H₂O₂ negatively regulates aluminum resistance via oxidation and degradation of the transcription factor STOP1

Abstract

Aluminum (Al) stress triggers the accumulation of hydrogen peroxide (H₂O₂) in roots. However, whether H₂O₂ plays a regulatory role in aluminum resistance remains unclear. In this study, we show that H₂O₂ plays a crucial role in regulation of Al resistance, which is modulated by the mitochondrion-localized pentatricopeptide repeat protein REGULATION OF ALMT1 EXPRESSION 6 (RAE6). Mutation in RAE6 impairs the activity of complex I of the mitochondrial electron transport chain, resulting in the accumulation of H₂O₂ and increased sensitivity to Al. Our results suggest that higher H₂O₂ concentrations promote the oxidation of SENSITIVE TO PROTON RHIZOTOXICITY 1 (STOP1), an essential transcription factor that promotes Al resistance, thereby promoting its degradation by enhancing the interaction between STOP1 and the F-box protein RAE1. Conversely, decreasing H₂O₂ levels or blocking the oxidation of STOP1 leads to greater STOP1 stability and increased Al resistance. Moreover, we show that the thioredoxin TRX1 interacts with STOP1 to catalyze its chemical reduction. Thus, our results highlight the importance of H₂O₂ in Al resistance and regulation of STOP1 stability in Arabidopsis (Arabidopsis thaliana).

IN A NUTSHELL

Background: Aluminum (Al) toxicity limits crop production on acid soils. The exposure of plants to Al stress triggers the accumulation of hydrogen peroxide (H₂O₂) in their roots. To combat Al toxicity, plants use the crucial transcription factor SENSITIVE TO PROTON RHIZOTOXICITY 1 (STOP1), which controls Al resistance. STOP1 undergoes various post-translational modifications, such ubiquitination, SUMOylation, and phosphorylation, to regulate its function.

Question: We wished to understand whether H₂O₂ plays a role in regulating Al signaling and Al resistance by affecting STOP1 through post-translational oxidation.

Findings: We identified the specific protein REGULATION OF ALMT1 EXPRESSION 6 (RAE6), which localizes to mitochondria and contains pentatricopeptide repeats. The mutation of RAE6 disrupted the activity of complex I in the mitochondrial electron transport chain. This defect led to increased accumulation of H₂O₂ in mitochondria and rendered rae6 seedlings more sensitive to Al toxicity. We determined that H₂O₂ promotes the oxidation of STOP1 at specific cysteine residues (C8, C27, and C185), which triggers its degradation by enhancing its interaction with the F-box protein RAE1. However, to counteract the negative effects of H₂O₂ on STOP1, another protein, thioredoxin TRX1, can interact with STOP1 and reverse its oxidation.

Next steps: While we learned that H₂O₂ negatively regulates the accumulation of STOP1, it is not the signal that activates another pathway composed of a series of kinases (the MEKK1–MKK1/1–MPK4 cascade), which positively promotes STOP1 accumulation via its phosphorylation. Further research is needed to understand how plants sense Al stress and transmit the signal to the MEKK1–MKK1/1–MPK4 cascade, ultimately leading to an increase in STOP1 protein levels.

Introduction

In acids soils, a portion of aluminum (Al) becomes solubilized into trivalent Al (Al³⁺), which can be toxic to root tip cells, inhibiting root growth and ultimately decreasing plant production. Since acid soils make up over 30% of world's arable land, Al toxicity has been recognized as the second most significant abiotic limitation to crop production, exceeded only by drought stress (von Uexkull and Mutert 1995; Kochian et al. 2015).

The phytotoxic Al³⁺ ion interacts strongly with oxygen donor compounds that contain carboxyl or phosphate groups, allowing it to target various sites in root cells for its toxicity, including the cell wall, plasma membrane, and symplastic components (Kochian 1995; Ma 2007; Poschenrieder et al. 2008). One early symptom and response to Al stress is the accumulation of reactive oxygen species (ROS) such as hydrogen peroxide (H₂O₂) in root tips, which has been observed several plant species, including pea (Pisum sativum) (Matsumoto and Motoda 2013), rice (Oryza sativa) (Ma et al. 2007), barley (Hordeum vulgare) (Tamas et al. 2004), and Melaleuca trees (Tahara et al. 2008). In non-photosynthesizing root cells, the mitochondrial electron transport chain is a significant source of ROS production (Moller et al. 2007). Physiological studies have shown that Al can induce H₂O₂ production in mitochondria by impairing mitochondrial function (Yamamoto et al. 2002; Huang et al. 2014; Zhan et al. 2014). Although H₂O₂ accumulation can be toxic to plant cells, it can also act as a signaling molecule to regulate plant stress responses (Castro et al. 2021; Mittler et al. 2022). However, it is currently unclear whether and how H₂O₂ plays a regulatory role in Al resistance.

Pentatricopeptide repeat (PPR) proteins are a group of modular proteins containing repeats of 35 amino acids; PRR protein is present in eukaryotes, with many more members in plants than other species (Schmitz-Linneweber and Small 2008; Barkan and Small 2014). The PPR family in plants can be divided into 2 subfamilies: P and PLS. P-type PPR proteins contain the canonical 35 amino-acid repeat (P) motif, while members of the PLS subfamily contain the P motif as well as longer (L) and short (S) variant PPR motifs (Fujii and Small 2011; Barkan and Small 2014). Most PPR proteins are located in mitochondria or chloroplasts, where they bind to 1 or several organellar transcripts to regulate their expression post-transcriptionally via RNA splicing, editing, cleavage, or translation (Barkan and Small 2014). Mutations in PPR genes that encode mitochondrion-localized PPR proteins frequently cause defects in embryo or seed development (Barkan and Small 2014), which can be attributed to an impairment of the mitochondrion electron transport chain.

Mitochondrial NADH-ubiquinone reductase (complex I), the major entry point of the electron transport chain, consists of 40 to 50 subunits, 9 of which are encoded by mitochondrial genes (Heazlewood et al. 2003; Sugiyama et al. 2005) that are frequently targeted by PPR proteins for promoting their RNA cleavage or editing (Yuan and Liu 2012; Zhu et al. 2012; Colas des Francs-Small et al. 2014; Leu et al. 2016; Wang et al. 2018; Yang et al. 2022). The Arabidopsis (Arabidopsis thaliana) genome encodes 450 PPR proteins, and different PPR members appear to show little functional redundancy (Barkan and Small 2014). Thus, many PPR proteins localizing to mitochondria remain to be identified and characterized. While some mutants in individual PPR genes have been reported to display defective responses to different stresses (Liu et al. 2010; Yuan and Liu 2012; Zsigmond et al. 2012; Zhu et al. 2014; Yang et al. 2022), the involvement of PPR proteins in Al resistance is currently unknown.

SENSITIVE TO PROTON RHIZOTOXICITY 1 (STOP1) is a C2H2-type zinc finger transcription factor that plays an essential role in regulating Al resistance and low pH tolerance (Iuchi et al. 2007). Recently, STOP1 has been demonstrated to be the master regulator for tolerance to a variety of other stresses, including low levels of phosphate, potassium, and oxygen (Balzergue et al. 2017; Mora-Macias et al. 2017; Enomoto et al. 2019; Wang et al. 2022), as well as salt and excess ammonium (Sadhukhan et al. 2019; Sadhukhan et al. 2021; Tian et al. 2021). STOP1 primarily regulates Al resistance by inducing the expression of ALUMINUM-ACTIVATED MALATE TRANSPORTER 1 (ALMT1), whose encoding transporter mediates malate secretion (Iuchi et al. 2007), although the increased expression of other STOP1 target genes, including MULTI-DRUG AND TOXIC COMPOUND EXTRUSION (MATE), ALUMINUM SENSITIVE 3 (ALS3), and GLUTAMATE DEHYDROGENASE 1 (GDH1) and GDH2, also contributes to Al resistance (Liu et al. 2009; Sawaki et al. 2009; Tokizawa et al. 2021). Recent research has shown that STOP1 is the subject of multiple layers of post-transcriptional regulation, including ubiquitination, SUMOylation, and phosphorylation (Zhang et al. 2019; Fang et al. 2020, 2021a, 2021b; Xu et al. 2021; Zhou et al. 2023); in addition, nucleocytoplasmic export of STOP1 mRNA is under regulation (Guo et al. 2020). Al stress can trigger the MITOGEN-ACTIVATED PROTEIN KINASE KINASE KINASE 1 (MEKK1)–MITOGEN-ACTIVATED PROTEIN KINASE KINASE 1/2 (MKK1/2)–MITOGEN-ACTIVATED PROTEIN KINASE 4 (MPK4) cascade, which phosphorylates STOP1 (Zhou et al. 2023). Phosphorylation of STOP1 prevents its interaction with the F-box protein REGULATION OF ALMT1 EXPRESSION (RAE1), which mediates STOP1 ubiquitination and degradation, leading to an increase in STOP1 accumulation. H₂O₂ can mediate oxidative modification of proteins, including transcription factors, to alter their activity or stability (Peleg-Grossman et al. 2010; Marinho et al. 2014; Tian et al. 2018). It is unclear whether STOP1 is also subjected to H₂O₂-mediated oxidative modification to regulate its function.

In this study, we identified the mitochondrion-localized PPR protein RAE6 involved in modulation of mitochondrial nad5 splicing. We show that mutation of RAE6 impairs complex I activity of the mitochondrion electron transport chain and consequently induces H₂O₂ accumulation. Increased H₂O₂ accumulation promotes STOP1 oxidation, leading to RAE1-mediated STOP1 degradation, ultimately resulting in increased Al sensitivity. Conversely, decreasing H₂O₂ content or blocking STOP1 oxidation increased STOP1 stability and Al resistance. Our results demonstrate the important role of H₂O₂ in regulating Al resistance and STOP1 stability.

Results

Identification of rae6 with lower STOP1 accumulation and Al resistance

We previously conducted a forward genetic screen on an ethyl methanesulfonate-mutagenized population in the proALMT1:LUC background, harboring the ALMT1 promoter driving the firefly luciferase (LUC) reporter gene. From this screen, we identified a series of rae (regulation of ALMT1 expression) mutants with altered LUC activity (Zhang et al. 2019), including the rae6 mutant with decreased LUC signal (Fig. 1A). A reverse transcription quantitative PCR (RT-qPCR) analysis showed that the expression of ALMT1, MATE, and ALS3, which are regulated by STOP1 (Iuchi et al. 2007; Liu et al. 2009; Sawaki et al. 2009), is lower in rae6 than in the wild-type (WT) control especially under +Al conditions (Fig. 1B). By contrast, the genes SENSITIVE TO AL RHIZOTOXICITY 1 (STAR1), ALS1, and STOP1, which are not regulated by STOP1, were expressed at a similar level in the WT and the rae6 mutant (Supplemental Fig. S1A), suggesting that mutation of RAE6 impairs STOP1 function.

Figure 1.

Mutation of RAE6 decreases the expression of STOP1-regulated genes, STOP1 stability, and Al resistance. A) Decreased LUC signal of proALMT1:LUC in the rae6 mutant. Scale bar, 1 cm. B) Lower expression of ALMT1, MATE, and ALS3 in the rae6 mutant. Seedlings of wild type (WT, Col-0) and rae6 were cultivated on gel medium soaked with no (−Al) or 0.75 mM Al (+Al) for 7 d and then the roots were collected for RT-qPCR analysis. C) Representative photographs (left panel) and quantitative data (right panel) of GUS levels in roots of WT and rae6 seedlings carrying the proSTOP1:STOP1-GUS transgene exposed to a 0.5 mM CaCl₂ solution containing no (−Al) or 30 μM Al (+Al) for 8 h. Scale bars, 200 μm. D) Representative immunoblot (left panel) and quantitative data (right panel) of STOP1-HA levels. WT and rae6 seedlings carrying the proSTOP1:STOP1-HA transgene were grown on soaked gel medium containing no (−Al) or 0.75 mM Al (+Al) for 7 d; the roots were then collected for immunoblot analysis. Actin served as a loading control. E) Decreased STOP1 stability in rae6. The roots of WT and rae6 seedlings harboring the proSTOP1:STOP1-3HA transgene were treated with 30 μM Al and/or 100 μM cycloheximide (CHX) alone for varying time intervals as indicated. F) Decreased malate release in rae6. WT and rae6 seedlings were treated with a 0.5 mM CaCl₂ solution containing no Al or 20 μM Al for 12 h; the root exudates were analyzed for malate concentrations (pmol plant⁻¹ h⁻¹). G) Representative photographs (left panel) and quantitative data (right panel) of Al resistance in WT, rae6, and hpr1. Seedlings were grown on gel medium soaked with no Al or 0.75, 1, or 1.25 mM Al for 7 d; root relative length was measured to evaluate Al resistance. Scale bars, 1 cm. H)rae6 accumulates more Al than WT in response to Al treatment. WT, rae6, and hpr1 seedlings were exposed to a 0.5 mM CaCl₂ solution containing no Al or 20 μM Al for 12 h and then the Al content in the roots was determined. Data shown in B to D), F), and H) are means ± Sd of 3 biological replicates; data shown in G) are means ± Sd of the relative root length from 19 to 34 seedlings. Asterisks indicate statistically different values (Student's t test, *P < 0.05, **P < 0.01, ***P < 0.001). n.s., not significantly different. Different lowercase letters indicate significantly different means (P < 0.05, ANOVA followed by Tukey test).

To investigate whether STOP1 protein accumulation is affected in rae6 mutant, we crossed proSTOP1:STOP1-GUS (encoding a fusion between STOP1 and the β- glucuronidase [GUS] reporter) and proSTOP1:STOP1-HA transgenic lines, which were produced previously (Zhang et al. 2019), to the rae6 mutant to introduce these transgenes into the mutant background. GUS staining and immunoblot analyses revealed that the STOP1-GUS and STOP1-HA protein levels are lower in the rae6 background compared to WT under both −Al and +Al conditions (Fig. 1, C and D). To examine whether mutation of RAE6 altered STOP1 stability, we treated roots with cycloheximide (CHX), a protein biosynthesis inhibitor, and exposed these roots to Al stress over time. An immunoblot analysis showed that STOP1-HA abundance decreases faster in rae6 than in WT (Fig. 1E), indicating that the rae6 mutation diminishes STOP1 stability.

Because of the lower expression of ALMT1 in rae6, we compared the level of malate secretion in WT and the mutant: We determined that the mutant secretes less malate than the WT under both −Al and +Al conditions (Fig. 1F). We then determined the Al resistance phenotype of rae6 by exposing seedlings of the WT, the rae6, and the Al-sensitive mutant control hyper recombination1 (hpr1, also named rae3) (Guo et al. 2020) to different Al concentrations. Under control (−Al) conditions, the root length of rae6 seedlings was similar to that of WT (Fig. 1G). However, under Al stress conditions, the relative root length of rae6 and hpr1 was much shorter than that of WT (Fig. 1G), indicating that rae6 seedlings are more sensitive to Al than the WT. In agreement with these results, the rae6 and hpr1 mutants accumulated more Al than the WT in response to Al treatment (Fig. 1H). These data demonstrate that the rae6 mutant decreases malate secretion and Al resistance.

In addition, the mutation of RAE6 also adversely affected plant development. The rae6 mutant displayed delayed flowering, shorter stature, and shorter siliques (Supplemental Fig. S1, B to D).

RAE6 encodes a mitochondrial-localized P-type PPR protein

We conducted a genetic analysis of rae6 using a segregating F₂ population derived from a cross between rae6 and Col-0. Among a total of 392 F₂ seedlings, 89 and 303 plants showed decreased and WT-like LUC signals, respectively. The ratio of the number of plants with a decreased LUC signal over the number of plants with a WT-like LUC signal fitted a 1:3 ratio (χ² = 1.1, P > 0.25), suggesting that the decreased LUC signal in rae6 is caused by a single recessive mutation.

To clone the RAE6 gene, we pooled genomic DNA from the 89 F₂ plants with decreased LUC signal for whole-genome sequencing. We used the genomic sequence from the WT as control, which was sequenced previously (Zhang et al. 2019). We located the candidate mutation to a small region of chromosome 1 through MutMap analysis (Abe et al. 2012) (Supplemental Fig. S2). To verify the results of the MutMap analysis, we developed 5 derived cleaved amplified polymorphic sequence (dCAPS) markers within the candidate region based on the detected mutations in rae6 relative to the WT sequence (Supplemental Table S1) and used the 89 F₂ plants to perform a marker linkage analysis. All 5 markers exhibited linkage to the mutant phenotype albeit to different extents, confirming that the RAE6 gene is located in this region. We mapped RAE6 between markers A1 and A3 with 5 and 3 remaining recombinants, respectively; notably, the A2 marker, based on a C-to-T substitution at the 1,981 bp position downstream from the start codon of At1g73710, showed complete linkage to the mutant phenotype (Supplemental Table S1). This C-to-T substitution results in an amino acid change from leucine to phenylalanine (L661F) in At1g73710 in the rae6 mutant. These results suggest that the L661F mutation in the protein encoded by At1g73710 may be responsible for the lower LUC signal and Al resistance observed in rae6.

To confirm that At1g73710 is RAE6, we attempted to construct a proAt1g73710:At1g73710 transgene for complementation of rae6. However, we were unable to clone the At1g73710 DNA into a plasmid, which might be attributed to the complex secondary structure of At1g73710 DNA and/or the toxicity of At1g73710 when present in bacteria. As an alternative, we used an adenine base editor (A-to-G) (Wei et al. 2021) to convert the T-to-C mutation detected in At1g73710 in rae6 back to the wild-type sequence of At1g73710. We obtained 1 plant (rae6^F661L) with the intended phenylalanine to leucine change (F661L) in At1g73710 in the rae6 background from a screen of ∼120 transgenic plants. An RT-qPCR analysis showed that the decreased expression of STOP1-regulated genes in rae6 returns to wild-type levels in rae6^F661L (Supplemental Fig. S3A). The lower Al resistance phenotype of rae6 was also rescued in rae6^F661L (Supplemental Fig. S3B). In addition, we ordered 1 T-DNA line (GK-375A05-017161) with a T-DNA insertion in At1g73710 (Supplemental Fig. S4A). However, we were unable to obtain homozygous mutant plants. Observation of seeds from a selfed heterozygous plant showed that 81 out of 316 seeds are aborted (Supplemental Fig. S4B), suggesting that the knockout of At1g73710 leads to embryo death. We further knocked down the transcript levels of At1g73710 via RNA interference (RNAi) (Supplemental Fig. S4C) to determine its function. Similar to rae6, the expression of STOP1-regulated genes and Al resistance were also lower in 2 At1g73710 RNAi lines compared to WT (Supplemental Fig. S4, C and D). Together, these results indicate that RAE6 is At1g73710, whose mutation causes a decrease in the expression of STOP1 target genes and Al resistance.

RAE6 encodes a P-type PPR family protein harboring 20 PPR motifs (Supplemental Fig. S4A). RT-qPCR analysis showed that RAE6 is expressed at similar levels in all tissues examined (Supplemental Fig. S5A), and Al stress and the rae6 mutation did not affect RAE6 expression (Supplemental Fig. S5B). To determine the tissue-specific expression pattern of RAE6, we constructed proRAE6:GUS transgenic lines by driving the GUS reporter gene with the RAE6 promoter. GUS staining analysis detected a GUS signal preferentially in root tips (Supplemental Fig. S5C). In shoots, we detected GUS staining in leaf vascular tissues and pollen (Supplemental Fig. S5C).

To investigate the subcellular localization of RAE6, we amplified a fusion construct consisting of proUBQ10:RAE6-GFP (encoding a fusion between RAE6 and the green fluorescent protein) and introduced the resulting PCR product into Arabidopsis protoplasts via transfection, since we could not clone RAE6 in a bacterial vector, as mentioned earlier. While we observed GFP fluorescence in both the cytoplasm and nucleus in protoplasts transfected with 35S:GFP, we determined that RAE6-GFP co-localizes with mitochondria, labeling with the fluorescent dye Mito Tracker (Fig. 2A), suggesting that RAE6 is localized to mitochondria.

Figure 2.

Subcellular localization of RAE6 and effect of the rae6 mutation on nad5 splicing and mitochondrial complex I accumulation. A) RAE6-GFP co-localizes with the mitochondrion fluorescent dye Mito Tracker in Arabidopsis protoplasts. Scale bars, 5 μm. B) RT-PCR analysis showing that the level of full-length nad5 transcript is lower in rae6 than in WT but returned to WT levels in rae6^F661L. C, D) Splicing diagram C) and splicing efficiency D) of nad5 transcripts in WT, rae6^F661L, and rae6. RT-qPCR analysis was carried out to calculate the relative proportions of spliced (s) to unspliced (u) forms of nad5 introns. E) Blue native PAGE gel analysis of the different mitochondrial respiratory complexes in WT, rae6^F661L, and rae6. F) Mutation of RAE6 decreases mitochondrial complex I activity in roots. Data shown in D) and F) are means ± Sd of 3 biological replicates. Different lowercase letters indicate significantly different means (P < 0.05, ANOVA test followed by Tukey test).

Mutation of RAE6 impairs the mitochondrial electron transport chain and induces H₂O₂ accumulation

A major function for P-type PPR proteins in chloroplasts and mitochondria is proposed to be the stabilization or splicing of specific organelle RNAs (Barkan and Small 2014). Given that RAE6 is localized to mitochondria (Fig. 2A), we compared the expression levels of 25 mitochondrial genes carrying no introns in WT and rae6 by RT-qPCR, which revealed that the transcript levels of these genes are similar between WT and rae6 (Supplemental Fig. S6A). We then compared the levels of full-length mature transcripts for 9 mitochondrial genes containing introns by end-point RT-PCR. We determined that the levels of full-length mature nad5 transcripts are markedly lower in rae6 compared to WT (Fig. 2B), whereas the levels of other transcripts did not show a significant difference between WT and the mutant (Supplemental Fig. S6B). To investigate whether the rae6 mutation affected the splicing of nad5 pre-RNA transcripts, which consists of 2 cis-spliced introns and 2 trans-spliced introns (Fig. 2C), we examined the splicing efficiencies of the 4 nad5 introns in WT and rae6: We established that their splicing efficiencies are all lower in rae6 compared to WT (Fig. 2D). The lower accumulation of mature nad5 transcript and the defective splicing of nad5 in rae6 was rescued to WT levels in the rae6^F661L line (Fig. 2, B and D).

Since nad5 encodes a subunit of mitochondrial complex I (NADH-ubiquinone oxidoreductase), we analyzed the accumulation of complex I in mitochondrial membrane proteins by blue native (BN)-PAGE: The abundance of complex I was decreased in rae6 compared to WT (Fig. 2E). In concert with the lower accumulation of complex I in the mutant, rae6 showed a decreased activity of complex I in comparison to the WT (Fig. 2F); these 2 phenotypes were again rescued back to WT levels in the rae6^F661L line (Fig. 2, E and F). Together, these results demonstrate that mutation of RAE6 decreases the abundance and activity of mitochondrial complex I, likely through impairing nad5 splicing.

Complex I is a major ROS production site in mitochondria; decreasing its function has been shown to induce ROS accumulation (Moller 2001; Yang et al. 2022). We thus compared ROS accumulation in WT and rae6 under −Al and +Al conditions by using the ROS fluorescent probe 2′,7′-dichlorofluorescin diacetate (H₂DCFDA): We observed an induction of ROS accumulation under Al stress in the WT and in the rae6 mutant regardless of Al stress, although treatment with Al further increased ROS accumulation (Fig. 3A). We further investigated where ROS accumulate by analyzing the co-localization of the ROS fluorescent probe with the mitochondria-labeling fluorescent dye Mito Tracker. We observed the accumulation of ROS within mitochondria following Al stress and in the rae6 mutant (Fig. 3B). To examine the specific types of ROS being accumulated, we stained roots for superoxide and H₂O₂ using nitroblue tetrazolium (NBT) and 3,3′-diaminobenzidine (DAB), respectively. These staining assays demonstrated that the rae6 mutant accumulates more H₂O₂ than WT, with Al stress also triggering H₂O₂ accumulation (Fig. 3C). However, we did not observe a significant difference in superoxide accumulation between WT and the mutant (Fig. 3D). Additionally, we measured the H₂O₂ production rate in isolated mitochondria and found that both Al stress and the rae6 mutant increase the rate of mitochondrial H₂O₂ production (Fig. 3E). Consistent with the increased mitochondrial H₂O₂ production, the expression of ALTERNATIVE OXIDASE (AOX) genes, whose expression can be triggered by mitochondrial ROS (Maxwell et al. 1999; Vanlerberghe 2013), was induced significantly in the mutant compared to WT (Fig. 3F). These findings underscore the finding that Al stress and the rae6 mutant can trigger mitochondrial H₂O₂ production.

Figure 3.

Rae6 accumulates more H₂O₂ than WT. A, B) Effect of the rae6 mutation on total ROS A) and mitochondrial ROS B) levels in the roots. Seven-day-old seedlings were transferred to soaked gel medium containing no Al or 0.75 mM Al for 12 h. ROS were visualized using the fluorescence probe H₂DCFDA, and mitochondria were stained using MitoTracker Red. White arrowheads indicate ROS accumulation in root mitochondria. Gray value plots on the right show the relative fluorescence along the dotted arrow line in the images. C, D) Effect of the rae6 mutation on H₂O₂C) and O₂⁻D) accumulation. DAB and NBT were used to stain H₂O₂ and O₂⁻, respectively. E) Effect of the rae6 mutation and Al stress on rates of mitochondrial H₂O₂ production. The roots of 10-d-old seedlings treated with either no AlCl₃or 0.75 mM AlCl₃-soaked gel medium for 12 h were subjected to mitochondrial isolation. The rates of H₂O₂ production in isolated mitochondria were determined by fluorometric monitoring of Amplex Red conversion to resorufin. F) Increased expression of AOX genes in rae6. Scale bars in A), C), and D) represent 200 μm; or 20 μm in B). Data shown are means ± Sd of 3 biological replicates. Different lowercase letters indicate significantly different means (P < 0.05, ANOVA followed by Tukey test). Asterisks indicate statistically different values (Student's t test, **P < 0.01, ***P < 0.001).

Elevated H₂O₂ weakens Al resistance of rae6 partly through limiting STOP1 accumulation

To investigate whether the increased H₂O₂ levels in rae6 cause the observed lower Al resistance and STOP1 accumulation, we applied glutathione (GSH) to scavenge H₂O₂ and characterized the resulting Al resistance and expression of STOP1-regulated genes. The supply of GSH significantly rescued the Al-sensitive phenotype of rae6, as measured by relative root length, and partially rescued the decreased expression of STOP1-regulated genes in rae6 (Supplemental Fig. S7, A and B). To corroborate this pharmacological result, we overexpressed CATALASE 2 (CAT2) encoding a H₂O₂-scavenging enzyme in WT and rae6 with or without a proSTOP1:STOP1-HA transgene. While overexpression of CAT2 in WT slightly increased Al resistance to a high Al concentration, CAT2 overexpression rescued the Al-sensitive phenotype of rae6 at all Al concentrations tested (Fig. 4A). Immunoblot and expression analyses revealed that the lower abundance of STOP1-HA and the decreased expression of STOP1-regulated genes in rae6 all return to WT levels upon the overexpression of CAT2 (Fig. 4, B and C). We also introduced a mutant in RESPIRATORY BURST OXIDASE HOMOLOG D (RBOHD) (Supplemental Fig. S8A), whose encoded protein is required for normal H₂O₂ production, into the rae6 background through crossing. The rbohD mutant had a marked decrease in RBOHD expression (Supplemental Fig. S8B) and largely rescued the increased Al sensitivity, lower accumulation of STOP1-HA, and the diminished expression of STOP1-regulated genes seen in rae6 when combined into the rbohD rae6 double mutant (Supplemental Fig. S8, C to E). These results indicate that the increased H₂O₂ accumulation causes the Al-sensitive phenotype of rae6 partially through decreasing STOP1 accumulation.

Figure 4.

Overexpression of CAT2 rescues the increased Al sensitivity and decreased STOP1 accumulation of rae6. A) Overexpression of CAT2 rescues the Al-sensitive phenotype of rae6. Seedlings of WT, rae6, 2 CAT2 overexpression lines (CAT2-OE), and 2 CAT2-OE in the rae6 background were cultivated on gel medium soaked with no Al or 0.75, 1, or 1.25 mM Al for 7 d; root relative length was measured to evaluate Al resistance. Scale bars, 1 cm. Values are means ± Sd (n = 18 to 27). B, C) Effect of CAT2 overexpression on STOP1 accumulation B) and the expression of STOP1-regulated genes C). Seedlings were grown on gel medium soaked with no Al or 0.75 mM Al for 7 d; the roots were collected for the determination of STOP1-HA accumulation and the expression of STOP1-regulated genes. Values are means ± Sd of 3 biological replicates. Different lowercase letters indicate significantly different means (P < 0.05, ANOVA test followed by Tukey test).

H₂O₂ induces STOP1 oxidation and negatively modulates STOP1 accumulation

H₂O₂ has well been documented to post-translationally regulate target proteins via cysteine oxidative modification (Moller et al. 2007; Tian et al. 2018; Lu et al. 2022). To examine whether H₂O₂ directly induces STOP1 oxidation, we treated recombinant STOP1-His with different concentrations of H₂O₂, followed by incubation with biotin-conjugated iodoacetamide (BIAM), which can compete with H₂O₂ to react with cysteine residues (Kim et al. 2000). An immunoblot analysis using an anti-biotin antibody showed that the abundance of biotin-labeled STOP1 decreases in the presence of H₂O₂ in a dose-dependent manner (Fig. 5A), suggesting that H₂O₂ promotes cysteine oxidation of STOP1. To determine whether STOP1 is oxidized in vivo, we extracted total proteins from the roots of WT and rae6 seedlings harboring the proSTOP1:STOP1-HA transgene, after they were pretreated with H₂O₂ or Al. We employed a biotin-switch assay (Li and Kast 2017) to detect oxidized STOP1: Both H₂O₂ and Al were effective in triggering STOP1 oxidation, and mutation of RAE6 increased STOP1 oxidation compared to WT under both normal and Al stress conditions (Fig. 5, B and C).

Figure 5.

STOP1 is oxidized in vitro and in vivo. A) Analysis of the oxidative modification of STOP1 by a BIAM-labeling assay. B) Effect of H₂O₂ and the rae6 mutation on STOP1 oxidation. Ten-day-old WT and rae6 seedlings carrying the proSTOP1:STOP1-HA transgene were treated with no or 1 mM H₂O₂ for 3 h, and then the roots were collected for the determination of oxidized STOP1-HA through the biotin-switch assay. C) Effect of Al and the rae6 mutation on STOP1 oxidation. WT and rae6 seedlings carrying the proSTOP1:STOP1-HA transgene were grown on gel medium soaked with no Al or 0.75 mM Al for 7 d, and the biotin-switch assay was used to determine the oxidation state of STOP1-HA. D) Effect of single or triple C-to-S mutations at the C8, C27, and C185 residues of STOP1 on the oxidation of recombinant purified STOP1-His treated with no or 1 mM H₂O₂ for 15 min. E to I) Effect of triple C-to-S mutations (3CS) at the C8, C27, and C185 residues of STOP1 on H₂O₂- E) and Al- F) induced oxidation of STOP1, STOP1 accumulation G), H₂O₂-induced degradation of STOP1 H), and Al resistance I). For H₂O₂ treatment, 9-d-old seedlings of proSTOP1:STOP1-HA and 2 proSTOP1:STOP1^3CS-HA (#1 and #2) transgenic lines in the stop1 background were exposed to no or 1 mM H₂O₂ for 3 h. For Al treatment, seedlings were grown on gel medium soaked with no Al or 0.75 mM Al for 7 d. For Al resistance evaluation, seedlings were grown on gel medium soaked with no Al or 0.75, 1, or 1.25 mM Al for 7 d. Scale bars, 1 cm. Data shown in G) represent the means ± Sd of 3 biological replicates; the data shown in I) represent the means ± Sd (n = 20 to 24). Different lowercase letters indicate significantly different means (P < 0.05, ANOVA followed by Tukey test).

To determine which cysteine sites in STOP1 are oxidized by H₂O₂, we performed a liquid chromatography-tandem mass spectrometry analysis. Accordingly, we subjected recombinant STOP1-His to sequential treatment with H₂O₂, N-ethylmaleimide (NEM), and BIAM, followed by trypsin digestion. This analysis revealed that the cysteine residues Cys-8, Cys-27, and Cys-185 are oxidized residues in STOP1 (Supplemental Fig. S9). Subsequently, we replaced each cysteine by serine residues individually at the C8, C27, or C185 positions to investigate the major oxidized sites in STOP1. However, the single C-to-S mutations did not significantly decrease STOP1 oxidation (Fig. 5D). We thus generated the triple mutant C8S, C27S, C185S (3CS thereafter); remarkably, we established that H₂O₂-induced STOP1 oxidation is significantly diminished in this variant (Fig. 5D), suggesting that all 3 cysteine residues are necessary for STOP1 oxidation in vitro.

To explore whether these 3 cysteine residues are oxidized in vivo, we introduced the proSTOP1:STOP1^3CS-3HA (STOP1^3CS) construct into the stop1 mutant background. We selected 2 STOP1^3CS lines with similar STOP1 expression levels as the proSTOP1:STOP1-3HA (STOP1) complementation line (Supplemental Fig. S10). A biotin-switch assay demonstrated that oxidized STOP1 cannot be detected in the 2 STOP1^3CS lines, even in the presence of Al or H₂O₂, whereas we readily detected oxidized STOP1 in the STOP1 line (Fig. 5, E and F). These results indicate that Al- or H₂O₂-induced STOP1 oxidation specifically occurs at the C8, C27, and C185 residues in Arabidopsis.

To determine whether oxidation of STOP1 modulates its function, we compared the expression of STOP1-regulated genes in the STOP1 and STOP1^3CS lines. We established that STOP1-regulated genes are expressed at higher levels in the STOP1^3CS lines than in the STOP1 line (Supplemental Fig. S10A). In addition, the STOP1^3CS lines accumulated more STOP1 than the STOP1 line (Fig. 5G). Moreover, H₂O₂ treatment rapidly promoted the degradation of STOP1, but not of STOP1^3CS (Fig. 5H). Phenotypic analysis showed that the STOP1^3CS lines are more tolerant to Al stress than the STOP1 line (Fig. 5I). Additionally, the introduction of the STOP1^3CS transgene into the rae6 mutant background via crossing returned the Al-resistant phenotype of rae6 back to WT level (Supplemental Fig. S10B). Taken together, these results suggest that Al stress triggers the oxidation of STOP1 at the C8, C27, and C185 residues, which negatively regulates STOP1 accumulation.

H₂O₂ promotes STOP1 degradation via enhancing its interaction with the F-box protein RAE1

To investigate whether H₂O₂ promotes STOP1 degradation via the 26S proteasome pathway, we treated the roots of WT and rae6 seedlings carrying the proSTOP1:STOP1-HA transgene with MG132, an inhibitor of the 26S proteasome, or maintained them on normal growth medium as control. The lower STOP1 accumulation in rae6 returned to WT levels after treatment with MG132 under both −Al and +Al conditions (Fig. 6A). The F-box protein RAE1 has been shown to play an important role in regulating STOP1 stability via the 26S proteasome pathway, with its homolog RAE1 HOMOLOG 1 (RAH1) also contributing to this regulation (Zhang et al. 2019; Fang et al. 2021b). We therefore generated a rae1 rae6 double mutant alone or with the proSTOP1:STOP1-HA transgene by crossing. Phenotypic analysis revealed that the rae1 mutation largely abolishes the increased Al sensitivity of rae6 (Fig. 6B). Immunoblot and expression analyses showed that the lower accumulation of STOP1-HA and decreased expression of STOP1-regulated genes in rae6 also return to WT levels in the rae1 rae6 double mutant (Fig. 6C; Supplemental Fig. S11). These results suggest that H₂O₂ promotes the degradation of STOP1 via the RAE1/RAH1-mediated 26S proteasome pathway.

Figure 6.

H₂O₂ promotes RAE1-mediated STOP1 degradation. A) Inhibition of rae6 mutation-induced STOP1 degradation by MG132, an inhibitor of the 26S proteasome. B) Mutation of RAE1 rescues the Al-sensitive phenotype of rae6. WT, rae6, rae1, and rae1 rae6 seedlings were grown on gel medium soaked with no Al or 0.75, 1, or 1.25 mM Al for 7 d. Scale bars, 1 cm. C) Mutation of RAE1 rescues the rae6 mutation-induced degradation of STOP1. WT, rae6, rae1, and rae1 rae6 seedlings carrying the proSTOP1:STOP1-HA transgene were grown on gel medium soaked with no Al or 0.75 mM Al for 7 d, and then the STOP1-HA level was determined by immunoblot analysis. D, E) Effect of H₂O₂ and triple C-to-S mutations (3CS) at the C8, C27, and C185 residues of STOP1 on the interaction between STOP1 and RAE1 lacking the F-box domain (RAE1ΔF) in pull-down D) and Co-IP E) assays. For the pull-down assay, recombinant GST-RAE1ΔF was incubated with STOP1-His or STOP1^3CS-His alone or with 1 mM H₂O₂ at 4 °C for 4 h. For the Co-IP assay, RAE1ΔF-HA was co-transfected with STOP1-FLAG or STOP1^3CS-FLAG in rae1 rah1 protoplasts. Crude protein extracts were incubated with anti-HA magnetic beads alone or with 1 mM H₂O₂ at 4 °C for 4 h. Data shown in B) represent the means ± Sd (n = 18 to 27); the data shown in C to E) represent the means ± Sd of 3 biological replicates. Different lowercase letters indicate significantly different means (P < 0.05, ANOVA followed by Tukey test).

To investigate how H₂O₂ promotes the RAE1-mediated degradation of STOP1, we tested the interaction between STOP1-His or STOP1^3CS-His and a version of RAE1 lacking the F-box domain (RAE1ΔF) fused to glutathione S-transferase (GST) in a pull-down assay. We showed that RAE1ΔF has a stronger interaction with STOP1 than with STOP1^3CS, and H₂O₂ treatment promoted the interaction between RAE1ΔF and STOP1 (Fig. 6D). To confirm this observation, we performed a co-immunoprecipitation (Co-IP) assay in Arabidopsis protoplasts. Application of H₂O₂ also promoted the co-immunoprecipitation of RAE1ΔF-HA with STOP1-FLAG, and the association of RAE1ΔF-HA with STOP1-FLAG was stronger than with STOP1^3CS-FLAG (Fig. 6E). These results suggest that H₂O₂-induced STOP1 oxidation enhances the interaction between STOP1 and RAE1, resulting in STOP1 degradation.

The thioredoxin TRX1 catalyzes STOP1 reduction

To counteract the H₂O₂-mediated thiol oxidation of STOP1 cysteine residues, reducing systems are required to catalyze the conversion of STOP1 from its oxidized to its reduced form. The thioredoxin (TRX) system is thought to play an important role in cellular thiol-redox control in eukaryotes (Garcia-Santamarina et al. 2014). While most plant TRXs localize to chloroplasts or mitochondria, h-type TRXs typically localize in the cytoplasm or nucleus (Meyer et al. 2012). A previous study showed that the h-type thioredoxin TRX1 positively regulates Al resistance (Nakano et al. 2020), but the underlying Al resistance mechanism was unknown. We hypothesized that TRX1 might catalyze the reduction of STOP1 to confer Al resistance. Therefore, we investigated whether TRX1 can interact with STOP1. A pull-down assay indicated that GST-TRX1, but not GST alone, can form bind to STOP1-His (Fig. 7A), indicating that TRX1 can directly interact with STOP1 in vitro. To investigate whether TRX1 can interact with STOP1 in planta, we performed a split-LUC assay in the leaves of Nicotiana benthamiana plants. We detected a LUC signal only for the combination of cLUC-TRX1 and STOP1-nLUC (Fig. 7B), reflecting the in planta interaction between TRX1 and STOP1. To investigate whether TRX1 and STOP1 formed a complex in Arabidopsis, we co-transfected Arabidopsis protoplasts with TRX1-HA or GFP-HA as control with STOP1-FLAG and performed a Co-IP. TRX1-HA, but not GFP-HA, was co-immunoprecipitated with STOP1-FLAG (Fig. 7C). To ascertain whether TRX1 indeed catalyzes the reduction of STOP1, we incubated recombinant STOP1 with H₂O₂ and/or TRX1, and then detected oxidized STOP1. We determined that TRX1 effectively decreases the H₂O₂-induced oxidation of STOP1 (Fig. 7D). Together, these findings show that TRX1 can interact directly with STOP1 to catalyze its reduction.

Figure 7.

The thioredoxin TRX1 catalyzes STOP1 reduction. A) TRX1 directly interacts with STOP1 in vitro. Glutathione agarose beads containing GST or GST-TRX1 were incubated with STOP1-His. Bound proteins were detected by immunoblot analysis using anti-His and anti-GST antibodies. B) Split-LUC assay showing the interaction of STOP1 with TRX1. The indicated pairs of constructs were co-infiltrated in N. benthamiana leaves, and then luciferase activity was determined. C) Co-immunoprecipitation of TRX1 with STOP1. STOP1-FLAG was co-transfected with TRX1-HA or GFP-HA in Arabidopsis protoplasts. Crude protein extracts were immunoprecipitated with anti-FLAG magnetic beads and then probed with anti-HA antibody. D) TRX1 decreases the H₂O₂-induced oxidation of STOP1. Recombinant STOP1-His, pretreated with 1 mM H₂O₂, was incubated alone or with a TRX system comprising 0.5 mM NADPH, 1 μM TRX reductase, and 5  μg TRX1 for 3 h. Subsequently, oxidized STOP1 was detected using the biotin-switch assay. E) Mutation of TRX1 decreases STOP1 accumulation. WT and trx1 seedlings carrying the proSTOP1:STOP1-HA transgene were grown on gel medium soaked with no Al or 0.75 mM Al for 7 d; the roots were then collected for immunoblot analysis. F) Overexpression of TRX1 rescues the decreased expression of STOP1-regulated genes in rae6. Seedlings of WT, TRX1 overexpression line (TRX1-OE), rae6, and rae6 TRX1-OE were cultivated on gel medium soaked with no Al or 0.75 Al for 7 d; the roots were then collected for RT-qPCR analysis of TRX1, ALMT1, MATE, and ALS3. G) Overexpression of TRX1 rescues the Al-sensitive phenotype of rae6. Seedlings were grown on gel medium soaked with no Al or 0.75, 1, or 1.25 mM Al for 7 d. Scale bars, 1 cm. H) Mutation and overexpression of TRX1 increases and decreases the level of STOP1 oxidation, respectively. WT, trx1, and TRX1-OE seedlings carrying the proSTOP1:STOP1-HA transgene were grown on gel medium soaked with no Al or 0.75 mM Al for 7 d; the biotin-switch assay was used to determine the oxidation state of STOP1-HA. Data shown in E) and F) are means ± Sd of 3 biological replicates; the data shown in G) represent the means ± Sd (n = 20 to 35). Asterisks indicate statistically different values (Student's t test, **P < 0.01). n.s., not significantly different. Different lowercase letters indicate significantly different means (P < 0.05, ANOVA followed by Tukey test).

To investigate whether TRX1 can regulate STOP1 function, we used a previously reported knockout line for TRX1 (Nakano et al. 2020). Expression analysis revealed that knockout of TRX1 has no effect on STOP1 expression (Supplemental Fig. S12A), but decreased the expression of STOP1-regulated genes under Al stress conditions (Supplemental Fig. S12A). We then introduced the proSTOP1:STOP1-HA transgene by crossing into the trx1 background to investigate the effect of the trx1 mutation on STOP1 accumulation. We observed that STOP1 abundance decreases in trx1 compared to WT under Al stress conditions, but not under control conditions (Fig. 7E). Accordingly, trx1 was more sensitive to Al than WT (Supplemental Fig. S12B), which is consistent with the previous report (Nakano et al. 2020). The lower expression of STOP1-regulated genes and Al-sensitive phenotype in trx1 could be rescued in 2 trx1 complementation lines (Supplemental Fig. S12, A and B). Furthermore, introduction of the STOP1^3CS transgene into the trx1 mutant background by crossing also rescued the lower STOP1 abundance and Al-sensitive phenotype of trx1 back to WT level (Supplemental Fig. S12, C and D). In addition, overexpression of TRX1 completely rescued the decreased expression of STOP1-regulated genes and the Al-sensitive phenotype of rae6 (Fig. 7, F and G). The biotin-switch assay detected a higher level of oxidized STOP1 in the trx1 mutant and lower in the TRX1 overexpression line than in the WT under Al stress conditions (Fig. 7H). Collectively, these findings suggest that TRX1 catalyzes STOP1 reduction to positively regulate STOP1 accumulation and Al resistance.

Discussion

H₂O₂ can be converted to the more reactive species OH•, which is toxic to cells, and it can also react with certain Cys residues of a protein to modulate its function (Garcia-Santamarina et al. 2014). Studies have shown that H₂O₂ is involved in the regulation of plant stress responses via oxidative post-translational modifications (oxi-PTMs) (Mittler et al. 2022). Nevertheless, it remains unknown whether H₂O₂ actively regulates Al resistance through oxi-PTMs. Our study shows that Al stress and/or the rae6 mutant can induce H₂O₂ accumulation in roots, leading to the oxidation of STOP1 on its C8, C27, and C185 residues. This oxidation of STOP1 enhances its interaction with the F-box protein RAE1, thereby promoting STOP1 degradation, which in turn leads to decreased expression of STOP1-regulated genes and diminished Al resistance. To counteract the H₂O₂-mediated oxidation of STOP1, we also demonstrated that the thioredoxin TRX1 can interact with STOP1 to catalyze its reduction (Fig. 8).

Figure 8.

Model for regulation of STOP1 oxidation and stability. The PPR protein RAE6, localized in the mitochondrion, plays a crucial role in regulating the splicing of nad5, which encodes a component of mitochondrial complex I. Mutation of RAE6 and/or exposure to Al stress result in the accumulation of mitochondrial H₂O₂. This H₂O₂ accumulation leads to the oxidation of STOP1 at the C8, C27, and C185 sites. Consequently, the interaction of STOP1 with the F-box protein RAE1 is enhanced, facilitating STOP1 degradation, which in turn leads to decreased expression of ALMT1 and MATE, and ultimately causes reduced Al resistance. Additionally, the thioredoxin TRX1 interacts with STOP1 to catalyze its reduction.

We identified RAE6 as an uncharacterized, mitochondrion-localized P-type PPR protein using a forward genetic screen. P-type PPR proteins typically participate in splicing, cleavage, and translation of organellar transcripts, while PLS-type PPR proteins are generally involved in RNA editing (Barkan and Small 2014). Our findings support this notion by demonstrating that RAE6 participates in the splicing of mitochondrial nad5 transcripts, which encode a subunit of mitochondrial complex I (Fig. 2). Maturation of the nad5 transcript requires 2 cis- and 2 trans-splicing events to remove the 4 introns (Bonen 2008). While previous studies showed that 2 P-type PPR proteins, ORGANELLE TRANSCRIPT PROCESSING 439 (OTP439) and TANG2, are involved in removing nad5 introns 2 and 3, respectively (Colas des Francs-Small et al. 2014), we showed here that RAE6 mediates splicing of all 4 introns of nad5 (Fig. 2D). Due to defective splicing of nad5, the rae6 mutant exhibited decreased mitochondrial complex I activity, resulting in defective mitochondrial electron transport chain (Fig. 2, E and F) and increased H₂O₂ accumulation (Fig. 3). Unlike the rae6 mutant, which exhibited delayed flowering, shorter plant stature, and shorter siliques under normal growth conditions (Supplemental Fig. S1, B to D), knockout of RAE6 results in seed abortion (Supplemental Fig. S4B). These findings suggest that RAE6 plays an essential role in plant development and that rae6 is a weak mutant allele.

Several studies have revealed that Al stress can induce H₂O₂ accumulation in root mitochondria in pea and peanuts (Arachis hypogaea) (Yamamoto et al. 2002; Huang et al. 2014; Zhan et al. 2014). In the present study, we also show that Al can induce H₂O₂ production in Arabidopsis root mitochondria (Fig. 3, B and E). Our results indicate that Al-triggered H₂O₂ accumulation in mitochondria is not caused by modulation of RAE6 expression or function, as Al stress did not affect RAE6 expression (Supplemental Fig. S5B), and mitochondrial H₂O₂ production rate was similarly induced by Al in WT and rae6 (Fig. 3E). Interestingly, we found that H₂O₂ induced by Al stress and in the rae6 mutant is involved in the modulation of nucleus-localized STOP1 stability and Al resistance, suggesting that H₂O₂ plays a crucial role in Al signaling and plant response to Al stress. Since H₂O₂ has a relatively long half-life in living cells and can cross cell membranes and migrate to different compartments to exert its signaling functions (Bienert et al. 2006), we hypothesized that H₂O₂ directly mediates the oxidation of STOP1 to regulate its function. Our results support this hypothesis, as we determined that H₂O₂ can directly catalyze the oxidation of STOP1 at residues C8, C27, and C185 and promote the degradation of STOP1 (Supplemental Fig. S9; Fig. 5). Moreover, we showed that the total level of H₂O₂, rather than the source of H₂O₂, determines STOP1 oxidation, as both overexpression of CAT2 encoding the peroxisomal H₂O₂-scavenging enzyme CATALASE2 and mutation of ROBHD encoding a plasma membrane-localized H₂O₂-producing oxidoreductase can rescue the lower accumulation of STOP1 and diminished Al resistance of rae6 (Fig. 4; Supplemental Fig. S8).

As a crucial intracellular signaling molecule, H₂O₂ can mediate oxidation of redox-sensitive transcription factors, transmitting a ROS signal into the transcriptional regulatory network (Garcia-Santamarina et al. 2014; Schmidt and Schippers 2015). The oxidation of various transcription factors can have different functional effects. For instance, while H₂O₂-induced oxidation of NONEXPRESSER OF PR GENES 1 (NPR1), an essential regulator of plant systemic acquired resistance, inhibits its localization to the nucleus (Mou et al. 2003), oxidation of the basic leucine zipper (bZIP) transcription factor bZIP16 or the cold-responsive C-REPEAT-BINDING TRANSCRIPTION FACTORs (CBFs) inhibits their DNA binding activities (Shaikhali et al. 2012; Lee et al. 2021). Moreover, Tian et al. (2018) reported that H₂O₂-triggered oxidation of the transcription factor BRASSINAZOLE-RESISTANT 1 (BZR1), a master regulator of brassinosteroid signaling, enhances its transcriptional activity by promoting its interaction with other transcription factors, including AUXIN RESPONSE FACTOR 6 (ARF6) and PIF4. By contrast, our findings indicate that STOP1 oxidation induced by H₂O₂ enhances its interaction with the F-box protein RAE1 and consequently promotes its degradation via the 26S proteasome pathway (Fig. 6).

The H₂O₂-mediated oxidation of thiol to disulfide in cysteine residues is reversible, primarily catalyzed by thioredoxins (TRXs) (Mittler et al. 2022). TRXs are thiol-disulfide oxidoreductases that reduce their target proteins by formation of an intermolecular mixed disulfide bond between the N-terminal cysteine of the CXXC motif of TRXs and their target proteins (Garcia-Santamarina et al. 2014; Mittler et al. 2022). Arabidopsis has various types of TRXs, among which only h-type TRXs are documented to mediate the reduction of transcription factors. For instance, TRX-h5 catalyzes the switch for BZR1 from its oxidized to its reduced form (Tian et al. 2018), while TRX-h2 reduces oxidized CBF oligomers to monomers (Lee et al. 2021). In this study, we demonstrated that the h-type TRX TRX1 directly interacts with STOP1 to catalyze its reduction (Fig. 7, A to D). Mutation of TRX1 increases STOP1 oxidation induced by Al stress and subsequently decreases STOP1 accumulation (Fig. 7, E and H), while overexpression of TRX1 counters rae6-induced STOP1 oxidation and degradation (Fig. 7, F and H). In a previous genome-wide association study, TRX1 was identified as an important locus explaining Al-tolerance variation in diverse Arabidopsis accessions (Nakano et al. 2020), suggesting that the selection of different TRX1 alleles with different expression levels might help regulate the oxidation and accumulation levels of STOP1.

The degradation of STOP1 induced by H₂O₂ appears to be detrimental to plant resistance against Al stress. However, Al stress can activate the MEKK1–MKK1/2–MPK4 cascade, which phosphorylates and stabilizes STOP1 (Zhou et al. 2023). This phosphorylation may counteract the oxidation of STOP1, ultimately resulting in increased STOP1 accumulation in response to Al stress. While an increase in STOP1 levels can enhance Al resistance, it can also inhibit plant growth (Fang et al. 2021b). H₂O₂-induced degradation of STOP1 may allow plants to fine-tune the regulation of STOP1 levels, striking a balance between Al resistance and plant growth.

Materials and methods

Plant materials and growth conditions

All genotypes used in this study were in the Arabidopsis (A. thaliana) accession Columbia-0 (Col-0) background. The previously identified and described genotypes include hpr1 (rae3), rae1 (rae1-1), proSTOP1:STOP1-GUS, stop1 proSTOP1:STOP1^{C8S,C27S,C185S}-3HA, and trx1 (Zhang et al. 2019; Guo et al. 2020; Nakano et al. 2020). The rbohD mutant (SALK_005253.55.50.x) was obtained from the Arabidopsis Biological Resource Center (ABRC).

Arabidopsis seeds were disinfected with 6% NaClO for 8 min and grown on solid half-strength Murashige and Skoog (MS) medium containing 1.2% (w/v) agar and 1% (w/v) sucrose for 7 d at 22 °C in a growth chamber (CU36L4, Percival). The seedlings were then examined for luciferase luminescence using a CCD imaging apparatus (Lumazone P1300B, Roper Scientific) or transferred to soil and grown at 22 °C under a photoperiod of 14 h of light (100 μmol/m²/s, Philips TL-D 26W/865 cool daylight tubes) and 10 h of darkness. For split-LUC assays, N. benthamiana seeds were sown on soil and grown under the same conditions as Arabidopsis.

Measurement of malate secretion and Al content

To measure malate secretion, WT and rae6 seeds were cultivated on 1.2% (w/v) agar medium containing half-strength MS medium and 1% (w/v) sucrose for 5 d. Seedlings were pretreated with a 2% (w/w) nutrient solution (Fujiwara et al. 1992), composed of 1.75 mM sodium phosphate buffer (pH 5.8), 1.5 mM MgSO₄, 2.0 mM Ca(NO₃)₂, 3.0 mM KNO₃, 67 μM Na₂EDTA, 8.6 μM FeSO₄, 10.3 μM MnSO₄, 30 μM H₃BO₃, 1.0 μM ZnSO₄, 24 nM (NH₄)₆Mo₇O₂₄, 130 nM CoCl₂, 1 μM CuS0₄, and 1% (w/v) sucrose, for 2 h at pH 4.8. Subsequently, the seedlings were treated with the same solution containing either no AlCl₃ or 10 μM AlCl₃ at pH 4.8 for 12 h in a 12-well plate with shaking. The root exudates were collected and concentrated using a freeze-dryer (Alpha 1-2 LDplus; Martin Christ Gefriertrocknungsanlagen GmbH, Osterode am Harz, Germany). The malate concentration in the root exudates was measured using the NAD⁺/NADH enzymatic cycling method (Hampp et al. 1984).

To determine Al content in roots, 3-wk-old wild-type, rae6, and hpr1 plants, grown in 1/10 Hoagland hydroponic solution, were pretreated with a 0.5 mM CaCl₂ solution for 6 h. Subsequently, the plants were treated with the same CaCl₂ solution alone or with 20 μM AlCl₃ for 12 h. Al content in the roots was determined using a recent method (Ligaba-Osena et al. 2017). Briefly, plant roots were immersed in a 0.5 mM citrate solution (pH 4.2) at 4 °C for 30 min to desorb Al bound in the apoplast and/or precipitated on the root surface. Afterward, the roots were rinsed 3 times with 18 MΩ water, gently blotted on paper towels, and carefully separated before transferred to preweighed test tubes. The tissues were then dried at 65 °C for 2 d, and their weights were measured. Subsequently, the tissues were digested with 2 mL of concentrated HNO₃ and diluted to a final volume of 10 mL with 2% HNO₃. The concentration of Al in the diluted solution was measured using inductively coupled plasma mass spectrometry (NexION300D; PerkinElmer, Waltham, Massachusetts, United States of America).

Evaluation of Al resistance

The method used to evaluate Al resistance was a modified version of the soaked gel medium method previously described (Larsen et al. 2005). The gel medium was prepared by combining the following components: 50 mL (pH 5.0) of 0.25 mM (NH₄)₂SO₄, 1 mM KNO₃, 0.2 mM KH₂PO₄, 2 mM MgSO₄, 1 mM Ca(NO₃)₂, 1 mM CaSO₄, 1 mM K₂SO₄, 1 mM MnSO₄, 5 mM H₃BO₃, 0.05 mM CuSO₄, 0.2 mM ZnSO₄, 0.02 mM NaMoO₄, 0.1 mM CaCl₂, 0.001 mM CoCl₂, 1% (w/v) sucrose, and 0.4% (w/v) Gellan gum (G1910; Sigma-Aldrich). The medium was allowed to solidify, after which 35 mL (pH 3.6) of the same nutrient solution (without Gellan gum) containing no AlCl₃ or 0.75, 1.0, or 1.25 mM AlCl₃ was added to soak the medium. After 2 d of soaking, the remaining solution was removed, and seedlings were grown on the gel medium for 7 d. Subsequently, the seedlings were imaged, and root lengths were measured using ImageJ software. Relative root length, calculated as a percentage of root length with Al treatment relative to root length without Al (×100), was used to assess Al resistance.

Cloning of RAE6

To perform genetic analysis and mapping-by-sequencing, the rae6 mutant was crossed to its wild type to construct an F₂ population. Eighty-nine F₂ plants exhibiting lower LUC signal from the proALMT1:LUC reporter were selected, and their genomic DNA was pooled for high-throughput DNA sequencing using an Illumina HiSeq4000 system, which was performed by a commercial company (Shanghai Hanyu Biotech Lab, Shanghai, China) (NCBI accession no. SRR9613722). The MutMap method (Abe et al. 2012) was employed to identify the candidate region for rae6. To validate the candidate region, a linkage analysis was conducted on the same 89 F₂ plants using 6 dCAPS markers. These markers were developed based on the single nucleotide polymorphisms between the wild type and rae6 within the candidate region. Among the dCAPS markers employed, marker A2, which is based on a C-to-T substitution in At1g73710, exhibited complete linkage to the mutant phenotype (Supplemental Table S1). This result suggests that At1g73710 is RAE6.

To confirm the association between At1g73710 and RAE6, an adenine base editor (A-to-G) (Wei et al. 2021) was used to convert the T-to-C mutation in At1g73710 in rae6 back to the wild-type sequence of At1g73710 by designing a 20 bp single guide RNA (sgRNA) targeting the mutation site (Supplemental Data Set 1). Among approximately 120 transgenic plants screened, 1 specific plant (rae6^F661L) was identified that carried the desired C-to-T mutation in At1g73710 within the rae6 background. This plant or its progeny was used for expression analysis and phenotypic analysis related to aluminum resistance. Furthermore, to generate knockdown lines for At1g73710 using RNAi, a portion of the At1g73710 coding sequence (201 bp) was amplified with a specific primer pair (Supplemental Data Set 1) and cloned into the pANDA vector by GATEWAY (11791020, Invitrogen) cloning. Two transgenic lines with lower transcript levels of At1g73710 were obtained and subjected to the analysis and phenotypic analysis related to aluminum resistance.

RT-PCR analysis

To analyze the expression of Al resistance genes, seedlings were cultivated on gel medium soaked with no AlCl₃ or 0.75 mM AlCl₃ for 7 d. The roots of approximately 30 seedlings were excised for RNA isolation in each of 3 separate sets of seedlings. Total RNA was extracted from each sample using TRIzol reagent (Beyotime), and DNase I treatment (Vazyme Biotech, Nanjing, China) was performed to eliminate DNA contamination. Approximately 1 μg of total RNA was used for first-strand cDNA synthesis using a HiScript 1st Strand cDNA Synthesis Kit (Vazyme Biotech, Nanjing, China). qPCR analysis was conducted using 1/10 of the cDNA products in combination with 2×Universal SYBR Green Fast qPCR Mix (ABclonal, Wuhan, China) on a CFX96 Touch real-time PCR detection system (Bio-Rad). The reference gene was UBQ10.

To investigate the expression of mitochondrial genes, the roots of 9-d-old seedlings grown on half-strength MS agar medium were excised for RNA extraction. First-strand cDNA synthesis was performed using random hexamer primers. qPCR and end-point RT-PCR methods were employed to assess the expression of mitochondrial genes without or with introns, respectively. For the analysis of splicing efficiency in the 4 introns of nad5, specific primers were designed to target intron–exon junctions (unspliced forms) or exon–exon junctions (spliced forms). The primers used for the expression analysis can be found in Supplemental Data Set 1.

Mitochondrion isolation, BN-PAGE, and mitochondrial complex I activity assay

Mitochondrion isolation was conducted following the protocol outlined by Kelly and Wiskich (1988). In brief, approximately 10 g fresh roots of 12-d-old seedlings were ground and homogenized at 4 °C using a Mitochondrial Isolation Kit (P0045; Biohao, Wuhan, China). Following homogenization, the sample was centrifuged at 1,000 × g for 10 min at 4 °C to eliminate the nucleus and cell debris. Subsequently, the mitochondria were precipitated and collected by centrifuging the supernatant at 16,000 × g for 10 min at 4 °C. For BN-PAGE, the procedure was conducted following a previously described method (Colas des Francs-Small et al. 2014). In brief, 75 μg of mitochondrial protein was treated with 1% (w/v) dodecylmaltoside, and the native protein complexes were separated on a 5% to 13.5% (w/v) acrylamide gradient gel. The activity of complex I was assessed using a CheKine Micro Mitochondrial complex I Activity Assay Kit (KTB1850; Abbkine, California, United States of America). This kit detects the activity of the mitochondrial respiratory chain complex by measuring the oxidation rate of NADH at 340 nm. Complex I catalyzes NADH dehydrogenation to generate NAD⁺, and the activity of the enzyme can be calculated directly by reading the oxidation rate of NADH at 340 nm.

Transient expression in protoplasts

Arabidopsis protoplasts were isolated from 14-d-old seedlings using a previously described method (Zhai et al. 2009). To investigate protein–protein interactions between STOP1 and TRX1, as well as between RAE1 lacking the F-box domain (RAE1ΔF) and intact of mutated STOP1, 2 mL of wild-type protoplasts (∼2 × 10⁶ cells/mL) was co-transfected with 50 μg of 35S:TRX1-2HA and either 50 μg of 35S:STOP1-2FLAG or 35S:GFP-2FLAG for the first experiment. For the second experiment, 2 mL of rae1 rah1 protoplasts was co-transformed with 50 μg of 35S:RAE1ΔF-3HA and either 50 μg of 35S:STOP1-2FLAG or 35S:STOP1^{C8S,C27S,C185S} (STOP1^3CS)-2FLAG. Total proteins were extracted using 200 μL of protein extraction buffer containing 20 mM Tris-HCl (pH 7.5), 300 mM NaCl, 5 mM MgCl₂, 50 μM MG132 (A2585; APExBIO, United States of America), 0.5% (v/v) NP-40, and 1×Complete Protease inhibitor tablets EDTA-free (5892791001, Roche). An aliquot of 20 μL of the protein extract was saved as an input control. The remaining cell extracts were diluted to 1 mL alone or with 1 mM H₂O₂ and then incubated with 20 μL of anti-FLAG or anti-HA M2 magnetic beads for 4 h at 4 °C with gentle rotation. The protein-bound beads were washed 4 times with 1 mL of protein extraction buffer, and the bound proteins were eluted with 1×SDS loading buffer for immunoblot analysis using anti-FLAG-HRP, 1:5,000 (A8592; Sigma-Aldrich) or anti-HA-HRP, 1:2,000 (12013819001, lot 44323100, Roche) antibodies.

Subcellular localization analysis

To investigate the subcellular localization of RAE6, a 1.7 kb promoter fragment of UBQ10, the full-length RAE6 coding sequence and GFP were separately amplified and then fused together using overlapping PCR (proUBQ10:RAE6-GFP). The PCR products of proUBQ10:RAE6-GFP were transfected into Arabidopsis mesophyll protoplasts, followed by cultivation in dark conditions for 16 h at 22 °C. Subsequently, the protoplasts were stained with 100 nM MitoTracker Red CMXRos for 20 min at 37 °C and observed using a confocal laser scanning microscope (ZEISS LSM880) to visualize the green and red fluorescence signals.

Pull-down assay

The coding sequences of TRX1 or RAE1ΔF and STOP1 or STOP1^{C8S,C27S,C185S} (STOP1^3CS) were individually cloned into the pGEX4T-1 and pET29a (+) vectors, respectively, resulting in the generation of GST-TRX1, GST-RAE1ΔF, STOP1-His, and STOP1^3CS-His recombinant plasmids. Each plasmid was then transformed into Escherichia coli strain BL21 (DE3) for production of the target proteins. The BL21 (DE3) cells were initially cultured at 37 °C to the OD600 at 0.6 and subsequently incubated at 16 °C for 16 h in the presence of 0.1 mM IPTG to induce protein production. For the TRX1 and STOP1 protein interaction assay, glutathione beads containing 20 μg of either GST or GST-TRX1 were incubated with 20 μg of STOP1-His at 4 °C for 4 h. In the protein interaction assay involving RAE1ΔF and STOP1 or STOP1^{C8S,C27S,C185S}, the glutathione beads containing 20 μg of GST-RAE1ΔF and either no or 1 mM H₂O₂ were incubated with 20 μg of STOP1-His or 20 μg of STOP1^3CS-His at 4 °C for 4 h. Following elution from the beads, the protein complexes were detected using immunoblotting with anti-GST or anti-His antibodies.

Split luciferase complementation (split-LUC) assays

The coding sequences of STOP1 and TRX1 were individually cloned into the pCAMBIA1-nLUC and pCAMBIA1-cLUC vectors, respectively. Subsequently, STOP1-nLUC was co-infiltrated with cLUC-TRX1 into fully expanded N. benthamiana leaves via Agrobacterium-mediated infiltration. The infiltrated plants were then cultivated under long-day conditions for 2 d, and LUC signals from the infiltrated leaves were examined.

Detection of ROS burst

Seven-day-old seedlings cultivated on half-strength MS agar medium were transferred to gel medium soaked with a solution containing no or 0.75 mM AlCl₃ for 12 h. For ROS detection, the roots of the seedlings were stained with 10 μM H₂DCFDA and 50 nM MitoTracker Red for 30 min and were subjected to fluorescence detection after washing twice with 18 MΩ water. For H₂O₂ detection, the roots were incubated in a freshly prepared staining solution composed of 1 mg/mL DAB and 0.1% (v/v) Tween 20 in 10 mM Na₂HPO₄ in the dark for 3 h. For O₂⁻ detection, 0.5 mg/mL NBT solution was used to stain the roots in the dark for 1 h. The stained roots were observed and photographed using a stereomicroscope (SZX12, Olympus) equipped with a camera (DP20, Olympus).

To determine the rate of mitochondrial H₂O₂ production, a previously described method for isolating root mitochondria was followed (Considine et al. 2003). Briefly, approximately 5 g roots of 10-d-old seedlings treated with no or 0.75 mM AlCl₃-soaked gel medium for 12 h were homogenized into 10 mL of extraction medium, consisting of 0.3 M mannitol, 50 mM TES-NaOH (pH 7.5), 0.5% (w/v) BSA, 0.5% (w/v) polyvinylpyrrolidone-40, 2 mM EGTA, and 20 mM cysteine. The resulting homogenate was centrifuged at 1,500 × g for 10 min at 4 °C. The supernatant was then subjected to another centrifugation at 18,000 × g for 10 min to collect the organelle pellet. This pellet was resuspended in wash buffer composed of 0.3 M mannitol and 20 mM TES-KOH (pH 7.5) and layered onto a stepped Percoll gradient. The Percoll gradient consisted of steps of 50, 28, and 20% (w/v) Percoll with 0.3 M mannitol as an osmoticum. After centrifugation at 43,000 × g for 30 min at 4 °C, mitochondria were recovered from the 28% to 50% Percoll interface. Further purification of mitochondria was performed using a second self-forming Percoll gradient consisting of 28% Percoll with 0.3 M sucrose as an osmoticum. Measurement of mitochondrial H₂O₂ was conducted using the peroxidase-dependent conversion of Amplex Red (ST010, Beyotime, China) to resorufin, following the method described by Smith et al. (2004). The assay mixture, with a final volume of 200 μL, contained 50 μM Amplex Red, 10 units superoxide dismutase, 1.2 units of horseradish peroxidase, 0.3 M mannitol, 10 mM TES-KOH (pH 7.5), 3 mM MgSO₄, 10 mM NaCl, 5 mM KH₂PO₄, 0.1% (w/v) BSA, 20 mM glucose, 0.3 mM NAD⁺, 0.1 mM ADP, 0.1 mM thiamin pyrophosphate, 0.15 units/mL of hexokinase, and purified mitochondria (100 μg protein equivalent). The background rate of H₂O₂ production was determined by measuring the accumulation rate of resorufin using fluorescence spectrophotometry (excitation at 570 nm, emission at 587 nm). Respiratory substrates including 10 mM pyruvate, 2.1 mM citrate, 1.3 mM succinate, 0.6 mM malate, 0.02 mM fumarate, and 0.02 mM isocitrate were then added, and the rate of H₂O₂ production was measured over a time period of 10 to 15 min after the addition of the substrates.

Determination of protein levels in roots

To assess STOP1 accumulation in different lines, seedlings were grown on gel medium soaked with no AlCl₃ or 0.75 mM AlCl₃ for 7 d. Subsequently, the roots were excised for protein extraction. To examine the effect of the rae6 mutation on STOP1 stability, 9-d-old seedlings were pretreated with a 0.5-mM CaCl₂ solution at pH 4.8 for 2 h and then exposed to the same CaCl₂ solution containing no Al or 30 μM AlCl₃ for 4 h. The seedlings subjected to Al treatment were further exposed to the same CaCl₂ solution containing 30 μM Al and 100 μM cycloheximide (A8244; ApexBio) for varying time intervals. At each time point, roots were harvested for protein extraction. To investigate the influence of MG132 on STOP1 accumulation, 9-d-old seedlings were exposed to either no Al or 30 μM Al for 3 h and then 50 μM MG132 with no Al or 30 μM Al for 3 h. Subsequently, the roots were collected for protein extraction. To explore the effect of H₂O₂ on STOP1 accumulation, 9-d-old seedlings were exposed to either no or 1 mM H₂O₂ for 3 h, after which the roots were excised for protein extraction. Total proteins were extracted using the protein extraction buffer mentioned above, and were then separated by 8% (w/v) SDS-PAGE. The resolved STOP1-3HA proteins were detected through standard immunoblot assays using HA-HRP antibody. Actin served as a loading control and was detected using an anti-Actin antibody (CW0264M; CoWin Biosciences).

GUS analysis

To determine the STOP1-GUS protein accumulation, 10-d-old seedlings from the wild type and rae6 carrying the proSTOP1:STOP1-GUS transgene were pretreated with a 0.5 mM CaCl₂ solution at pH 4.8 for 2 h and then exposed to the same CaCl₂ solution containing either no Al or 30 μM Al for 8 h. Subsequently, the roots were collected and ground to a fine powder, which was then mixed with GUS extraction buffer comprising 50 mM sodium phosphate, pH 7.0, 10 mM EDTA, 0.1% (w/v) SDS, 0.1% (v/v) Triton X-100, 10 mM 2-mercaptoethanol, and 25 mg/mL PMSF. The mixed solution was centrifuged at 13,523 × g for 10 min at 4 °C, and subsequently a portion of the supernatant was mixed with GUS extraction buffer containing 1 mM 4-methylumbelliferyl β-D-glucuronide (A602251; Sangon Biotech) at 37 °C. After 1 h, a solution of 0.2 M Na₂CO₃ was added to halt the reaction. The fluorescence of the fluorochrome 4-methylumbelliferone in the reaction was measured, with excitation and emission wavelengths of 350 and 455 nm, respectively. GUS activity in each sample was normalized to total protein content.

To generate proRAE6:GUS transgenic lines, a 1.5 kb promoter fragment of RAE6 was amplified and cloned upstream of the GUS reporter gene in the pORE-R2 vector. Different tissues from 1 proRAE6:GUS transgenic line were stained with a commercialized GUS staining solution (S102; O’Biolab, Beijing, China). Stained tissues were observed and photographed using a stereomicroscope (SZX12, Olympus) equipped with a camera (DP20, Olympus).

Determination of STOP1 oxidation

The oxidation modification assay of Arabidopsis proteins was performed as previously described (Tian et al. 2018). For H₂O₂ treatment, seedlings grown on half-strength MS agar medium for 10 d remained on the same medium or were treated with 1 mM H₂O₂ for 3 h. For Al treatment, the seedlings were grown on gel medium soaked with no AlCl₃ or 0.75 mM AlCl₃ for 7 d. The roots of the seedlings were harvested and ground to a fine powder in liquid nitrogen. Total proteins were extracted using extraction buffer, consisting of 20 mM HEPES (pH 8.0), 40 mM KCl, 5 mM EDTA, 0.5% (v/v) Triton X-100, 1% (w/v) SDS, 1 mM PMSF, and 1× protease inhibitor cocktail. The protein extracts were incubated with 100 mM NEM in EBR buffer at room temperature for 30 min with frequent vortexing to block free thiols. The samples were then precipitated with 1 volume of acetone and washed 3 times with 50% (v/v) acetone. The pellets were dissolved in 500 μL EBR buffer containing 20 mM DTT and were incubated at 37 °C for 30 min to reduce the oxidized thiols. DTT was removed by protein precipitation with 1 volume of acetone, and the pellets were resuspended in 500 μL EBR buffer. The supernatant was labeled with 100 μM BIAM at room temperature for 1 h in the dark, and proteins were precipitated with 1 volume of acetone to remove free BIAM. The BIAM-treated proteins were then dissolved in 250 μL extraction buffer and diluted with 750 μL incubation buffer, composed of 20 mM HEPES (pH 8.0), 40 mM KCl, 5 mM EDTA, 0.25% (v/v) Triton X-100, 1 mM PMSF, and 1 × protease inhibitor cocktail. After centrifugation at 12,000 × g for 5 min at 4 °C, the supernatant was added to 40 μL of streptavidin beads and incubated at 4 °C overnight. The beads were washed 5 times with the incubation buffer, and the proteins were eluted with 40 μL of 2 × SDS sample buffer. The samples were then separated on 8% (w/v) SDS-PAGE gels. An aliquot of proteins before incubation with streptavidin beads was used as total STOP1 protein control. The anti-HA-HRP antibody was used to detect oxidized or total STOP1-HA protein.

For in vitro reduction analysis of STOP1 by TRX1, recombinant STOP1-His and GST-TRX1 were purified from E. coli. Subsequently, STOP1-His, pretreated with H₂O₂, was incubated at room temperature for 3 h alone or with a TRX system. The TRX system comprised 0.5 mM NADPH, 1 μM TRX reductase (T7915; Sigma-Aldrich), and 5 μg GST-TRX1. To precipitate the proteins, 1 volume of acetone was added, and the mixture was kept at −20 °C for 20 min, followed by centrifugation at 5,000 × g for 5 min at 4 °C. The resulting pellet was washed 3 times with 50% (v/v) acetone and then dissolved in 500 μL EBR buffer. The in vitro oxidation of STOP1 was detected using the same method as described for the aforementioned in vivo oxidation detection.

LC–MS/MS analysis

Recombinant STOP1 was treated with 1 mM H₂O₂ or DTT for 30 min at room temperature, and was then incubated with 100 mM NEM in EBR buffer at room temperature for 30 min with frequent vortexing to block free thiols. The sample was then precipitated by ice-cold acetone and labeled with 100 μM BIAM at room temperature for 1 h in the dark. The labeled protein was in-gel digested with trypsin, 1:30 (V5280, Promega). The resulting sample was redissolved in a 0.1% (v/v) formic acid solution and adsorbed onto an Easy column with an injection needle (SenSil C18-AQ, 75 μm × 25 cm, 1.6 μm) from Fresh Bioscience. Separation was carried out using an EASY-nLC 1200 liquid phase system (Thermo Scientific). Mobile phase A consisted of a 0.1% (v/v) formic acid aqueous solution, while mobile phase B consisted of an 80% (v/v) acetonitrile with 0.1% (v/v) formic acid. Gradient elution was performed at a flow rate of 250 nL/min for 75 min. The samples were separated by chromatography and then analyzed by mass spectrometry using a Q EXACTIVE HF instrument (Thermo Scientific). The source voltage was set to 2.4 kV, and the parent ions and secondary fragments of the peptides were detected and analyzed using a high-resolution Orbitrap. For primary mass spectrometry, the scanning range was set from 300 to 1,650 m/z with a scanning resolution of 45,000. For secondary mass spectrometry, the scanning range was fixed to 100 m/z, and the scanning resolution of the Orbitrap was set to 30,000. The data acquisition mode employed a data-dependent acquisition program, which selected the 20 parent ions with the highest signal intensity for higher-energy collisional dissociation. A fragmentation energy of 32% was used for fragmentation. Likewise, secondary mass spectrometry was performed successively. To enhance mass spectrum utilization efficiency, the automatic gain control was set to 200,000 the maximum injection time was set to 100 ms, and a dynamic exclusion time of 30 s was implemented during series mass spectrum scanning to prevent repeated scanning of parent ions.

In vitro BIAM-labeling and STOP1 oxidation assays

For the BIAM-labeling assay, recombinant STOP1-His and STOP1^3CS-His proteins were treated with different concentrations of H₂O₂ at room temperature for 15 min. The proteins were precipitated by adding 1 volume of acetone at −20 °C for 20 min and then centrifuged at 5,000 × g for 5 min at 4 °C. The pellets were washed 3 times with 50% (v/v) acetone and dissolved in 500 μL of labeling buffer containing 50 mM MES-NaOH (pH6.5), 100 mM NaCl, 1% (v/v) Triton X-100, and 100 μM BIAM. They were then incubated at room temperature in the dark for 1 h. The labeling reactions were terminated by adding β-mercaptoethanol to a final concentration of 20 mM. The reaction mixtures were precipitated by adding 1 volume of acetone at −20 °C for 20 min and then centrifuged at 5,000 × g for 5 min at 4 °C. The pellets were dissolved in 40 μL of SDS sample buffer, and subjected to separate on SDS-PAGE. Proteins labeled with BIAM were detected using HRP-conjugated streptavidin (SLP001H, Smart-Lifesciences).

For in vitro oxidation analysis, recombinant STOP1-His, STOP1^C8S-His, STOP1^C27S-His, STOP1^C185S-His, and STOP1^3CS-His proteins were purified from E. coli and then treated with 1 mM H₂O₂ or 1 mM DTT at room temperature for 15 min. The proteins were precipitated by adding 1 volume of acetone at −20 °C for 20 min and then centrifuged at 5,000 × g for 5 min at 4 °C. The pellet was washed 3 times with 50% (v/v) acetone and dissolved in 500 μL EBR buffer. The in vitro oxidation state of STOP1 was then detected using the same method as the aforementioned in vivo oxidation detection.

Quantification and statistical analysis

ImageJ software was used to quantify the band intensities from immunoblots or the intensity of fluorescence and chemical staining. Statistical analysis between 2 groups was conducted using a 2-tailed Student's t test; when more than 2 groups were involved, a 1-way analysis of variance (ANOVA) followed by Tukey test was performed using GraphPad Prism 9.00 software. Statistical data are provided in Supplemental Data Set 2.

Accession numbers

Gene sequence data in this article can be found in the Arabidopsis Information Resource (TAIR) database under the following accession numbers: At1g34370 (STOP1), At1g08430 (ALMT1), At1g51340 (MATE), At2g37330 (ALS3), At1g67940 (STAR1), At5g39040 (ALS1), At1g73710 (RAE6), At2g37330 (ALS3), At4g35090 (CAT2), At5g47910 (RBOHD), At5g01720 (RAE1), and At3g51030 (TRX1). Whole-genome sequence data of pooled F₂rae6 mutant plants produced in this study were deposited at the National Center for Biotechnology Information (accession number: SRR17165106).

Acknowledgments

We thank Prof. Ian Small from The University of Western Australia for commenting on RAE6 function.

Author contributions

X.W., Y.Z., and C.-F.H. designed the experiments. X.W., Y.Z., W.X., W.R., Y.Z., and H.Z. performed the experiments. S.D. commented on the research. X.W. and C.-F.H. wrote the manuscript.

Supplemental data

The following materials are available in the online version of this article.

Supplemental Figure S1. Effect of the rae6 mutation on the expression of Al resistance genes and plant development.

Supplemental Figure S2. MutMap analysis of the rae6 mutant.

Supplemental Figure S3. Genetic rescue of the rae6 mutant.

Supplemental Figure S4. Effect of RAE6 knockout or knockdown on plant development, the expression of STOP1-regulated genes, and Al resistance.

Supplemental Figure S5. Expression pattern of RAE6.

Supplemental Figure S6. Effect of the rae6 mutation on the expression of mitochondrial genes.

Supplemental Figure S7. Exogenous supply of GSH partially rescues the Al-sensitive phenotype and expression of STOP1-regulated genes of rae6.

Supplemental Figure S8. Mutation of RBOHD rescues the Al-sensitive phenotype, and decreased STOP1 accumulation and expression of STOP1-regulated genes in rae6.

Supplemental Figure S9. Mass spectrometry analysis of the tryptic fragments of STOP1-His.

Supplemental Figure S10. Triple C-to-S mutations (3CS) of C8, C27, and C185 of STOP1 increase the expression of STOP1-regulated genes and rescue the Al-sensitive phenotype of rae6.

Supplemental Figure S11. Mutation of RAE1 rescues the rae6-induced decreased expression of STOP1-regulated genes.

Supplemental Figure S12. Rescue of trx1 deficiency by wild-type TRX1 and triple C-to-S mutations (3CS) of C8, C27, and C185 in STOP1.

Supplemental Table S1. List of mutations in the rae6 mutant and linkage analysis.

Supplemental Data Set 1. Primers used in this study.

Supplemental Data Set 2. Summary of statistical analyses in all figures.

Funding

This work was supported by the National Natural Science Foundation of China (Grant No. 32170261 to C.-F.H.) and by National Key Laboratory of Plant Molecular Genetics.

Dive Curated Terms

The following phenotypic, genotypic, and functional terms are of significance to the work described in this paper:

• BZR1 Gramene: AT1G75080

• BZR1 Araport: AT1G75080

• PLS Gramene: AT4G39403

• PLS Araport: AT4G39403

• ARF6 Gramene: AT1G30330

• ARF6 Araport: AT1G30330

• PIF4 Gramene: AT2G43010

• PIF4 Araport: AT2G43010

• CAT2 Gramene: AT4G35090

• CAT2 Araport: AT4G35090

• ALS3 Gramene: AT2G37330

• ALS3 Araport: AT2G37330

• UBQ10 Gramene: At4g05320

• UBQ10 Araport: At4g05320

• GDH1 Gramene: AT5G18170

• GDH1 Araport: AT5G18170

• GDH2 Gramene: AT5G07440

• GDH2 Araport: AT5G07440

• STOP1 Gramene: AT1G34370

• STOP1 Araport: AT1G34370

• ALMT1 Gramene: AB081803

• ALMT1 Araport: AB081803

• NPR1 Gramene: At1g64280

• NPR1 Araport: At1g64280

• RBOHD Gramene: At5g47910

• RBOHD Araport: At5g47910

• NADPH CHEBI: CHEBI:16474

• ADP CHEBI: CHEBI:16761

• CHX Gramene: cycloheximide

• CHX Araport: cycloheximide

Author notes

Xiang Wei and Yifang Zhu contributed equally to this work.

The author responsible for distribution of materials integral to the findings presented in this article in accordance with the policy described in the Instructions for Authors (https://academic.oup.com/plcell/pages/General-Instructions) is: Chao-Feng Huang ([email protected]).

Conflict of interest statement. None declared.