction and reoxidation of a sample of isolated E. coli membranes containing the alternative oxidase generated an EPR signal characteristic of a mixed-valent Fe(II)/Fe(III) binuclear iron center. The high anisotropy of the signal, the low value of the g-average tensor, and a small exchange coupling (-J) suggest that the iron center is hydroxo-bridged. A reduced membrane preparation yielded a parallel mode EPR signal with a g-value of about 15. In AOX containing a mutation of a putative glutamate ligand of the diiron center (E222A or E273A) the EPR signals are absent. These data provide evidence for an antiferromagnetic-coupled binuclear iron center, and together with the conserved sequence motif, identify the alternative oxidase as belonging to the growing family of diiron carboxylate proteins. The alternative oxidase is the first integral membrane protein in this family, and adds a new catalytic activity (ubiquinol oxidation) to this group of enzymatically diverse proteins.     

Site-directed mutagenesis reveals the essentiality of the conserved residues in the putative diiron active site of the trypanosome alternative oxidase.

Trypanosoma brucei possesses a non-cytochrome, salicylhydroxamic acid (SHAM)-sensitive ubiquinol:oxygen oxidoreductase known as trypanosome alternative oxidase (TAO). TAO and similar SHAM-sensitive alternative oxidases (AOXs) contain 2-3 conserved diiron-binding motifs (EXXH). Site-directed mutagenesis of residues H165A, E214A, E266A, and H269L within the conserved EXXH motif abolished the ability of TAO to complement the heme-deficient Escherichia coli strain GE1387. These mutations also reduced the growth of this E. coli auxotroph to about 85% of the control cells containing wild type TAO. In contrast, mutation of residues outside the EXXH motifs, e.g. V205A, L243A, C261A, and V271A, had little effect on complementation, and the reduction in the cell growth was about 5-10%. Mutations of the putative iron-binding residues within the EXXH motifs of TAO abolished the ability to confer SHAM-sensitive respiration to E. coli heme mutant, whereas mutations of the non-conserved/non-iron binding residues resulted in 20-30% reduction of SHAM-sensitive respiration of the E. coli auxotroph. Immunoblot analysis of the total cellular protein of transformed E. coli revealed that the expression level of mutated and wild type TAO (35 kDa) remained unaltered. Mutation at C261A produced a truncated but functional protein of 28 kDa. The addition of ortho-phenanthroline to the growth medium produces a non-functional TAO. The effect of ortho-phenanthroline on the activity of TAO was completely alleviated by the addition of iron in the medium, which suggests that iron is needed for the activity of TAO. This work demonstrates that His-165, Glu-214, Glu-266, and His-269 and the presence of iron are essential for the activity of TAO.     

Reactions of nitric oxide with the reduced non-heme diiron center of the soluble methane monooxygenase hydroxylase

The soluble methane monooxygenase system from Methylococcus capsulatus (Bath) catalyzes the oxidation of methane to methanol and water utilizing dioxygen at a non-heme, carboxylate-bridged diiron center housed in the hydroxylase (H) component. To probe the nature of the reductive activation of dioxygen in this system, reactions of an analogous molecule, nitric oxide, with the diiron(II) form of the enzyme (Hred) were investigated by both continuous and discontinuous kinetics methodologies using optical, EPR, and M?ssbauer spectroscopy. Reaction of NO with Hred affords a dinitrosyl species, designated Hdinitrosyl, with optical spectra (lambdamax = 450 and 620 nm) and M?ssbauer parameters (delta = 0.72 mm/s, DeltaEQ = 1.55 mm/s) similar to those of synthetic dinitrosyl analogues and of the dinitrosyl adduct of the reduced ribonucleotide reductase R2 (RNR-R2) protein. The Hdinitrosyl species models features of the Hperoxo intermediate formed in the analogous dioxygen reaction. In the presence of protein B, Hdinitrosyl builds up with approximately the same rate constant as Hperoxo ( approximately 26 s-1) at 4 degrees C. In the absence of protein B, the kinetics of Hdinitrosyl formation were best fit with a biphasic A --> B --> C model, indicating the presence of an intermediate species between Hred and Hdinitrosyl. This result contrasts with the reaction of Hred with dioxygen, in which the Hperoxo intermediate forms in measurable quantities only in the presence of protein B. These findings suggest that protein B may alter the positioning but not the availability of coordination sites on iron for exogenous ligand binding and reactivity.     

Strain-specific difference in amino acid sequences of trypanosome alternative oxidase

Cyanide-insensitive trypanosome alternative oxidase (TAO) is the terminal oxidase of the respiratory chain of long slender bloodstream forms of the African trypanosome, which causes sleeping sickness in human and nagana in cattle. TAO has been targeted for the development of anti-trypanosomal drugs because it does not exist in the host. The cDNA for TAO has been cloned from Trypanosoma brucei brucei EATRO110 strain and has been used for further characterization. In this study, we found amino acid sequence of the C-terminal part of TAO from the strain that we are using, T. b. brucei TC221, is considerably different from that of the EATRO110 strain.     

Compelling EPR evidence that the alternative oxidase is a diiron carboxylate protein

The alternative oxidase is a respiratory chain protein found in plants, fungi and some parasites that still remains physically uncharacterised. In this report we present EPR evidence from parallel mode experiments which reveal signals at approximately g=16 in both purified alternative oxidase protein (g=16.9), isolated mitochondrial membranes (g=16.1), and in trypanosomal AOX expressed in Escherichia coli membranes (g=16.4). Such signals are indicative of a dicarboxylate diiron centre at the active site of the enzyme. To our knowledge these data represent the first EPR signals from AOX present in its native environment.     

Compelling EPR evidence that the alternative oxidase is a diiron carboxylate protein

The alternative oxidase is a respiratory chain protein found in plants, fungi and some parasites that still remains physically uncharacterised. In this report we present EPR evidence from parallel mode experiments which reveal signals at approximately g = 16 in both purified alternative oxidase protein (g = 16.9), isolated mitochondrial membranes (g = 16.1), and in trypanosomal AOX expressed in Escherichia coli membranes (g = 16.4). Such signals are indicative of a dicarboxylate diiron centre at the active site of the enzyme. To our knowledge these data represent the first EPR signals from AOX present in its native environment.     

Glycolate oxidase is an alternative source for H2O2 production during plant defense responses and functions independently from NADPH oxidase

The photorespiratory enzyme glycolate oxidase (GOX) was found to be involved in nonhost resistance by regulating plant defense responses through the production of H₂O₂. Silencing of a gene encoding NADPH oxidase (AtRBOHD) in the gox mutants did not further increase susceptibility to a nonhost pathogen, P. syringae pv tabaci, although it caused an increase in bacterial growth in the Atgox1 and Atgox3 mutant backgrounds. In order to confirm this finding, we created double homozygous knockouts AtrbohD x Atgox1 and AtrbohD x Atgox3 to evaluate symptom development and bacterial growth. Here we show that there is no additive effect of disease symptoms or bacterial growth in the AtrbohD x Atgox1 and AtrbohD x Atgox3 double mutants when compared with individual mutants. Slight additive effect observed previously upon silencing of AtRBOHD in Atgox1 and Atgox3 mutants was most likely due to cross-silencing of AtRBOHF. These results further prove that GOX plays a role in nonhost resistance independent of NADPH oxidase.     

Overproduction of highly active trypanosome alternative oxidase in Escherichia coli heme-deficient mutant

Cyanide-insensitive trypanosome alternative oxidase (TAO) is the terminal oxidase of the respiratory chain of long slender bloodstream forms of the African trypanosome, which causes sleeping sickness in humans and nagana in cattle. TAO has been targeted for the development of anti-trypanosomal drugs, because it does not exist in the host. In this study, we established a system for overproduction of highly active TAO in Eschericia coli heme-deficient mutant. Kinetic analysis of recombinant enzyme and TAO in Trypanosoma brucei brucei mitochondria revealed that recombinant TAO retains the properties of native enzyme, indicating that recombinant TAO is quite valuable for further biochemical study of TAO.     

The ferroxidase reaction of ferritin reveals a diferric mu-1,2 bridging peroxide intermediate in common with other O2-activating non-heme diiron proteins

Ferritins are ubiquitous proteins that concentrate, store, and detoxify intracellular iron through oxidation of Fe2+ (ferroxidation), followed by translocation and hydrolysis to form a large inorganic mineral core. A series of mutagenesis, kinetics, and spectroscopic studies of ferritin led to the proposal that the oxidation/translocation path involves a diiron protein site. Recent stopped-flow absorption and rapid freeze-quench M?ssbauer studies have identified a single peroxodiferric species as the initial transient intermediate formed in recombinant frog M ferritin during rapid ferroxidation [Pereira, S. A., Small, W., Krebs, C., Tavares, P., Edmondson, D. E., Theil, E. C., and Huynh, B. H. (1998) Biochemistry 37, 9871-9876]. To further characterize this transient intermediate and to establish unambiguously the peroxodiferric assignment, rapid freeze-quenching was used to trap the initial intermediate for resonance Raman investigation. Discrete vibrational modes are observed for this intermediate, indicating a single chromophore in a homogeneous state, in agreement with the M?ssbauer conclusions. The frequency at 851 cm-1 is assigned as nu(O-O) of the bound peroxide, and the pair of frequencies at 485 and 499 cm-1 is attributed, respectively, to nus and nuas of Fe-O2-Fe. Identification of the chromophore as a micro-1,2 bridged diferric peroxide is provided by the isotope sensitivity of these Raman bands. Similar peroxodiferric intermediates have been detected in a mutant of the R2 subunit of ribonucleotide reductase from Escherichia coli and chemically reduced Delta9 stearoyl-acyl carrier protein desaturase (Delta9D), but in contrast, the ferritin intermediate is trapped from the true reaction pathway of the native protein. Differences in the Raman signatures of these peroxide species are assigned to variations in Fe-O-O-Fe angles and may relate to whether the iron is retained in the catalytic center or released as an oxidized product.     

Biphasic regulation of H2O2 on angiogenesis implicated NADPH oxidase

ROS (reactive oxygen species) take an important signalling role in angiogenesis. Although there are several ways to produce ROS in cells, multicomponent non‐phagocytic NADPH oxidase is an important source of ROS that contribute to angiogenesis. In the present work, we examined the effects of H₂O₂ on angiogenesis including proliferation and migration in HUVECs (human umbilical vein endothelial cells), new vessel formation in chicken embryo CAM (chorioallantoic membrane) and endothelial cell apoptosis, which is closely related to anti‐angiogenesis. Our results showed that H₂O₂ dose‐dependently increased the generation of O₂ ^? (superoxide anion) in HUVECs, which was suppressed by DPI (diphenylene iodonium) and APO (apocynin), two inhibitors of NADPH oxidase. H₂O₂ at low concentrations (10 µM) stimulated cell proliferation and migration, but at higher concentrations, inhibited both. Similarly, H₂O₂ at 4 nmol/cm² strongly induced new vessel formation in CAM, while it suppressed at high concentrations (higher than 4 nmol/cm²). Also, H₂O₂ (200～500 µM) could stimulate apoptosis in HUVECs. All the effects of H₂O₂ on angiogenesis could be suppressed by NADPH oxidase inhibitors, which suggests that NADPH oxidase acts downstream of H₂O₂ to produce O₂ ^? and then to regulate angiogenesis. In summary, our results suggest that H₂O₂ as well as O₂ ^? mediated by NADPH oxidase have biphasic effects on angiogenesis in vitro and in vivo.     

Adrenaline stimulates H2O2 generation in liver via NADPH oxidase

It is known that adrenaline promotes hydroxyl radical generation in isolated rat hepatocytes. The aim of this work was to investigate a potential role of NADPH oxidase (Nox) isoforms for an oxidative stress signal in response to adrenaline in hepatocytes. Enriched plasma membranes from isolated rat liver cells were prepared for this purpose. These membranes showed catalytic activity of Nox isoforms, probably Nox 2 based on its complete inhibition with specific antibodies. NADPH was oxidized to convert O(2) into superoxide radical, later transformed into H(2)O(2). This enzymatic activity requires previous activation with either 3 mM Mn(2+) or guanosine 5'-0-(3-thiotriphosphate) (GTPgammaS) plus adrenaline. Experimental conditions for activation and catalytic steps were set up: ATP was not required; S(0.5) for NADPH was 44 microM; S(0.5) for FAD was 8 microM; NADH up to 1 mM was not substrate, and diphenyleneiodonium was inhibitory. Activation with GTPgammaS plus adrenaline was dose- and Ca(2+)-dependent and proceeded through alpha(1)-adrenergic receptors (AR), whereas beta-AR stimulation resulted in inhibition of Nox activity. These results lead us to propose H(2)O(2) as additional transduction signal for adrenaline response in hepatic cells.     

Prevention of H2O2 generation by monoamine oxidase protects against CNS O2 toxicity.

Toxicity to the central nervous system (CNS) by hyperbaric oxygen (HBO) presumably relates to increased production of reactive oxygen species. The sites of generation of reactive oxygen species during HBO, however, have not been fully characterized in the brain. We investigated the relationship between regional generation of hydrogen peroxide (H2O2) in the brain in the presence of an irreversible inhibitor of catalase, aminotriazole (ATZ), and protection from CNS O2 toxicity by a monoamine oxidase (MAO) inhibitor, pargyline. At 6 ATA of oxygen, pargyline significantly protected rats from CNS O2 toxicity whereas ATZ enhanced O2 toxicity. In animals pretreated with ATZ, HBO inactivated 21-40% more catalase than air exposure in the six brain regions studied. Because ATZ-mediated inactivation of catalase was H2O2 dependent, the decrease in catalase activity during hyperoxia was proportional to the intracellular production of H2O2. Pargyline, administered 30 min before HBO, inhibited MAO by greater than 90%, prevented ATZ inhibition of catalase activity during HBO, and reversed the augmentation of CNS O2 toxicity by ATZ. These findings indicate that H2O2 generated by MAO during hyperoxia is important to the pathogenesis of CNS O2 toxicity in rats.     

Adrenaline stimulates H2O2 generation in liver via NADPH oxidase

It is known that adrenaline promotes hydroxyl radical generation in isolated rat hepatocytes. The aim of this work was to investigate a potential role of NADPH oxidase (Nox) isoforms for an oxidative stress signal in response to adrenaline in hepatocytes. Enriched plasma membranes from isolated rat liver cells were prepared for this purpose. These membranes showed catalytic activity of Nox isoforms, probably Nox 2 based on its complete inhibition with specific antibodies. NADPH was oxidized to convert O₂ into superoxide radical, later transformed into H₂O₂. This enzymatic activity requires previous activation with either 3 mM Mn²⁺ or guanosine 5′-0-(3-thiotriphosphate) (GTPγS) plus adrenaline. Experimental conditions for activation and catalytic steps were set up: ATP was not required; S_0.5 for NADPH was 44 μM; S_0.5 for FAD was 8 μM; NADH up to 1 mM was not substrate, and diphenyleneiodonium was inhibitory. Activation with GTPγS plus adrenaline was dose- and Ca²⁺-dependent and proceeded through α₁-adrenergic receptors (AR), whereas β-AR stimulation resulted in inhibition of Nox activity. These results lead us to propose H₂O₂ as additional transduction signal for adrenaline response in hepatic cells.     

Site-directed mutagenesis of selected residues at the active site of aryl-alcohol oxidase, an H2O2-producing ligninolytic enzyme

Aryl-alcohol oxidase provides H(2)O(2) for lignin biodegradation, a key process for carbon recycling in land ecosystems that is also of great biotechnological interest. However, little is known of the structural determinants of the catalytic activity of this fungal flavoenzyme, which oxidizes a variety of polyunsaturated alcohols. Different alcohol substrates were docked on the aryl-alcohol oxidase molecular structure, and six amino acid residues surrounding the putative substrate-binding site were chosen for site-directed mutagenesis modification. Several Pleurotus eryngii aryl-alcohol oxidase variants were purified to homogeneity after heterologous expression in Emericella nidulans, and characterized in terms of their steady-state kinetic properties. Two histidine residues (His502 and His546) are strictly required for aryl-alcohol oxidase catalysis, as shown by the lack of activity of different variants. This fact, together with their location near the isoalloxazine ring of FAD, suggested a contribution to catalysis by alcohol activation, enabling its oxidation by flavin-adenine dinucleotide (FAD). The presence of two aromatic residues (at positions 92 and 501) is also required, as shown by the conserved activity of the Y92F and F501Y enzyme variants and the strongly impaired activity of Y92A and F501A. By contrast, a third aromatic residue (Tyr78) does not seem to be involved in catalysis. The kinetic and spectral properties of the Phe501 variants suggested that this residue could affect the FAD environment, modulating the catalytic rate of the enzyme. Finally, L315 affects the enzyme k(cat), although it is not located in the near vicinity of the cofactor. The present study provides the first evidence for the role of aryl-alcohol oxidase active site residues.     

Reduced nicotinamide adenine dinucleotide oxidase activity and H2O2 formation of Mycoplasma pneumoniae

Cell-free extracts of Mycoplasma pneumoniae showed two distinct reduced nicotinamide adenine dinucleotide (NADH(2)) oxidase activities in the supernatant fraction. By ammonium sulfate fractionation and polyacrylamide gel electrophoresis, one activity not requiring flavine co-factors was precipitated by 50 to 70% ammonium sulfate concentration and identified with a slower-moving band on acrylamide gel electrophoresis; a second NADH(2) oxidase activity was flavine mononucleotide (FMN) dependent and associated with a more rapidly moving band; it could only be partially precipitated by ammonium sulfate concentrations ranging from 50 to 100%. Studies with alternate electron acceptors indicated the presence of a menadione, a 2,6-dichlorophenol indophenol and a very weak ferricyanide oxido-reductase activity, but no cytochrome c oxido-reductase, in the cell-free preparations. The NADH(2) oxidase activities of all fractions were relatively cyanide insensitive and were only minimally inhibited by flavoprotein and other respiratory chain inhibitors. H(2)O(2) formation was negligible unless FMN, but not flavine adenine dinucleotide (FAD), was added to the crude NADH(2) oxidase system; upon fractionation and electrophoresis, the H(2)O(2) formation was associated with the FMN-dependent, more rapidly moving NADH(2) oxidase band. This FMN-dependent NADH(2) oxidase-H(2)O(2) generating system may be a mechanism for the H(2)O(2) formation observed during glucose oxidation in the intact organism.     

CD and MCD spectroscopic studies of the two Dps miniferritin proteins from Bacillus anthracis: role of O2 and H2O2 substrates in reactivity of the diiron catalytic centers

DNA protection during starvation (Dps) proteins are miniferritins found in bacteria and archaea that provide protection from uncontrolled Fe(II)/O radical chemistry; thus the catalytic sites are targets for antibiotics against pathogens, such as anthrax. Ferritin protein cages synthesize ferric oxymineral from Fe(II) and O(2)/H(2)O(2), which accumulates in the large central cavity; for Dps, H(2)O(2) is the more common Fe(II) oxidant contrasting with eukaryotic maxiferritins that often prefer dioxygen. To better understand the differences in the catalytic sites of maxi- versus miniferritins, we used a combination of NIR circular dichroism (CD), magnetic circular dichroism (MCD), and variable-temperature, variable-field MCD (VTVH MCD) to study Fe(II) binding to the catalytic sites of the two Bacillus anthracis miniferritins: one in which two Fe(II) react with O(2) exclusively (Dps1) and a second in which both O(2) or H(2)O(2) can react with two Fe(II) (Dps2). Both result in the formation of iron oxybiomineral. The data show a single 5- or 6-coordinate Fe(II) in the absence of oxidant; Fe(II) binding to Dps2 is 30× more stable than Dps1; and the lower limit of K(D) for binding a second Fe(II), in the absence of oxidant, is 2-3 orders of magnitude weaker than for the binding of the single Fe(II). The data fit an equilibrium model where binding of oxidant facilitates formation of the catalytic site, in sharp contrast to eukaryotic M-ferritins where the binuclear Fe(II) centers are preformed before binding of O(2). The two different binding sequences illustrate the mechanistic range possible for catalytic sites of the family of ferritins.     

Azide binding to the trinuclear copper center in laccase and ascorbate oxidase.

Azide binding to the blue copper oxidases laccase and ascorbate oxidase (AO) was investigated by electron paramagnetic resonance (EPR) and pulsed electron-nuclear double resonance (ENDOR) spectroscopies. As the laccase : azide molar ratio decreases from 1:1 to 1:7, the intensity of the type 2 (T2) Cu(II) EPR signal decreases and a signal at g approximately 1.9 appears. Temperature and microwave power dependent EPR measurements showed that this signal has a relatively short relaxation time and is therefore observed only below 40 K. A g approximately 1.97 signal, with similar saturation characteristics was found in the AO : azide (1:7) sample. The g < 2 signals in both proteins are assigned to an S = 1 dipolar coupled Cu(II) pair whereby the azide binding disrupts the anti-ferromagnetic coupling of the type 3 (T3) Cu(II) pair. Analysis of the position of the g < 2 signals suggests that the distance between the dipolar coupled Cu(II) pair is shorter in laccase than in AO. The proximity of T2 Cu(II) to the S = 1 Cu(II) pair enhances its relaxation rate, reducing its signal intensity relative to that of native protein. The disruption of the T3 anti-ferromagnetic coupling occurs only in part of the protein molecules, and in the remaining part a different azide binding mode is observed. The 130 K EPR spectra of AO and laccase with azide (1:7) exhibit, in addition to an unperturbed T2 Cu(II) signal, new features in the g parallel region that are attributed to a perturbed T2 in protein molecules where the anti-ferromagnetic coupling of T3 has not been disrupted. While these features are also apparent in the AO : azide sample at 10 K, they are absent in the EPR spectra of the laccase : azide sample measured in the range of 6-90 K. Moreover, pulsed ENDOR measurements carried out at 4.2 K on the latter exhibited only a reduction in the intensity of the 20 MHz peak of the 14N histidine coordinated to the T2 Cu(II) but did not resolve any significant changes that could indicate azide binding to this ion. The lack of T2 Cu(II) signal perturbation below 90 K in laccase may be due to temperature dependence of the coupling within the trinuclear : azide complex.     

A probe for NADH and H2O2 amperometric detection at low applied potential for oxidase and dehydrogenase based biosensor applications

Modified screen-printed electrodes for amperometric detection of H(2)O(2) and nicotinamide adenine dinucleotide (NADH) at low applied potential are presented in this paper. The sensors are obtained by modifying the working electrode surface with Prussian Blue, a well known electrochemical mediator for H(2)O(2) reduction. The coupling of this sensor with phenazine methosulfate (PMS) in the working solution gives the possibility of measuring both NAD(P)H and H(2)O(2). PMS reacts with NADH producing PMSH, which in the presence of oxygen, gives an equimolar amount of H(2)O(2). This allows the measurement of both analytes with similar sensitivity (357 mA mol(-1)L cm(-2) for H(2)O(2) and 336 mA mol(-1)L cm(-2) for NADH) and LOD (5x10(-7)mol L(-1) for H(2)O(2) and NADH) and opens the possibility of a whole series of biosensor applications. In this paper, results obtained with a variety of dehydrogenase enzymes (alcohol, malic, lactate, glucose, glycerol and glutamate) for the detection of enzymatic substrates or enzymatic activity are presented demonstrating the suitability of the proposed method for future biosensor applications.     

Respiratory burst oxidase homologue‐dependent H2O2 and chloroplast H2O2 are essential for the maintenance of acquired thermotolerance during recovery after acclimation

Thermotolerance is improved by heat stress (HS) acclimation, and the thermotolerance level is “remembered” by plants. However, the underlying signalling mechanisms remain largely unknown. Here, we showed NADPH oxidase‐mediated H₂O₂ (NADPH‐H₂O₂), and chloroplast‐H₂O₂ promoted the sustained expression of HS‐responsive genes and programmed cell death (PCD) genes, respectively, during recovery after HS acclimation. When spraying the NADPH oxidase inhibitor, diphenylene iodonium, after HS acclimation, the NADPH‐H₂O₂ level significantly decreased, resulting in a decrease in the expression of HS‐responsive genes and the loss of maintenance of acquired thermotolerance (MAT). In contrast, compared with HS acclimation, NADPH‐H₂O₂ declined but chloroplast‐H₂O₂ further enhanced during recovery after HS over‐acclimation, resulting in the reduced expression of HS‐responsive genes and substantial production of PCD. Notably, the further inhibition of NADPH‐H₂O₂ after HS over‐acclimation also inhibited chloroplast‐H₂O₂, alleviating the severe PCD and surpassing the MAT of HS over‐acclimation treatment. Due to the change in subcellular H₂O₂ after HS acclimation, the tomato seedlings maintained a constant H₂O₂ level during recovery, resulting in stable and lower total H₂O₂ levels during a tester HS challenge conducted after recovery. We conclude that tomato seedlings increase their MAT by enhancing NADPH‐H₂O₂ content and controlling chloroplast‐H₂O₂ production during recovery, which enhances the expression of HS‐responsive genes and balances PCD levels, respectively.