N-glycosylation of SnRK2s affects NADPH maintenance in peroxisomes during prolonged ABA signalling

Lu, Junyao; Li, Ning; Li, Gaojian; Tian, Ziang; Shi, Lianping; Wang, Yan; Cai, Yingao; Zhang, Kaiyuan; Sun, Wanting; Wang, Danyang; Lin, Jinxin; Huang, Jinguang; Wu, Changai; Yan, Kang; Zhang, Shizhong; Zheng, Chengchao; Yang, Guodong

doi:10.1038/s41467-024-50720-3

Download PDF

Article
Open access
Published: 05 August 2024

N-glycosylation of SnRK2s affects NADPH maintenance in peroxisomes during prolonged ABA signalling

Nature Communications volume 15, Article number: 6630 (2024) Cite this article

6435 Accesses
12 Citations
Metrics details

Subjects

Abstract

Unfavourable conditions, such as prolonged drought and high salinity, pose a threat to the survival and agricultural yield of plants. The phytohormone ABA plays a key role in the regulation of plant stress adaptation and is often maintained at high levels for extended periods. While much is known about ABA signal perception and activation in the early signalling stage, the molecular mechanism underlying desensitization of ABA signalling remains largely unknown. Here we demonstrate that in the endoplasmic reticulum (ER)-Golgi network, the key regulators of ABA signalling, SnRK2.2/2.3, undergo N-glycosylation, which promotes their redistribution from the nucleus to the peroxisomes in Arabidopsis roots and influences the transcriptional response in the nucleus during prolonged ABA signalling. On the peroxisomal membrane, SnRK2s can interact with glucose-6-phosphate (G6P)/phosphate translocator 1 (GPT1) to maintain NADPH homeostasis through increased activity of the peroxisomal oxidative pentose phosphate pathway (OPPP). The resulting maintenance of NADPH is essential for the modulation of hydrogen peroxide (H₂O₂) accumulation, thereby relieving ABA-induced root growth inhibition. The subcellular dynamics of SnRK2s, mediated by N-glycosylation suggest that ABA responses transition from transcriptional regulation in the nucleus to metabolic processes in the peroxisomes, aiding plants in adapting to long-term environmental stress.

Unveiling the regulatory role of GRP7 in ABA signal-mediated mRNA translation efficiency regulation

Article Open access 26 April 2025

The abscisic acid signaling negative regulator OsPP2C68 confers drought and salinity tolerance to rice

Article Open access 25 February 2025

A dual function of SnRK2 kinases in the regulation of SnRK1 and plant growth

Article 19 October 2020

Introduction

The phytohormone abscisic acid (ABA) controls stress tolerance and development in plants^1,2,3,4. After ABA perception by the soluble receptors pyrabactin resistance 1 (PYR1)/PYR1-like (PYL)/regulatory components of ABA receptors (RCAR)^5,6, sucrose nonfermenting-1 (Snf1)-related kinase 2 (SnRK2) members (mainly SnRK2.2, SnRK2.3 and SnRK2.6) are released from the protein phosphatase type 2 C (PP2C)-SnRK2 inactivation complex^7,8,9. Subsequently, activated SnRK2s phosphorylate and activate downstream effectors, such as transcription factors in the nucleus¹⁰, ion channel and NADPH oxidases (NOXs)/respiratory burst oxidase homologs (RBOHs) on the plasma membrane^11,12,13. The SnRK2 kinases function as the central hub in this well-established early ABA signalling^14,15,16,17. A common feature of signal-response regulatory modules is that the continuous presence of stimulation often results in attenuated responsiveness to subsequent challenges in the long-term. This ability to adapt is referred to as signalling desensitisation^18,19. Pretreatment with salt stress or water-deficit conditions significantly dampens ABA signalling, a phenomenon that has been observed for almost 40 years^20,21. However, the underlying molecular mechanism of desensitisation in long-term ABA signalling remains largely unknown.

Asparagine (Asn/N)-linked glycosylation (N-glycosylation) catalysed by a group of N-glycan processing enzymes (NPEs) is one of the most prominent and abundant co/posttranslational modifications for secretory and membrane proteins within the ER-Golgi network in organisms across all domains^22,23,24. When a nascent glycoprotein is present in the ER lumen, the oligosaccharyltransferase complex (OST) attaches an N-glycan from dolichol-pyrophosphate-linked lipid anchor to a specific N residue in the NX(S/T) motif (the third one may also be cystine occasionally). Subsequently, N-glucosidase I (GI) and GIIα/β heterodimer remove the outmost two glucose residues. GII-trimmed glycoprotein is captured by calnexin/calreticulin (CNX/CRT) for quality control. Successfully folded glycoprotein can be further processed to form a mature N-glycan structure by NPEs sequentially, such as mannose-removing enzymes α-mannosidases (MNSs), β1,2-N-acetylglucosaminyl-transferase I (GnTI), Golgi-α-mannosidase II (GMII), and fucosyltransferases (FUC)^24,25. Based on genetic analyses, several NPEs were reported to be required for plant tolerance to abiotic stresses. Defective alleles of GIIα, display sensitivity to high-temperature stress²⁶. The loss-of-function mutants of NPEs are hypersensitive to high salt stress, such as mutants of the catalytic subunit of OST, STT3A (staurosporine and temperature sensitivity 3 A isoform)²⁷, Complex glycan 1(CGL1/GnTI)^28,29, GMII^28,30, MNS1 and MNS2³¹, β1,2-xylosyltransferase (XYLT)²⁸, and α1,3-Fucosyltransferase 11 and 12 (FUT11/FUT12)^28,32. Here, we unexpectedly found that the nonsecretory proteins SnRK2.2/2.3 are modified by NPEs with an N-glycan addition at its C-terminal domain II, which positively regulates desensitisation of ABA signalling in Arabidopsis.

Results

NPEs promote desensitisation of the transcriptional ABA response

Through screening of Arabidopsis mutants in response to exogenous ABA, an allele embedding a T-DNA insertion in the β-subunit of glucosidase II (gIIβ−2, Supplementary Fig. 1a) exhibited hypersensitivity to ABA-mediated repression of primary root growth and ABA-induced detached leaf senescence (Fig. 1a, b and Supplementary Fig. 2a, b). Consistently, two more loss-of-function alleles, gIIβ−3 and gIIβ−4, and their F1 hybrids all showed hypersensitivity to ABA treatment (Supplementary Fig. 2c, d). A genomic DNA fragment harbouring GIIβ and its native promoter fully rescued the defective morphology of gIIβ−2, suggesting that the mutations in the GIIβ are responsible for the enhanced ABA sensitivity (Fig. 1a, b).

**Fig. 1: NPEs promote desensitisation of ABA transcriptional response.**

As GIIβ is involved in the N-glycosylation pathway, a complex process employing sequential actions of many enzymes that encompasses the biosynthesis, transfer and modification of sugar moieties in ER and Golgi apparatus^24,33. The enhanced sensitivity to ABA of gIIβ alleles prompted us to check the effects of other NPEs in response to ABA. Several previously reported high salt-sensitive mutants were tested. Loss-of-function mutant of STT3A, stt3a-2 exhibited hypersensitivity to ABA treatment (Fig. 1c, d). In contrast, STT3A overexpression (STT3A-OE) obviously reduced sensitivity to ABA treatment compared with that of the WT. Moreover, mutants of GIIα and three NPEs, CGL1, GMII, and FUT11, residing in the Golgi apparatus, namely, rsw3, cgl1-T, hgl1-2 and fut11²⁴, all exhibited hypersensitivity to ABA (Fig. 1e, f). The results of the morphological analyses suggest that the NPEs involved in the N-glycosylation pathway negatively regulate the ABA responses.

GIIβ was induced by prolonged ABA treatment, and the GIIβ protein also accumulated in an ABA dose-dependent manner (Supplementary Fig. 1c, f, g). The temporal expression pattern of the STT3A gene was similar to that of GIIβ (Supplementary Fig. 1d). Tissue expression analysis showed that GIIβ was significantly expressed in the roots (Supplementary Fig. 1e). To further determine the biological roles of NPEs in the ABA signalling pathway, we investigated the transcription of ABA-responsive marker genes in stt3a-2 roots, as the transcriptional response is the most important and conserved ABA response present from semiterrestrial algae to angiosperms³⁴. NCED3, P5CS1, and RAB18 all displayed a biphasic expression pattern (rapid activation followed by a gradual decline from 4 h to 72 h) in the WT roots. Intriguingly, in the stt3a-2, the biphasic inflection point was delayed from 3 h to 4 h for P5CS1 and to 6 h for NCED3 and RAB18. In contrast, this biphasic inflection point shifted forward to 2 h for P5CS1 and NCED3 in the STT3A-OE line, suggesting that NPEs positively regulate desensitisation of the ABA transcriptional response during prolonged ABA signalling (Fig. 1g–i).

To investigate the genome-wide effect of NPEs on ABA transcription dynamics, we performed a time-course RNA sequencing (RNA-seq) analysis with Col-0/WT and stt3a-2 harvested at 0 h to 72 h after application of ABA (see the “Methods”). The transcriptomics datasets of ABA-treated Col-0 were first subjected to k-means clustering analysis to determine the temporal expression patterns with 0 h as control³⁵. The results showed that 8 out of 20 clusters containing 12,337 genes exhibit ABA-induced biphasic expression patterns over time. Among them, 5271 ABA-upregulated genes composing 4 clusters (clusters 1, 8, 13, and 18) displayed rapid induction in response to short-term ABA application, followed by a gradual attenuation during prolonged ABA treatment (Fig. 1j, k and Supplementary Fig. 3a). The expression patterns of NCED3, P5CS1, and RAB18 from the transcriptome were similar to those from RT-PCR (Supplementary Fig. 3b). The expression profiles from stt3a-2 were further retrieved to compare with the WT datasets, and the results showed that the biphasic switch to desensitisation of 1632 out of 5271 genes was delayed in stt3a-2 (Fig. 1j, k and Supplementary Fig. 3a). For example, ABA-responsive genes whose associated Gene Ontology (GO) terms were most highly enriched in responses to vesicle-mediated transport in cluster 1 were rapidly induced and peaked at 6 h in the WT, and the desensitisation inflection point of 689 out of 1288 genes shifted from 6 h to 24 h in stt3a-2 (Fig. 1j, k). A total of 7066 genes in the remaining 4 clusters (clusters 2, 6, 7, and 14) were rapidly downregulated in response to short-term ABA treatment (1–3 h), followed by the gradual increase in the WT background. 840 out of 1499 genes in cluster 6, showed a delayed biphasic inflection point from 1 h in WT to 2 h in stt3a-2 after ABA application (Supplementary Fig. 3c). With respect to the control, a biphasic expression pattern of ABA-responsive genes was not found in the no-ABA treatment datasets (Col-0, ck; Fig. 1k and Supplementary Fig. 3a, c). Moreover, our RNA-seq datasets significantly overlap with the ABA-induced transcriptomes previously reported (Supplementary Fig. 4)³⁶. In summary, the clustering analysis results reveal that the NPE STT3A promotes genome-wide desensitisation of the ABA transcriptional response during prolonged ABA signalling.

SnRK2.2 and SnRK2.3 are N-glycosylated

To further decipher the molecular mechanisms of NPEs acting on ABA transcriptional desensitisation, we examined the pair-wise interaction between NPEs and key components of the ABA-signalling pathway. Yeast two-hybrid (Y2H) assays showed that GIIβ interacts with SnRK2.2/2.3/2.6 but not with 14 PYLs, 2 PP2Cs (ABA insensitive 1 (ABI1) and ABI2), 2 downstream transcription factors (ABI5 and ABA-responsive element binding factor 2 (ABF2)), and SnRK2.10, an ABA-unresponsive SnRK2 kinase³⁷ (Fig. 2a). Coimmunoprecipitation (Co-IP) and Pull-down assays further validated that SnRK2.2/2.3/2.6 can directly interact with GIIβ (Fig. 2b and Supplementary Fig. 5a). Firefly luciferase complementation imaging (LCI) assays showed that GIIβ, STT3A, GIIα, CGL1 and FUT11 all exclusively bind to SnRK2s, and the binding activities were strongly enhanced under prolonged ABA treatment (Fig. 2c and Supplementary Fig. 5b–f). In general, N-glycosylation is thought to be processed in the ER and Golgi apparatus²⁴. Bimolecular fluorescence complementation (BiFC) assays showed that the interactions between SnRK2.2/2.3/2.6-YFP^C and GIIβ-YFP^N took place in the ER (Supplementary Fig. 5g). Collectively, our findings suggest that NPEs directly interact with SnRK2.2/2.3/2.6 in the ER-Golgi network.

**Fig. 2: SnRK2.2/2.3 are N-glycosylated during prolonged ABA signalling.**

The interactions between NPEs and SnRK2s suggest the occurrence of two possibilities: NPEs are phosphorylated by SnRK2 kinases, and/or SnRK2s are modified by N-glycosylation. Phos-tag-based assays showed that GIIβ cannot be phosphorylated by GST-SnRK2.2 in vitro (Supplementary Fig. 6a). To verify the second possibility, we transiently co-expressed SnRK2.3 and STT3A in tobacco leaves, as STT3A could induce a larger molecular weight difference of SnRK2.2/2.3 than other NPEs. Western blots revealed that STT3A coexpression produced a larger band shift of approximately 2-3 kDa (about one N-glycan³⁸, Fig. 2d). Subsequently, PNGase F and Endo H digestions confirmed that the band shift was due to N-glycan modification (Fig. 2e)³⁹. Like SnRK2.3, SnRK2.2 can also be N-glycosylated, but SnRK2.6 cannot in our tests (Fig. 2f, g). As the apparent molecular masses of SnRK2.6 and N-glycosylated SnRK2.2/2.3 are very similar to each other, we verified the N-glycosylation of endogenous SnRK2.2/2.3 in the snrk2.6 background. Upon ABA treatment, glycosylated SnRK2.2/2.3 gradually increased in abundance over time (12 h to 2 d, Fig. 2h). However, only a portion of the N-glycan was removed by PNGase F and Endo H digestions, indicating that SnRK2.2/2.3 N-glycans vary in structure in Arabidopsis in vivo (Fig. 2i). Taken together, these findings suggest that SnRK2.2/2.3 are modified with N-glycan in a prolonged ABA-dependent manner.

As N-glycosylation plays crucial roles in many biological processes, including protein folding, protein stability, and protein targeting²⁴, we checked whether the subcellular localisation of SnRK2.2/2.3 was affected by N-glycosylation. SnRK2.2/2.3-GFP was mainly located in the nucleus to activate transcriptional responses, as previously reported. However, when coexpressed with GIIβ, SnRK2.2/2.3-GFP unexpectedly underwent nuclear export (or suppressed nuclear import) in tobacco leaves (Supplementary Fig. 6b, e)^7,34. Quantitative analysis of the SnRK2.2-GFP fluorescence signal showed that the localisation change of SnRK2.2-GFP was tightly associated with GIIβ coexpression (Supplementary Fig.6c, d). Similarly, co-expression of STT3A-mCherry also led to SnRK2.2/2.3-GFP subcellular change and their co-localisation in the ER, suggesting that N-glycosylation of SnRK2.2/2.3 induces its subcellar change in transiently expressed Nicotiana benthamiana leaves (Fig. 2j and Supplementary Fig.6f–i).

Based on these observations, we hypothesised that the N-glycosylation site mutation would disrupt the subcellular change of SnRK2.2/2.3. A total of 11 N-site mutations of SnRK2.3 were coexpressed with STT3A-mCherry in N. benthamiana leaves, and only the SnRK2.3^N323A still was retained in the nucleus (without change) upon STT3A-mCherry coexpression (Fig. 2j and Supplementary Fig. 6g–i). Western blots showed that the SnRK2.3^N323A mutation led to abolition of the N-glycosylation of SnRK2.3, suggesting that the N³²³ was the bona fide N-glycosylation site (Fig. 2k). The identified SnRK2.3^N323 site resides in a noncanonical N-glycosylation motif NRC⁴⁰, which is also present in SnRK2.2 but not in SnRK2.6. This motif resides in the C-terminal domain II of SnRK2s, an extended conformation protruding out of the three-dimensional structure required for ABA activation (Supplementary Fig. 7a)⁴¹. Moreover, the identified N-glycosylation site of SnRK2.2/2.3 is highly conserved across mammal AMP-activated protein kinases (AMPKs) and yeast Snf1 (Supplementary Fig. 7b).

SnRK2s are epistatic to NPEs in ABA signalling

We generated the quadruple mutant gIIβ−2/snrk2.2/2.3/2.6, and the morphological analysis revealed that loss of SnRK2.2/2.3/2.6 fully suppressed the sensitivity phenotype of gIIβ−2 to ABA in terms of both primary root length and ABA-induced leaf senescence (Supplementary Fig. 8a–d). Moreover, stt3a-2, rsw3, cgl1-T, hgl1-2, and fut11 were introduced into the snrk2.2/2.3 double mutant background to generate triple mutants, respectively. As expected, snrk2.2/2.3 fully repressed the ABA-hypersensitive phenotype of stt3a-2, rsw3, cgl1-T, hgl1-2, and fut11 respectively (Fig. 2l, m and Supplementary Fig. 8e, f). These genetic assays suggest that SnRK2s are epistatic to NPEs in the ABA signalling pathway.

To investigate the biological role of the N-glycosylation of SnRK2 kinases, we generated transgenic lines harbouring SnRK2.3^N323A or SnRK2.3^WT in snrk2.2/2.3 background, respectively. Upon ABA application, SnRK2.3^WT and SnRK2.3^N323A all recovered the hyposensitivity of the snrk2.2/2.3 double mutant (Supplementary Fig. 8g, h). Moreover, the eight SnRK2.3^N323A lines were more sensitive to ABA than the SnRK2.3^WT lines, resembling the morphological differences observed between the NPE mutants and the WT (Fig. 1a–f and Supplementary Fig. 8g, h). Taken together, these findings suggest that the N-glycosylation of SnRK2.2/2.3 desensitises the ABA responses.

SnRK2.2/2.3 redistribute to the peroxisome membrane

According to the subcellular change of SnRK2s in N. benthamiana leaves, we investigated its dynamics in Arabidopsis. Both SnRK2.2-GFP and SnRK2.3-GFP reporter lines complemented the ABA hyposensitivity of the snrk2.2/2.3 double mutant and exhibited protein accumulation patterns similar to that of endogenous SnRK2s, representing rapid induction in response to short-term ABA and subsequent degradation within 6–72 h after ABA treatment (Supplementary Fig. 9a–e)⁴². Microscopy observation showed that strong nuclear localisation of SnRK2.2/2.3-GFP (little diffusion in the cytosol) was detected at 2 h of ABA treatment in root tip cortical cells⁷. However, accompanied by a decrease in total fluorescence in the cells over time, a gradual translocation from the nucleus to vesicle/organelle-like structures in the cytoplasm occurred under prolonged ABA treatment for both SnRK2.2-GFP and SnRK2.3-GFP (Fig. 3a, b). With the aid of red fluorescent protein (RFP)-tagged organelle marker genes and organelle staining (see the “Method”), SnRK2.2/2.3-GFP were found to translocate onto the peroxisomal membrane under prolonged ABA signalling in Arabidopsis roots (Fig. 3c and Supplementary Fig.10a–e). However, this process was not blocked by nuclear export inhibitor leptomycin B (LMB), suggesting that the translocation of SnRK2s might consist of two steps: nuclear SnRK2 decrease and peroxisomal membrane redistribution (Supplementary Fig. 10f).

**Fig. 3: N-glycosylation positively controls SnRK2.2/2.3’s redistribution from the nucleus to the peroxisomes.**

For comparative observations, the SnRK2.2-GFP reporter was introduced into WT, stt3a-2 and STT3A-OE backgrounds, respectively. Intriguingly, SnRK2.2-GFP’s redistribution was delayed in stt3a-2. A stronger fluorescent signal was present in the nucleus and less in the peroxisomes in stt3a-2 compared with the WT at the same time points after ABA treatment, such as at 6 h and 48 h (Fig. 3d, e). In contrast, the SnRK2.2-GFP fluorescent signal was detected more rapidly/earlier in the peroxisomes, while a weaker nuclear fluorescent signal was observed in the STT3A-OE background. For example, at 3 h, an obvious peroxisomal signal was induced in STT3A-OE, whereas it was present at 6 h after ABA treatment in the WT (Fig. 3d, e). Moreover, stronger nuclear fluorescence and less peroxisomal localisation were detected for SnRK2.3^N323A-GFP compared with SnRK2.3^WT-GFP, such as at 48 h and 72 h after ABA treatment (Fig. 3f–h). On the other hand, the expression patterns of RAB18 and NCED3 showed that the biphasic switch of the ABA transcriptional response from activation to desensitisation was also delayed in SnRK2.3^N323A plants (Fig. 3i, j). The findings suggest that desensitisation of the ABA transcriptional response is orchestrated by SnRK2 redistribution from the nucleus to the peroxisomes.

Furthermore, we sought to determine whether the molecular mechanisms regulated by the N-glycosylation of SnRK2s are required for plant adaptation to salt stress. First, gIIβ mutants exhibited enhanced sensitivity to 125 mM NaCl, as did mutants deficient in STT3A, GIIα/RSW3, CGL1 and FUT11 (Supplementary Fig. 9f–i)^24,27,31. In contrast, overexpression of GIIβ and STT3A relieved the growth inhibition induced by high salt (Supplementary Fig. 9j, k). SnRK2.3^N323A plant also resulted in hypersensitivity to high-salinity stress (Supplementary Fig. 9l, m). These morphological assays suggested that N-glycosylation of SnRK2 kinases plays positive biological roles in the response to high-salinity stress. Second, we investigated the subcellular dynamics of SnRK2s under high-salt and osmotic stresses. The results showed that SnRK2.2-GFP was also redistributed to the peroxisomes during prolonged salinity treatment, as well as during prolonged osmotic stress treatment (Supplementary Fig. 10g, h). These findings indicate that plants may exploit SnRK2 subcellular dynamics to better adapt to high-salt and osmotic stresses.

N-glycosylation of SnRK2s modulates H₂O₂ accumulation

The desensitisation of the transcriptional response appears to be mainly related to the decrease of nuclear SnRK2s, prompting an interesting question concerning the biological role of peroxisome-redistributed SnRK2.2/2.3 achieved by N-glycosylation during prolonged ABA signalling. It is well known that peroxisomes play key roles in the regulation of reactive oxygen species (ROS) homoeostasis⁴³, and H₂O₂ is essential for ABA signal transduction^12,44. To elucidate the biological role of peroxisomal SnRK2s, we investigated the H₂O₂ accumulations during prolonged ABA application in the WT and N-glycosylation-defective mutants. H₂O₂ was induced in the WT roots after ABA application and peaked at 48 h, and then declined until 168 h (Fig. 4a). The biphasic pattern of H₂O₂ level was further confirmed with 3,3’-diaminobenzidine (DAB) staining and the H₂O₂ fluorescent indicators 3’-(p-hydroxyphenyl) fluorescein (HPF) and BES-H₂O₂-Ac (see the “Methods”) in both Col-0 and SnRK2.3^WT root tip (Supplementary Fig. 11a–f). Intriguingly, further investigations showed that the switch to the desensitisation phase of ABA-induced H₂O₂ was delayed from 48 h to 96 h in stt3a-2 and SnRK2.3^N323A, each of which exhibited significantly higher H₂O₂ levels during 96–168 h after ABA treatment, suggesting that N-glycosylation of SnRK2s inhibits H₂O₂ accumulation under prolonged ABA signalling in root tip (Fig. 4a and Supplementary Fig. 11a–f). As a control, the superoxide O₂^•- level did not show significant divergence at most time points of prolonged ABA signalling in different genotypes (stt3a-2 vs. Col-0; SnRK2.3^N323A vs. SnRK2.3^WT; Supplementary Fig. 12a–d).

**Fig. 4: SnRK2-GPT1 module regulates NADPH homoeostasis during prolonged ABA signalling.**

H₂O₂ is required for ABA-modulated stomatal closure^12,44, senescence of leaves⁴⁵ and primary root growth inhibition¹². To verify the biological role of H₂O₂ scavenging mediated by N-glycosylation of SnRK2s, we carried out morphological analyses with H₂O₂ scavengers. Applications of the antioxidant reduced glutathione (GSH) significantly repressed the ABA-induced root growth inhibition and reduced root meristem cell numbers both in Col-0 and the SnRK2.3^WT line, supporting the notion that H₂O₂-mediated oxidative signalling plays key roles in ABA-induced plant growth inhibition⁴⁶. Moreover, the recovery of root growth was more pronounced in stt3a-2 and SnRK2.3^N323A plants after treatment with GSH or H₂O₂-scavenger KI (Supplementary Fig. 13a–f). For example, the primary root length of stt3a-2 is 62% that of Col-0 upon ABA treatment, whereas GSH application ultimately restored this percentage to 90% (Supplementary Fig. 13a, b). These findings suggest that glycosylated SnRK2s relieve ABA-induced growth inhibition by reducing H₂O₂ content during prolonged ABA signalling, which helps plants to better adapt to long-term unfavourable conditions.

The SnRK2-GPT1 module regulates NADPH maintenance

Plant peroxisomes mainly counter H₂O₂ overload with catalases (CATs) in the matrix and ascorbate peroxidases (APXs) involved in the ascorbate-glutathione cycle with reducing power from NADPH on the membrane⁴³. To identify the H₂O₂ scavenger regulated by N-glycosylation of SnRK2.2/2.3, we investigated the potential SnRK2-interacting partners among 13 published peroxisomal isozymes involved in H₂O₂-scavenging and NADPH generation (Supplementary Fig. 14a). The results of LCI assays showed that only glucose-6-phosphate (G6P)/phosphate translocator 1 (GPT1) exhibited reliable interaction with SnRK2.2/2.3 (APX3 binds to SnRK2.2/2.3 in an ABA-independent manner, and pull-down assay failed to show a direct interaction between them. Supplementary Fig. 14a, b). GPT1 was previously reported to be translocated to the peroxisomal membrane under stress conditions⁴⁷. Time-course Co-IP analyses showed that short-term ABA-induced very weak interactions between GPT1 and SnRK2s before 6 h, however, stronger binding activity was observed at 24–72 h after ABA treatment (Fig. 4b). BiFC assays showed that the interactions between SnRK2.2/2.3 and GPT1 occur on the peroxisomal membrane (Fig. 4c and Supplementary Fig. 14c). Moreover, GPT1 protein level was found to be induced by prolonged ABA treatment (Supplementary Fig. 14d). These findings suggest that SnRK2s bind to GPT1 on the peroxisomal membrane during prolonged ABA signalling.

As GPT1 is thought to be involved in the peroxisomal oxidative pentose phosphate pathway (OPPP) for NADPH generation (Supplementary Fig. 14a)⁴⁷, we investigated the dynamics of NADPH accumulation. In the Col-0 background, ABA induces an increase of NADPH/NADP⁺ before 24 h, which is followed by a decline in the middle, then a re-accumulation from 48 h to 168 h, indicating two NADPH-producing phases induced by short-term and long-term ABA respectively (Fig. 4d). Intriguingly, compared with the WT (Col-0 and SnRK2.3^WT), the deficiency of N-glycosylation in stt3a-2 and SnRK2.3^N323A delayed the phasic switch of NADPH/NADP⁺ and apparently repressed the later accumulation (96 h later) of NADPH/NADP⁺, suggesting that N-glycosylated SnRK2s function in NADPH maintenance during prolonged ABA signalling when they are redistributed to the peroxisomes (Fig. 4d). As controls, the endogenous GSH/GSSG and NADH/ NAD⁺ levels did not significantly differ in stt3a-2 or SnRK2.3^N323A compared with the Col-0 and SnRK2.3^WT, respectively, although their accumulation was indeed affected by ABA application (Supplementary Fig. 15a, b). Furthermore, applications of OPPP inhibitors, polydatin and glucosamine-6-phosphate (GN6P), dramatically decreased the prolonged ABA-induced NADPH/NADP⁺ accumulation and negated the significant difference between N-glycosylation-deficient mutants and the WT (stt3a-2 vs. Col-0 and SnRK2.3^N323A vs. SnRK2.3^WT, Supplementary Fig. 15c). Correspondingly, the significant difference in H₂O₂ accumulation between N-glycosylation-deficient mutants and the WT no longer occurred upon inhibitor applications (Supplementary Fig. 15d–f). The findings suggest that N-glycosylated SnRK2.2/2.3 regulate NADPH maintenance relying on OPPP activity, sequentially leading to H₂O₂ modulation during prolonged ABA signalling.

To verify whether GPT1 regulates NADPH concentrations during prolonged ABA treatment, we attempted to measure the NADPH/NADP⁺ level in a mutant lacking GPT1. As the gpt1(-/-) homozygous mutant exhibits an embryo-lethal phenotype⁴⁸, artificial miRNAs of GPT1 (amiR-GPT1) were employed. The expression of amiR-EV (an artificial miRNA expression vector) or amiR-GPT2 led to significant decreases in NADPH/NADP⁺ levels and corresponding increases in H₂O₂ levels after 96 h ABA treatment in stt3a-2 and SnRK2.3^N323A compared with Col-0 and SnRK2.3^WT, respectively (Supplementary Fig. 16a–f). However, the significant differences in both NADPH/NADP⁺ and H₂O₂ levels between N-glycosylation-deficient mutants and the WT (stt3a-2 vs. Col-0, SnRK2.3^N323A vs. SnRK2.3^WT) were negated when amiR-GPT1s were expressed (amiR-GPT1-1 and amiR-GPT1-2 in Supplementary Fig. 16a–f and i). Taken together, these findings suggest that the SnRK2-GPT1 module orchestrates NADPH homoeostasis and H₂O₂ scavenging during prolonged ABA signalling on the peroxisomes.

Using GST-SnRK2.2/2.3 and GPT1 N-terminus (GPT1^N) purified from Escherichia coli, we carried out in vitro phosphorylation assays. The results showed that GPT1^N was phosphorylated by SnRK2.2/2.3 but not at a previously identified phosphorylation site, GPT1Ser27 (Supplementary Fig. 14g)⁴⁷. Subsequent LC-MS/MS and site mutation analyses validated that the phosphorylation occurred on the Ser32 site of GPT1 (Fig. 4e and Supplementary Fig. 14e, f, h). Furthermore, in the amiR-GPT1-1 background, GPT1^WT, GPT1^S32A, and GPT1^S32D with the anti-amiR-GPT1-1 sequence (in which the amiR-GPT1-1 binding site was replaced with synonymous codons, as shown in Fig. 4f) were expressed in Arabidopsis roots, respectively. During ABA treatments, the nonphosphorylation-mimicking mutant GPT1^S32A exhibited a significantly decreased NADPH/NADP⁺ ratio and increased H₂O₂ levels compared with those of GPT1^WT, mainly in the prolonged ABA signalling stage (Fig. 4g, h and Supplementary Fig. 16g, h). In contrast, expression of the phosphorylation-mimicking mutant GPT1^S32D gave rise to significantly increased NADPH accumulations and decreased H₂O₂ levels accordingly during prolonged ABA signalling in both amiR-EV and amiR-GPT1-1 backgrounds (Fig. 4g, h). Taken together, our findings suggest that phosphorylation of GPT1 at the Ser32 site catalysed by SnRK2s controls NADPH homoeostasis on the peroxisomal membrane, which alleviates H₂O₂-mediated oxidative stress during prolonged ABA signalling.

Finally, to verify whether the NADPH maintenance regulated by N-glycosylation is due to the peroxisome targeting of SnRK2.2/2.3 during prolonged ABA signalling, we constructed two fixed SnRK2.3 s specifically located on the peroxisomal membrane (SnRK2.3-mPTS^PEX26) and in the nucleus (SnRK2.3-NLS) respectively, and they possessed normal kinase activity but different protein expression patterns (Supplementary Fig. 17a–f). When expressed in snrk2.2/2.3 double mutant background, SnRK2.3-mPTS^PEX26 only exhibited one NADPH-accumulation phase induced by long-term ABA treatment, while SnRK2.3^WT displayed two NADPH-producing phases (Fig. 4i). Moreover, SnRK2.3-mPTS^PEX26 possessed significantly higher NADPH level than SnRK2.3^WT, which repressed H₂O₂ to very low level during prolonged ABA signalling (Fig. 4j). SnRK2.3-NLS also showed one NADPH-accumulation phase but induced by short-term ABA (before 24 h), lacking the NADPH maintenance phase although its NADPH level was not lower than SnRK2.3^WT during prolonged ABA signalling (before 168 h, Fig. 4i). The accumulation level of H₂O₂ was significantly higher in SnRK2.3-NLS than that in SnRK2.3^WT (Fig. 4j). Moreover, the transcriptional expression of ABA-induced genes RAB18 and NCED3 was enhanced by both 2 h and 48 h ABA treatments in SnRK2.3-NLS, whereas almost no any obvious expression of the two genes were tested in SnRK2.3-mPTS^PEX26 (Supplementary Fig. 17g, h). Totally, the findings support the notions that SnRK2s activate ABA-transcriptional response in the nucleus, and are redistributed to the peroxisomes to maintain the NADPH accumulation for relieving H₂O₂ stress during prolonged ABA signalling.

Discussion

Here, we revealed a molecular mechanism by which N-glycosylated SnRK2.2/2.3 are subcellularly redistributed to the peroxisomes and control prolonged ABA-signalling for long-term stress acclimation in plants. As master regulators of ABA signalling, SnRK2.2/2.3/2.6 were previously found to be located in the nucleus to activate the transcriptional response, and SnRK2.6 has also been shown to interact with RBOHD/F to activate H₂O₂ signalling on the plasma membrane under short-term ABA treatment^10,11,13. According to our results, SnRK2.2/2.3 were unexpectedly found to be N-glycosylated in the ER-Golgi network during prolonged ABA signalling, which ultimately orchestrates the gradual redistribution of these proteins from the nucleus onto the peroxisomal membrane, leading to desensitisation of the transcriptional response in the nucleus. Peroxisomal SnRK2.2/2.3 kinases subsequently directly interact with the G6P transporter GPT1 to enhance peroxisomal OPPP activity in controlling NADPH maintenance, which is critical for redox rebalancing under long-term stress conditions (Supplementary Fig. 18). Our findings suggest that N-glycosylation of SnRK2.2/2.3 regulates desensitisation of the ABA response, which does not involve just deactivation of ABA signalling, but actually appears to reflect a transition state from transcriptional reprogramming gradually switching to metabolic regulation for NADPH maintenance in the peroxisomes.

N-glycosylation of SnRK2s is proposed here to be a basic principle promoting the cellular state transition during prolonged ABA signalling. The N-glycosylation is specific for secretory and membrane proteins processed in the ER-Golgi network^22,23. How SnRK2s are associated with ER-Golgi needs further investigation. The subcellular dynamics of SnRK2s function as a direct cellular effector to regulate prolonged ABA signalling. SnRK2s’ redistributions from the nucleus to the peroxisomes was mainly detected in the heterotrophic root tissue of Arabidopsis. Nuclear export inhibitor LMB did not block this process (Supplementary Fig. 10f), indicating that it is not directly translocation from the nucleus to peroxisomes. Combined with the results from CHX and MG132 treatments, this process may involve nuclear decrease and peroxisomal membrane re-sorting, both of which are regulated by N-glycosylation of SnRK2s (Fig.3a–h and Supplementary Fig. 10f).

NADPH accumulation underwent two distinct accumulation phases, indicating that different molecular mechanisms were involved in NADPH generation during prolonged ABA signalling. Our observations suggested that the SnRK2-GPT1 module mainly regulated the latter generation phase (Fig. 4d, g, i). In plants, the photosynthetic electron transfer chain is a major resource of NADPH in leaves, however at night, the OPPP located in chloroplast and cytosol provides NADPH, and also in heterotrophic tissues^49,50. Both phases of NADPH generation induced by ABA are thought to be a result of OPPP activity in Arabidopsis roots, which is supported by OPPP inhibitor assays (Supplementary Fig. 15c). Considering H₂O₂ dynamics, the initial increase in NADPH abundance is proposed to be provided by cytosolic OPPP for ABA-induced H₂O₂ production via the activity of RBOHD/F^12,50, whereas the latter NADPH accumulation regulated by SnRK2-GPT1 comes from peroxisomal OPPP pathway, which is involved in relieving H₂O₂-mediated growth inhibition during prolonged ABA signalling (Fig. 4a, d, g–j and Supplementary Fig. 13a–f). The temporal expression pattern of the GPT1 protein also supports this hypothesis, as the decrease in GPT1 before 24 h-ABA application indicates that more G6P is retained in the cytosol, whereas the subsequent GPT1 induction implies increased peroxisomal OPPP activity during prolonged ABA signalling (Supplementary Fig. 14d).

To summarise, peroxisomes are highly dynamic in terms of their size, shape, abundance and metabolic processing in response to environmental changes and are thus thought to be cellular sensors that respond to metabolic changes and abiotic stresses^43,51,52. as plant-specific homologues of yeast Snf1 and mammalian AMPK, which are key regulators of the limited energy status in response to stresses, SnRK2 kinases were found to directly orchestrate NADPH homoeostasis in the peroxisomes during prolonged ABA signalling. Our findings would shed light on the molecular mechanisms involved in plant acclimation under long-term stress conditions.

Methods

Plant materials, growth conditions

Arabidopsis thaliana plants used in this study were in the Columbia-0 (Col-0) background. Seeds were sterilised for 6 h with chlorine before being placed on half-strength Murashige and Skoog (1/2 × MS) medium (1.5% sucrose, 0.8% agar, pH 5.8–5.85) with or without ABA. The plates were stratified at 4 °C for 2 d in the dark and were then transferred to a growth chamber at 23 °C with a long-day photoperiod cycle (16- h light/8- h dark). Phenotypes were assessed 7-14 d after the transfer of plants to the growth chamber. Arabidopsis thaliana plants and Nicotiana benthamiana plants were grown in the greenhouse under long-day conditions (22 °C, 16- h light/8- h dark). For the experiments with OPPP inhibitors, seedlings were grown for 4 days before being transferred to the inhibitor-containing medium, and the subsequent experiments were carried out after 4 days of growth. The concentrations of GN6P (Sigma) and polydatin (Macklin) were 3 μM and 10 μM, respectively. For the salt treatment, 2-day-old seedlings were transferred to 1/2 × MS plates with 125 mM NaCl or 100 mM NaCl. Phenotypes were assessed 7–14 d, and the data of root length was counted.

Transfer DNA insertion lines, gIIβ−2 (SALK_120812)⁵³, gIIβ−3 (SALK_093735)⁵³, gIIβ−4 (SALK_039458)⁵³, stt3a-2 (CS800052)²⁷, rsw3 (SALK_124837C)²⁶, fut11 (SALK_087481), cgl1-T (SALK_073650) and hgl1-2 (SALK_141821)²⁸ were obtained from the Arabidopsis Biological Resource Centre (ABRC). snrk2.2/2.3/2.6 was obtained by crossing snrk2.2 (GABI-Kat 807G04) snrk2.3 (SALK_107315) with snrk2.6 (SALK_008068)⁸, and stt3a-2/snrk2.2/2.3, rsw3/snrk2.2/2.3, fut11/snrk2.2/2.3, cgl1-T/snrk2.2/2.3, hgl1-2/snrk2.2/2.3 were obtained by crossing stt3a-2, rsw3, fut11, cgl1-T, hgl1-2 with snrk2.2/2.3. In addition, gIIβ−2/snrk2.2/2.3/2.6 screened the F2 for quadruple homozygous progeny, after gIIβ−2 and snrk2.2/2.3/2.6 were crossed. Complementation lines (com-1, com-2) were generated by transforming gIIβ−2 with pGIIβ::GIIβ-GFP by Agrobacterium-mediated transformation. Overexpression lines under the control of the 35 S promoter were generated by transforming Col-0. After all the above transgenic lines were homozygous and the expression levels of the genes were verified, phenotypic observations and biochemical experiments were performed.

Quantitative real-time PCR

RNA was extracted from 0.1 g seedlings by Trizol Reagent (TaKaRa), followed by treatment with RNase-free DNase I (TaKaRa) at 42 °C for 2 min. For testing gene expressions, 1 μg total RNA was used for reverse transcription using reverse transcriptase (TaKaRa) according to the manufacturer’s instructions. Real-time PCR was performed using SYBR Green real-time PCR master mix (TaKaRa) on a CFX96 real-time PCR detection system (Bio-Rad). UBQ10 (AT4G05320) and ACTIN2 (AT3G18780) were used as the internal reference for all reactions.

Detached leaf senescence and measurement of chlorophyll content

Detached Arabidopsis thaliana rosette leaves of 3-week-old were soaked in water with or without 50 μM ABA. After 3 days (22 °C, Dark), the rosette leaves were photographed in a glass petri dish filled with water. The Chlorophyll of leaves was extracted using 80% acetone and the supernatant was taken after centrifugation (18514 × g, 10 min). Chlorophyll quantified the absorbance at 663 nm and 645 nm, respectively. Then, the content of chlorophyll was calculated by the previously established formula as (20.21 × OD₆₄₅ + 8.02 × OD₆₆₃)/1000 × V/W [“V” indicates volume of extraction buffer (ml) and “W” indicates fresh weight of leaves (g)]⁵⁴.

Yeast two-hybrid analysis

The coding sequences (CDS) of genes were amplified using DNA polymerase (Vazyme, P505-d1) using primers listed in Supplementary Table 1 and cloned individually into pGBKT7 or pGADT7. Various combinations of the bait and prey constructs (the empty vectors as controls) were then co-transformed into yeast strain Yeast Gold (Clontech). Competent yeast cells were prepared according to the protocol described below. Yeast cells from a frozen yeast stock were streaked on a YPDA agar plate and grew at 30 °C until colonies appeared. A colony of suitable size (2–4 mm) is selected and inoculated into the 3 ml YPDA medium at 30 °C for 12 h. The 5 µl culture was transferred to the 50 ml YPDA in a 250 ml flask. When the OD₆₀₀ reached 0.15-0.3, the yeast cells were collected through centrifugation at 700 × g for 5 min and resuspended into 100 ml of fresh YPDA medium. After the OD₆₀₀ reached 0.4-0.5, the cells were centrifuged (700 × g, 5 min). The pellet was resuspended into sterile, deionized H₂0. Subsequently, the cells were collected again, followed by resuspension in 1.5 ml of 1.1 × TE/LiAc. The cell suspensions were transferred to 1.5 ml microcentrifuge tubes and centrifuged at high speed for 15 seconds. Finally, the pellet was resuspended with 600 µl of 1.1 × TE/LiAc. The 600 µl of competent cells were aliquoted into sterile centrifuge tubes, each containing 50 µl of competent cells.

The construction of successfully recombined pGBKT7 and pGADT7 with carrier DNA were combined in a sterile tube. Next, the DNA mixture was added to competent cells and mixed gently. PEG/LiAc was added and incubated at 30 °C. After 30 min, DMSO was added, and the mixture was placed in a 42 °C water bath for 15 min. After centrifuged at high speed for 15 s, the pellet was resuspended in YPDA. After 3 h incubation at 30 °C, yeast cells were plated onto the synthetic dropout medium lacking leucine (Leu) and tryptophan (Trp). Yeast cells were selected for growth for 2 days at 30 °C and inoculated by growing on QDO medium (minimal media quadruple dropouts, SD medium with -Leu/-Trp/-Ade/-His, Clontech), containing 40 μg/ml x-α-gal. After 2-3 days of growth at 30 °C, Photos of yeast growth were taken using a camera (Nikon).

Co-immunoprecipitation analysis

For interaction between GIIβ and SnRK2.2/2.3. The samples of 0.5 g 7-day-old pGIIβ::GIIβ-GFP transgenic lines (p35S::GFP transgenic lines as negative control) were grounded to powder in the frozen state and homogenised in 2 ml extraction buffer (50 mM Tris-HCl, pH 7.4, 150 mM NaCl, 1 mM MgCl₂, 10 μM EDTA, 0.2% NP-40, 1 × protease inhibitor, 5 mM DTT and 0.5 mM PMSF). Centrifugation (13520 × g, 10 min, 4 °C) was performed, and the supernatant was incubated with 10 μl GFP antibody (Abcam) at 4 °C. After 4 h, the mixture was incubated with 20 μl protein A/G beads (22202-20, BeaverBeads^TM) for 8 h at 4 °C. Beads were washed five times with ice-cold washing buffer (50 mM Tris-HCl, pH 7.4, 400 mM NaCl, 1 mM MgCl₂, 0.02% NP-40, 1 × protease inhibitor, 5 mM DTT and 0.5 mM PMSF) and the proteins bound to the beads were released by 5 min boiling. Samples were separated by SDS–PAGE and analysed by immunoblot using GFP antibody (1:1000, Abcam, ab290) and SnRK2.2/2.3/2.6 antibody (1:1000, Agrisera, AS142783) sequentially.

For interaction between GPT1and SnRK2.2/2.3, the pGPT1::GPT1-YFP transgenic lines were treated with 50 μM ABA for the corresponding time after seven days of growth. Roots of transgenic lines were grounded to powder in the frozen state and homogenised in extraction buffer. In addition to using 50 μl GFP-Trap agarose (Chromotek, gta-20) instead of GFP antibody, the following operations were the same as above.

Firefly luciferase complementation imaging (LCI) assays

The full-length coding sequences of STT3A, GIIα, GIIβ, CGL1, FUT11, GPT1, GPT2, CAT2, PXN were cloned individually into pCambia1300-nLUC vector and RBOHD, RBOHF, APX3, MDAR1, MDAR4, PICDH, NADK3 were cloned individually into pCambia1300-cLUC vector and the full-length coding sequences of SnRK2.2, SnRK2.3 and SnRK2.6 were cloned into pCambia1300-cLUC vector or pCambia1300-nLUC. All the constructs were transformed into the A. tumefaciens strain GV3101. GV3101 containing the recombinant plasmid was inoculated into 3 ml of YEP medium (kan⁺). After culturing at 28 °C for 12 h, the strains were transferred again. The 1 ml of bacterial broth was added to 10 ml of fresh YEP medium (Kan⁺ 10 mM MES and 20 μM acetosyringone) after 12 h. When the OD₆₀₀ value was 1.0, the cells were collected and resuspended with MMA solution (10 mM MES pH 5.6, 10 mM MgCl₂, and 200 μM acetosyringone) to achieve an OD₆₀₀ value of 1.5. P19 plasmid was co-infiltrated to suppress gene silencing. After the resuspended cells were incubated in the dark for 3 h, equal volumes of different GV3101 were mixed and injected into N. benthamiana leaves. Infected tobacco was incubated in the dark for 8 hours and grown under normal growth conditions. For ABA treatment, 12 h after GV3101 was injected, 50 μM ABA was injected into leaves (0.05% DMSO as a control). After 2.5 days, 1 mM luciferin solution was sprayed on the leaves. Then, luciferase activity was detected using a low-light cooled CCD imaging apparatus (Lumina II, USA).

Bimolecular fluorescence complementation (BiFC) analysis

The full-length coding sequences of SnRK2.2, SnRK2.3 and SnRK2.6 were cloned into pCambia1300-YFP^C vector, and the full-length coding sequences of GIIβ and GPT1 were cloned into pCambia1300-YFP^N vector. Agrobacterium tumefaciens (GV3101) was cultured according to the method mentioned above. GV3101 carrying the vectors were infiltrated into N.benthamiana leaves. After 2.5 days, reconstituted YFP signals, ER marker and dsRed signals of dsRed-mPTS^PEX26 were imaged on a Zeiss LSM880 confocal laser scanning microscope (Zeiss, Germany).

GST pull-down assay

The full-length CDS of SnRK2.2, SnRK2.3 and SnRK2.6 were cloned into pGEX4T-3 and the CDS of APX3 (1–771 bp, APX3ΔTM) were cloned into pET30a. All PCR primers were listed in Supplementary Table 1. All fusion proteins (GST, GST-SnRK2.2, GST-SnRK2.3, GST-SnRK2.6 and APX3ΔTM-His) were produced in E. coli Rosetta strain cells with the induction conditions (0.5 mM isopropyl β-D-thiogalactopyranoside, 16 °C for 20 h). Recombinant proteins were purified according to the manufacturer’s protocol (Beyotime, P2233 and P2253). GIIβ−3 × FLAG proteins were produced by transforming 293 T cells with pCMV-FLAG. To pull-down GIIβ−3 × FLAG, 100 μg of GST-SnRK2.2, GST-SnRK2.3 and GST-SnRK2.6 incubated with GIIβ−3 × FLAG in pull-down buffer (Beyotime, P2233 and P2253). Then 10 μl Glutathione Sepharose 4B beads were added and incubated for 8 h at 4 °C. After incubation, the beads were washed 5 times with washing buffer (Beyotime, P2233 and P2253), and the proteins were released by boiling in 5 × SDS sample buffer. GIIβ−3 × FLAG was detected by immunoblotting with monoclonal anti-FLAG antibody (1:1000, 20543-1-AP, proteintech^TM). For pull-down APX3ΔTM-His, the following operations were the same as above, and APX3ΔTM-His was detected by immunoblotting with anti-His antibody (1:1000, Transgen Biotech, HT501-01). GST-tagged fusion proteins were detected by using an anti-GST antibody (1:1000, Transgen Biotech, HT601-01).

In vitro kinase assay

In vitro phosphorylation assay was carried out according to a previous study⁵⁵. For phosphorylation assay of GIIβ, 0.5 μg GST-SnRK2.2 and 0.2 μg GIIβ−3 × FLAG (ABF2-His as positive control) were added to kinase buffer (20 mM Tris-HCl pH 7.4, 100 mM NaCl, 12 mM MgCl₂, 1 mM DTT, and 1 mM ATP). After the reaction for 1 h or 2 h at 25 °C, the 5 × SDS loading buffer was added to stop the reaction. Samples were separated by SDS–PAGE with Phos-tag. GIIβ−3 × FLAG and ABF2-His were detected by using anti-FLAG antibody and anti-His antibody.

For phosphorylation assay of GPT1, the N-terminal region of GPT1(1–96 amino acid, GPT1^N) and mutants of GPT1 N-terminal region (GPT1^N,S27A, GPT1^N,S30A, GST-GPT1^N,S32A) were amplified using DNA polymerase (Vazyme, P505-d1) using primers listed in Supplementary Table 1 and cloned into pGEX4T-3. The expressed fusion protein was used for subsequent experiments in E. coli Rosetta strain cells. 0.5 μg purified GST-SnRK2s (GST-SnRK2.2 or GST-SnRK2.3) were incubated with purified 0.2 μg GST-GPT1^N (or GST-GPT1^N,S27A, GST-GPT1^N,S30A, GST-GPT1^N,S32A) and were added to kinase buffer. After 2 h at 30 °C, The reaction was stopped by adding 5 × SDS loading buffer. For phosphatase treatment, CIAP was added to the buffer after the kinase reaction. Samples were separated by SDS–PAGE and analysed by immunoblot. Phosphoserine/threonine antibody (1:1000, EMC biosciences, PP2551) was used to detect phosphorylated GPT1.

N-glycosylation analysis of SnRK2s

For N-glycosylation analysis, we mutated the asparagine at sites 288 and 323 of SnRK2.3 to alanine (SnRK2.3^N288A, SnRK2.3^N323A) using primers (Supplementary Table. 1) and fused the sequence of His tag to the 3’end of SnRK2.2/2.3/2.6/SnRK2.3^N288A/SnRK2.3^N323A sequence through primers (Supplementary Table. 1). The full-length SnRK2.2/2.3/2.6/SnRK2.3^N288A/SnRK2.3^N323A -His sequence was cloned into the pMDC83 binary vector driven by a 2 × 35 S promoter. The 35 S::SnRK2.2/2.3/2.6/SnRK2.3^N288A/SnRK2.3^N323A -His sequence was amplified by PCR and cloned into the pCambia1300-mCherry containing pSuper::STT3A-mCherry to form two expression boxes. STT3A-mCherry and SnRK2.2/2.3/2.6/SnRK2.3^N288A/SnRK2.3^N323A -His were transiently co-expressed in tobacco leaves. After the leaf protein was extracted with extraction buffer (50 mM Tris-HCl, pH 7.4, 150 mM NaCl, 1 mM MgCl₂, 10 μM EDTA, 0.2% NP-40, 1 × protease inhibitor, 5 mM DTT and 0.5 mM PMSF), the protein sample was separated in 6% SDS-PAGE gel and detected through immunoblots using anti-His antibody. For N-glycan analysis in Arabidopsis thaliana, proteins were extracted from 7-day-old snrk2.6 with or without 50 μM ABA and proteins were separated in 6% SDS-PAGE gel and detected through immunoblots using SnRK2.2/2.3/2.6 antibody (1:000, Agrisera, AS142783).

For peptide-N-glycosidase F (PNGase F, PNG F) and Endoglycosidase H (Endo H) treatment³⁸, Proteins extracted from plants were subjected to deglycosylation treatment with PNGase F or Endo H (New England Biolabs). Refering to the manufacturer’s protocol, 10 × glycoprotein denaturing buffer was added to extracted proteins and heated at 100 °C for 10 min. After returning to room temperature, the reaction buffer and endoglycosidases were added to the sample. The mixture was incubated for 1 h at 37 °C before being boiled in 5 × SDS loading buffer. The protein mixture was detected by immunoblot analyses using specific antibodies after 6% SDS-PAGE.

Measurement of ROS in plants

Before measuring ROS content, seeds were germinated on 1/2 × MS medium and grown to a 4-day-old seedling stage. The seedlings were then transferred onto an agarose-solidified growth medium supplemented with ABA. Col-0 and stt3a-2 were treated with 3 μM ABA for 7 days. 35 S::SnRK2.3^WT-GFP/snrk2.2/2.3 (SnRK2.3^WT) and 35 S::SnRK2.3^N323A-GFP/snrk2.2/2.3 (SnRK2.3^N323A) transgenic lines were treated with 1 μM ABA for 7 days. After the corresponding time, ROS content was measured according to the following methods.

For DAB (3’, 3’- diaminobenzidine) staining to detect H₂O₂, seedlings were incubated in 1 mg/ml DAB (Sigma-Aldrich, and adjust pH to 3.8 with NaOH). After removing the air from the seedlings with a vacuum pump (− 850 Mbar, 30 min), the seedlings were kept at 28 °C and protected from light for 8 h. After staining, the dyeing solution was discarded, and the ethanol: lactic acid: glycerin (3:1:1) was added. Then the seedlings were decolorised for 5 min in boiling water and soaked in absolute ethanol. For NBT (nitroblue tetrazolium) staining to detect superoxide, the seedlings were incubated in 0.5 mg/ml NBT (Sigma-Aldrich). The next steps were the same as DAB staining. The roots were then photographed using a Carl Zeiss AX10 fluorescence microscope.

For BES-H₂O₂-Ac and HPF staining to detect H₂O₂, the method refer to Hironaka Tsukagoshi et al.⁵⁶. Seedlings were incubated in 50 μM of BES-H₂O₂-Ac (WAKO) and 5 μM HPF (Maokang Bio-Technology) for 30 min in the dark. In addition, for staining to detect superoxides, 10 μM DHE (Maokang Bio-Technology) was used to stain seedlings after ABA treatment. Images were acquired using an LSM880 confocal laser scanning microscope (Zeiss, Germany).

For the measurement of H₂O₂ content, the kit (Geruisi) was used. 0.1 g of roots were frozen in liquid nitrogen and ground to a powder, and then 1 ml of pre-cooled acetone was added. After standing for 30 min, the samples were centrifuged (12000 rpm, 4 °C, 10 min). 50 μl solution 1 and 100 μl solution 2 mixed with supernatant and incubated at 25 °C for 10 min. After centrifugation (13520 × g, 25 °C, 10 min), the 700 μl solution 3 was added to the pellet. After the pellet dissolved, the supernatant was transferred to a cuvette, and the absorbance was measured at 415 nm (AS). After the sample was replaced with acetone, the measured absorbance value was the absorbance value of the control (A_CK). ΔA = A_S-A_CK. For the standard curve line, the standards were diluted with acetone to different concentrations (0, 0.4, 0.8, 1, 1.2, 1.4, 1.6, 1.8, 2.0 μmol/ml). The absorbance values were determined according to the method described above. The standard curve equation: y = 0.9491x + 0.0078. The H₂O₂ content in the sample was obtained according to the following formula: H₂O₂ content (μmol/g) = [(y-0.0078)/0.9491]/(W × V1 × V) × D. W (Sample quality, g), V1 (Sample volume in the reaction system, 0.5 ml), V (Extract volume, 1 ml), D (Dilution factor).

Measurement of NADH/NAD⁺, NADPH/NADP⁺ and GSH/GSSG

Before measurement, seeds were germinated on 1/2 × MS medium and grown to a 4-day-old seedling stage. The seedlings were then transferred onto an agarose-solidified growth medium supplemented with ABA. Col-0 and stt3a-2 were treated with 3 μM ABA for 7d. 35 S::SnRK2.3^WT-GFP/snrk2.2/2.3 (SnRK2.3^WT) and 35 S::SnRK2.3^N323A-GFP/snrk2.2/2.3 (SnRK2.3^N323A) transgenic lines were treated with 1 μM ABA for 7days. The content was measured using the kit (Beyotime). For measurement of NADH/NAD⁺, 0.1 g roots were fully ground with liquid nitrogen. The extraction solution (Beyotime, S0175-5) was added and incubated at 4 °C for 20 min. After centrifuging the sample, alcohol dehydrogenase solution was added to the supernatant. After incubating at 37 °C for 10 min in the dark, chromogenic solution (Beyotime, S0175-2) was added. the absorbance value was measured at 450 nm after 37 °C for 30 min in the dark. The resulting content is the total amount of NAD⁺ and NADH. The supernatant, after centrifugation, was heated at 60 °C for 30 min and subjected to subsequent operations. The resulting content is the total amount of NADH. For NAD⁺, the total content minus NADH was equal to NAD⁺. NADH/NAD⁺ was the ratio of NADH content to NAD⁺ content.

For the measurement of NADPH/NADP⁺, the method was similar to the above. 0.1 g roots were fully ground and extraction solution (Beyotime, S0179-4) was added and incubated at 4 C for 20 min. After centrifugation, G6PDH solution was added to the supernatant. After 37 °C for 10 min, chromogenic solution (Beyotime, S0179-2) was added. After 37 °C for 30 min in the dark, and the absorbance was measured at 450 nm. The resulting content is the total amount of NADP⁺ and NADPH. The supernatant, after centrifugation was heated at 60 °C for 30 min to subject to subsequent operations. The resulting content is the total amount of NADPH. For NADP⁺, the total content minus NADPH was equal to NADP⁺. NADPH/NADP⁺ was the ratio of NADPH content to NADP⁺ content.

For measurement of GSH/GSSG, according to the method provided by the supplier (Beyotime), GSSG stock solution, DTNB stock solution, protein removal reagent solution, NADPH stock solution, glutathione reductase solution, total glutathione detection working solution, GSH removal auxiliary solution and GSH removal working solution were prepared. For measurement of total glutathione, 0.1 g roots were fully ground with liquid nitrogen. The 1 ml protein removal reagent solution was added and incubated at 4 °C for 20 min. After centrifuging at 4 °C for 10 min, 150 μl total glutathione detection working solution was added to 10 μl the supernatant. After 25 °C for 5 min, 50 μl NADPH solution (0.5 mg/ml) was added. After 25 °C for 25 min, the absorbance was measured at 412 nm. For measurement of GSSG, 20 μl GSH removal auxiliary solution was added to the 100 μl supernatant. After fully mixed, 4 μl GSH removal working solution was added. Then, 150 μl total glutathione detection working solution was added. The next steps were consistent with the above process. The concentration of total glutathione was calculated according to the following formula: The concentration of total glutathione (μM) = (A₄₁₂− 0.2) ÷ 0.1 × 2; The concentration of GSSG (μM) = (A₄₁₂− 0.2) ÷ 0.1; The concentration of GSH (μM) = total glutathione (μM)- GSSG (μM) × 2. GSH/GSSG was the ratio of GSH content to GSSG content.

Agrobacterium rhizogenes mediated hairy root transformation for Arabidopsis

Arabidopsis seeds were placed on MS medium (pH 5.8–5.85) and the plates were transferred to a growth chamber at 23°C under a long-day (16-h light/8-h dark) photoperiod cycle. After the seedlings were four days old, the roots were completely removed and Agrobacterium rhizogenes (OD₆₀₀ = 0.3–0.5) containing the target vector was co-cultured with the root-depleted seedlings in the infection buffer (MS, 100 μM AS). After 15 min incubation on 28 °C shaker, Air-dried seedlings were placed on MS medium with timentin for growth. It was worth noting that the incision of the seedling was inserted into the medium. After the hairy roots were about 1 cm in length, seedlings were transferred to fresh MS medium for vertical cultivation. After four days of growth, ABA treatment was performed. According to the experimental requirements, hairy roots are used for subsequent experiments.

For amiR-GPT1 and amiR-GPT2, the DNA fragment was amplified by PCR using the RS300 plasmid as a template and primers (Supplementary Table 1) designed by the website (WMD3-Web MicroRNA Designer). After that, the DNA fragment was fused to the pMDC32 vector. All the constructs were transformed into the Agrobacterium rhizogenes. By the above method, Agrobacterium infects plant material (Col-0, stt3a-2, SnRK2.3^WT, SnRK2.3^N323A). After four days of growth, the seedlings were then transferred onto an agarose-solidified growth medium supplemented with 3 μM ABA or 1 μM ABA. After 7 days, the NADP⁺, NDAPH and H₂O₂ content was measured. The pMDC32 without exogenous DNA fragment served as a control.

For pGPT1::GPT1^WT-GFP (GPT1^WT), pGPT1::GPT1^S32A-GFP (GPT1^S32A) and pGPT1::GPT1^S32D-GFP (GPT1^S32D) construct, 1 kb upstream and 2.1 kb downstream sequences of the GTP1 coding region were cloned into pMDC107. We mutated the target site of amiR-GPT1-1 by primers (Supplementary Table 1) to construct pGPT1::GPT1^WT-GFP, and then further mutated the phosphorylation site (S32A,S32D) using pGPT1::GPT1^WT-GFP as a template by primers (Supplementary Table 1). All the constructs were transformed into the Agrobacterium rhizogenes. By the above method, two kinds of Agrobacterium (containing amiRGPT1-1 vector and GPT1^WT/GPT1^S32A/GPT1^S32D vector) were mixed in equal volumes to infect Col-0. After four days of growth, the seedlings were then transferred onto an agarose-solidified growth medium supplemented with 3 μM ABA. After 7 days, the NADP⁺, NDAPH and H₂O₂ content was measured.

For the pSnRK2.3::SnRK2.3^WT-GFP (SnRK2.3^WT-GFP) construct, 2020 bp upstream and 2033 bp downstream sequences of the SnRK2.3 coding region were cloned into pMDC107. For pSnRK2.3::SnRK2.3-NLS-GFP (SnRK2.3-NLS) construct, The NLS sequence (GGAGGAGTGGATCTAAAAGCACCTAAGAAGAAACGGAAGGTCGGC) was divided into two parts. 1–21 bp sequences were added to the 3’ primer of SnRK2.3 to form primer NLS-R-1. 4–42 bp sequences and the 3’ homologous sequence of pMDC107 formed primer NLS-R-2. The pSnRK2.3::SnRK2.3-NLS sequence was cloned into pMDC107 using pSnRK2.3::SnRK2.3^WT-GFP as a template by primers (Promoter-F, NLS-R-1 and NLS-R-2). For pSnRK2.3::SnRK2.3-GFP-mPTS^PEX26 (SnRK2.3-mPTS^PEX26) construct, pSnRK2.3::SnRK2.3 sequence was cloned using pSnRK2.3::SnRK2.3^WT-GFP as a template by primers (Promoter-F and SnRK2.3-g-R). GFP sequence was cloned using pSnRK2.3::SnRK2.3^WT-GFP as a template by primers (GFP-F and GFP-F-R). mPTS^PEX26 sequence was cloned using cDNA of Col-0 as a template by primers (PEX26-F and PEX26-R). pSnRK2.3::SnRK2.3, GFP and mPTS^PEX26 sequence were were fused to pMDC107 vector through homologous recombination. All PCR primers are listed in Supplementary Table 1. All the constructs were transformed into the Agrobacterium rhizogenes. By the above method, Agrobacterium infects snrk2.2/2.3. After four days of growth, the seedlings were then transferred onto an agarose-solidified growth medium supplemented with 1 μM ABA. After 7 days, the NADP⁺, NDAPH and H₂O₂ content was measured. The pMDC107 without exogenous DNA fragment served as a control.

Confocal laser scanning microscopy

Imaging data and micrographs of this study were done with an LSM880 confocal laser scanning microscope (Zeiss, Germany). GFP (eGFP, YFP) was excited with a 488 nm argon laser, and emission was collected with a 505–530 nm band pass filter. RFP (mCherry), PI (propidium iodide) and FM4-64 were excited with a 561 nm diode laser, and emission was detected with a 575–615 nm band pass filter. Fluorescence of MDC was excited with a 355 nm diode laser, and emission was detected with a 508 nm⁵⁷. DAPI was excited with a 364 nm argon laser, and emission was detected with a 440–470 nm. CFP was detected at 464–486 nmafter excitation a 458 nm laser beam⁵⁸. BES-H₂O₂-Ac was excited with a 495 nm argon laser, emission was collected with a 515–535 nm band pass filter, and HPF (Hydroxyphenyl fluorescein) was excited with a 492 nm argon laser, emission was collected with a 515 nm. DHE (Dihydroethidium) was excited with a 535 nm argon laser, emission was collected with a 610 nm. CD3-959 was used as ER marker⁵⁹, and CFP-HDEL was used as ER marker⁵⁸. SYP32-mCherry was used as a Golgi apparatus marker. dsRed-mPTS^PEX26 as peroxisome marker⁶⁰. For the dsRed-mPTS^PEX26 construct, the mPTS^PEX26 sequence (encoding the last 90 amino acids of PEX26) was cloned using cDNA of Col-0 as a template by primers (mPTS^PEX26-F and mPTS^PEX26-R, Supplementary Table 1) and fused to pMDC43 vector by using DNA ligases. Measurement of the fluorescence intensity was conducted with Fiji (ImageJ).

LC-MS/MS for identification Phosphorylation of site of GPT1

For mass spectrometry analysis, protein samples were separated using SDS-PAGE electrophoresis and detected by coomassie brilliant blue staining. The SDS-PAGE gel at 33 kD was collected into a clean centrifuge tube. 500 μl destain solution was added to a centrifuge tube and shaken at room temperature until the gel discoloured. After discarding the supernatant, 75% acetonitrile (200 μl) was added to the centrifuge tube and shaken at room temperature. After 30 min, 500 μl H₂O were added to the centrifuge tube and shaken at room temperature for 1 h. After discarding the supernatant, 50 mM NH₄HCO₃ (300 μl) were added to the centrifuge tube and shaken at room temperature for 5 min. After discarding the supernatant, 50 mM NH₄HCO₃ (20 μl) and 2 μl trypsin (1 μg/μl, Promega) were added to the centrifuge tube, and the gel was crushed. After centrifugation (100 g, 1 min), the sample was digested at 37 °C. After 8 min, 200 μl acetonitrile was added to the centrifuge tube and shaken at room temperature for 5 min. After centrifugation (100 × g, 1 min), the supernatant was transferred to a new centrifuge tube. This step was repeated once, ultimately resulting in 400 μl supernatant. The obtained supernatant was vacuum-dried at 60 °C and dissolved in 0.1% formic acid (100 μl). Samples were subjected to mass spectrometry detection by using a mass spectrometer (Orbitrap Fusion™ Lumos™ Tribrid™, Thermo Scientific).

The MS data were analysed using Proteome Discoverer 2.3 (Thermo Scientific). Proteins were identified by searching MS and MS/MS data of peptides against the GPT1(AT5G54800) amino acid sequence. In addition, phosphorylation was set as a variable modification (+ 79.966 Da). The minimal peptide length was specified to be 6 amino acids. The maximum peptide length was specified to be 144 amino acids. A maximum of two miscleavages was allowed. The precursor mass tolerance was set to 100 ppm, whereas fragment MS tolerance was set to 0.6 Da.

Transcriptome analysis

Total RNA was extracted from 7-day-old seedlings of different genotypes at 0, 1, 2, 3, 4, 6, 12, 24, 48, and 72 h after H₂O or 50 μM ABA treatment (spraying on the plants). For each time point, three biological replicates were prepared. The isolated RNAs were used to construct NGS libraries using the TruSeq RNA Library Prep Kit and sequenced in 150-nt paired-end mode with Illumina high-throughput sequencing platform NovaSeq 6000 (Novogene, Beijing, China). Raw high-throughput sequencing data have been deposited to the NCBI database under BioProject id: PRJNA940344.

RNA-seq reads were mapped to the Arabidopsis TAIR10 genome using Hisat2 (version 2.1.0)⁶¹ with parameters “--max-intronlen 15000 -k 1”. After quantified with featureCounts⁶², the expression level of each gene was calculated by the fragments per kilobase of exons per million fragments mapped (FPKM). Genes which expressed in at least one sample were kept and clustered by the fuzzy c-means algorithm of MFuzz package⁶³ based on their expression profiles across different samples. Then, Gene Ontology (GO) enrichment analysis was performed with the R package clusterProfiler⁶⁴ with an a p-value of 0.05.

To explore the correlation between every clusters with the previous experiments³⁶, we retrieved the RNA-seq data under accession GSE74864 (GSM1936759, GSM1936760, GSM1936761, GSM1936762, GSM1936763, GSM1936764, GSM1936765, GSM1936766, GSM1936767). After mapping and quantification as previously described, we used DESeq2⁶⁵ to characterise differentially expressed genes ( | log2 fold change | ≥ 1, P-value ≤ 0.05). Then the hypergeometric test was used to calculate the P-value between every cluster with the up- or down-regulated genes from previous experiments.

Statistics and reproducibility

Two-tailed paired t tests were performed when comparing two categories. Biological replicates and appropriate statistical tests for each data were indicated in the figure legends and Source Data file. Significance indicated as *P-value < 0.05, **P-value < 0.01, ***P-value < 0.001 and “ns” means no significant difference.

Reporting summary

Further information on research design is available in the Nature Portfolio Reporting Summary linked to this article.

Data availability

Sequence data for the genes described in this article can be found in the Arabidopsis Information Resource (https://www.arabidopsis.org) under the following accession numbers: STT3A (AT5G19690), STT3B (AT1G34130), GIIβ (AT5G56360), GIIα (AT5G63840), FUT11 (AT3G19280), CGL1 (AT4G38240), GMII (AT5G14950), PYR1 (AT4G17870), PYL1 (AT5G46790), PYL2 (AT2G26040), PYL3 (AT1G73000), PYL4 (AT2G38310), PYL5 (AT5G05440), PYL6 (AT2G40330), PYL7 (AT4G01026), PYL8 (AT5G53160), PYL9 (AT1G01360), PYL10 (AT4G27920), PYL11 (AT5G45860), PYL12 (AT5G45870), PYL13 (AT4G18620), ABI1 (AT4G26080), ABI2 (AT5G57050), ABF2 (AT1G45249), ABI5 (AT2G36270), SnRK2.2 (AT3G50500), SnRK2.3 (AT5G66880), SnRK2.6 (AT4G33950), P5CS1 (AT2G39800), RAB18 (AT5G66400), RD29B (AT5G52300), NCED3 (AT3G14440), EF1-α (AT1G18070), ACT2 (AT3G18780), PEX6 (AT1G03000), APX3 (AT4G35000), CAT2 (AT4G35090), MDAR1 (AT3G52880), MDAR4 (AT3G27820), RBOHD (AT5G47910), RBOHF (AT1G64060), NADK3 (AT1G78590), ICDH (AT1G54340), PXN (AT2G39970), G6PD1 (AT5G35790), PGD2 (AT3G02360), GPT1 (AT5G54800), GPT2 (AT1G61800), PEX26 (AT3G10572). The structural models of SnRK2.2/2.3 are generated by AlphaFold2 (www.alphafold. ebi.ac.uk). The time-course RNA-seq datasets and mass spectrometry data generated in this study have been deposited in the NCBI Gene Expression Omnibus (GEO) under accession PRJNA940344, and ProteomeXchange Consortium with identifier PXD053098, respectively. The previously published RNA-seq data used in this study are available in the GEO under accession GSE74864. Source data are provided with this paper.

References

Zhu, J.-K. Abiotic stress signaling and responses in plants. Cell 167, 313–324 (2016).
Article CAS PubMed PubMed Central Google Scholar
Gupta, A., Rico-Medina, A. & Cano-Delgado, A. I. The physiology of plant responses to drought. Science 368, 266–269 (2020).
Article ADS CAS PubMed Google Scholar
Weng, J. K., Ye, M., Li, B. & Noel, J. P. Co-evolution of hormone metabolism and signaling networks expands plant adaptive plasticity. Cell 166, 881–893 (2016).
Article CAS PubMed Google Scholar
Hauser, F., Li, Z., Waadt, R. & Schroeder, J. I. SnapShot: abscisic acid signaling. Cell 171, 1708–1708.e1700 (2017).
Article CAS PubMed PubMed Central Google Scholar
Park, S. Y. et al. Abscisic acid inhibits type 2C protein phosphatases via the PYR/PYL family of START proteins. Science 324, 1068–1071 (2009).
Article ADS CAS PubMed PubMed Central Google Scholar
Ma, Y. et al. Regulators of PP2C phosphatase activity function as abscisic acid sensors. Science 324, 1064–1068 (2009).
Article ADS CAS PubMed Google Scholar
Fujita, Y. et al. Three SnRK2 protein kinases are the main positive regulators of abscisic acid signaling in response to water stress in Arabidopsis. Plant Cell Physiol. 50, 2123–2132 (2009).
Article CAS PubMed Google Scholar
Fujii, H. & Zhu, J. K. Arabidopsis mutant deficient in 3 abscisic acid-activated protein kinases reveals critical roles in growth, reproduction, and stress. Proc. Natl. Acad. Sci. USA 106, 8380–8385 (2009).
Article ADS CAS PubMed PubMed Central Google Scholar
Raghavendra, A. S., Gonugunta, V. K., Christmann, A. & Grill, E. ABA perception and signalling. Trends Plant Sci. 15, 395–401 (2010).
Article CAS PubMed Google Scholar
Yoshida, T. et al. AREB1, AREB2, and ABF3 are master transcription factors that cooperatively regulate ABRE-dependent ABA signaling involved in drought stress tolerance and require ABA for full activation. Plant J. 61, 672–685 (2010).
Article CAS PubMed Google Scholar
Geiger, D. et al. Activity of guard cell anion channel SLAC1 is controlled by drought-stress signaling kinase-phosphatase pair. Proc. Natl. Acad. Sci. USA 106, 21425–21430 (2009).
Article ADS CAS PubMed PubMed Central Google Scholar
Kwak, J. M. et al. NADPH oxidase AtrbohD and AtrbohF genes function in ROS-dependent ABA signaling in Arabidopsis. EMBO J. 22, 2623–2633 (2003).
Article CAS PubMed PubMed Central Google Scholar
Sirichandra, C. et al. Phosphorylation of the Arabidopsis AtrbohF NADPH oxidase by OST1 protein kinase. FEBS Lett. 583, 2982–2986 (2009).
Article CAS PubMed Google Scholar
Miyazono, K. et al. Structural basis of abscisic acid signalling. Nature 462, 609–614 (2009).
Article ADS CAS PubMed Google Scholar
Fujii, H. et al. In vitro reconstitution of an abscisic acid signalling pathway. Nature 462, 660–664 (2009).
Article ADS CAS PubMed PubMed Central Google Scholar
Yang, G., Yu, Z., Gao, L. & Zheng, C. SnRK2s at the crossroads of growth and stress responses. Trends Plant Sci. 24, 672–676 (2019).
Article CAS PubMed Google Scholar
Lin, Z. et al. Initiation and amplification of SnRK2 activation in abscisic acid signaling. Nat. Commun. 12, 2456 (2021).
Article ADS CAS PubMed PubMed Central Google Scholar
Sibley, D. R. & Lefkowitz, R. J. Molecular mechanisms of receptor desensitization using the beta-adrenergic receptor-coupled adenylate cyclase system as a model. Nature 317, 124–129 (1985).
Article ADS CAS PubMed Google Scholar
Lemmon, M. A., Freed, D. M., Schlessinger, J. & Kiyatkin, A. The dark side of cell signaling: positive roles for negative regulators. Cell 164, 1172–1184 (2016).
Article CAS PubMed PubMed Central Google Scholar
Larosa, P. C., Handa, A. K., Hasegawa, P. M. & Bressan, R. A. Abscisic Acid accelerates adaptation of cultured tobacco cells to salt. Plant Physiol. 79, 138–142 (1985).
Article CAS PubMed PubMed Central Google Scholar
Peng, Z.-Y. & Weyers, J. D. B. Stomatal sensitivity to abscisic acid following water deficit stress. J. Exp. Bot. 45, 835–845 (1994).
Article CAS Google Scholar
Hebert, D. N., Lamriben, L., Powers, E. T. & Kelly, J. W. The intrinsic and extrinsic effects of N-linked glycans on glycoproteostasis. Nat. Chem. Biol. 10, 902–910 (2014).
Article CAS PubMed PubMed Central Google Scholar
Kowarik, M. et al. N-linked glycosylation of folded proteins by the bacterial oligosaccharyltransferase. Science 314, 1148–1150 (2006).
Article ADS CAS PubMed Google Scholar
Strasser, R. Plant protein glycosylation. Glycobiology 26, 926–939 (2016).
Article CAS PubMed PubMed Central Google Scholar
Nagashima, Y., von Schaewen, A. & Koiwa, H. Function of N-glycosylation in plants. Plant Sci. 274, 70–79 (2018).
Article CAS PubMed Google Scholar
Burn, J. E. et al. The cellulose-deficient Arabidopsis mutant rsw3 is defective in a gene encoding a putative glucosidase II, an enzyme processing N-glycans during ER quality control. Plant J. 32, 949–960 (2002).
Article CAS PubMed Google Scholar
Koiwa, H. et al. The STT3a subunit isoform of the Arabidopsis oligosaccharyltransferase controls adaptive responses to salt/osmotic stress. Plant Cell 15, 2273–2284 (2003).
Article CAS PubMed PubMed Central Google Scholar
Kang, J. S. et al. Salt tolerance of Arabidopsis thaliana requires maturation of N-glycosylated proteins in the Golgi apparatus. Proc. Natl. Acad. Sci. USA 105, 5933–5938 (2008).
Article ADS CAS PubMed PubMed Central Google Scholar
Frank, J., Kaulfurst-Soboll, H., Rips, S., Koiwa, H. & von Schaewen, A. Comparative analyses of Arabidopsis complex glycan1 mutants and genetic interaction with staurosporin and temperature sensitive3a. Plant Physiol. 148, 1354–1367 (2008).
Article CAS PubMed PubMed Central Google Scholar
Strasser, R. et al. Molecular cloning and characterization of Arabidopsis thaliana Golgi alpha-mannosidase II, a key enzyme in the formation of complex N-glycans in plants. Plant J. 45, 789–803 (2006).
Article CAS PubMed Google Scholar
Liu, C. et al. Trimming of N-Glycans by the Golgi-Localized alpha-1,2-Mannosidases, MNS1 and MNS2, Is Crucial for Maintaining RSW2 Protein Abundance during Salt Stress in Arabidopsis. Mol. Plant 11, 678–690 (2018).
Article CAS PubMed Google Scholar
Strasser, R., Altmann, F., Mach, L., Glossl, J. & Steinkellner, H. Generation of Arabidopsis thaliana plants with complex N-glycans lacking beta1,2-linked xylose and core alpha1,3-linked fucose. FEBS Lett. 561, 132–136 (2004).
Article CAS PubMed Google Scholar
Zielinska, D. F., Gnad, F., Schropp, K., Wisniewski, J. R. & Mann, M. Mapping N-glycosylation sites across seven evolutionarily distant species reveals a divergent substrate proteome despite a common core machinery. Mol. Cell 46, 542–548 (2012).
Article CAS PubMed Google Scholar
Shinozawa, A. et al. SnRK2 protein kinases represent an ancient system in plants for adaptation to a terrestrial environment. Commun. Biol. 2, 30 (2019).
Article PubMed PubMed Central Google Scholar
Ikeuchi, M. et al. Wounding triggers callus formation via dynamic hormonal and transcriptional changes. Plant Physiol. 175, 1158–1174 (2017).
Article CAS PubMed PubMed Central Google Scholar
Sun, H. X., Li, Y., Niu, Q. W. & Chua, N. H. Dehydration stress extends mRNA 3’ untranslated regions with noncoding RNA functions in Arabidopsis. Genome Res. 27, 1427–1436 (2017).
Article CAS PubMed PubMed Central Google Scholar
Soma, F. et al. ABA-unresponsive SnRK2 protein kinases regulate mRNA decay under osmotic stress in plants. Nat. Plants 3, 16204 (2017).
Article CAS PubMed Google Scholar
Shen, J., Ding, Y., Gao, C., Rojo, E. & Jiang, L. N-linked glycosylation of AtVSR1 is important for vacuolar protein sorting in Arabidopsis. Plant J. 80, 977–992 (2014).
Article CAS PubMed Google Scholar
Maley, F., Trimble, R. B., Tarentino, A. L. & Plummer, T. H. Jr. Characterization of glycoproteins and their associated oligosaccharides through the use of endoglycosidases. Anal. Biochem. 180, 195–204 (1989).
Article CAS PubMed Google Scholar
Matsui, T. et al. N-glycosylation at noncanonical Asn-X-Cys sequences in plant cells. Glycobiology 21, 994–999 (2011).
Article CAS PubMed Google Scholar
Yoshida, R. et al. The regulatory domain of SRK2E/OST1/SnRK2.6 interacts with ABI1 and integrates abscisic acid (ABA) and osmotic stress signals controlling stomatal closure in Arabidopsis. J. Biol. Chem. 281, 5310–5318 (2006).
Article CAS PubMed Google Scholar
Ali, A. et al. Rheostatic control of ABA signaling through HOS15-mediated OST1 degradation. Mol. Plant 12, 1447–1462 (2019).
Article CAS PubMed Google Scholar
Pan, R., Liu, J., Wang, S. & Hu, J. Peroxisomes: versatile organelles with diverse roles in plants. N. Phytol. 225, 1410–1427 (2020).
Article Google Scholar
Pei, Z. M. et al. Calcium channels activated by hydrogen peroxide mediate abscisic acid signalling in guard cells. Nature 406, 731–734 (2000).
Article ADS CAS PubMed Google Scholar
Hung, K. T. & Kao, C. H. Hydrogen peroxide is necessary for abscisic acid-induced senescence of rice leaves. J. Plant Physiol. 161, 1347–1357 (2004).
Article CAS PubMed Google Scholar
Yang, L. et al. ABA-mediated ROS in mitochondria regulate root meristem activity by controlling PLETHORA expression in Arabidopsis. PLoS Genet. 10, e1004791 (2014).
Article PubMed PubMed Central Google Scholar
Baune, M. C. et al. The Arabidopsis plastidial glucose-6-phosphate transporter GPT1 is dually targeted to peroxisomes via the endoplasmic reticulum. Plant Cell 32, 1703–1726 (2020).
Article CAS PubMed PubMed Central Google Scholar
Niewiadomski, P. et al. The Arabidopsis plastidic glucose 6-phosphate/phosphate translocator GPT1 is essential for pollen maturation and embryo sac development. Plant Cell 17, 760–775 (2005).
Article CAS PubMed PubMed Central Google Scholar
Kruger, N. J. & von Schaewen, A. The oxidative pentose phosphate pathway: structure and organisation. Curr. Opin. Plant Biol. 6, 236–246 (2003).
Article CAS PubMed Google Scholar
Baxter, C. J. et al. The metabolic response of heterotrophic Arabidopsis cells to oxidative stress. Plant Physiol. 143, 312–325 (2007).
Article CAS PubMed PubMed Central Google Scholar
Hu, J. et al. Plant peroxisomes: biogenesis and function. Plant Cell 24, 2279–2303 (2012).
Article CAS PubMed PubMed Central Google Scholar
He, A., Dean, J. M. & Lodhi, I. J. Peroxisomes as cellular adaptors to metabolic and environmental stress. Trends Cell Biol. 31, 656–670 (2021).
Article CAS PubMed PubMed Central Google Scholar
von Numers, N. et al. Requirement of a homolog of glucosidase II beta-subunit for EFR-mediated defense signaling in Arabidopsis thaliana. Mol. Plant 3, 740–750 (2010).
Article Google Scholar
Li, H. et al. The plant ESCRT component FREE1 shuttles to the nucleus to attenuate abscisic acid signalling. Nat. Plants 5, 512–524 (2019).
Article CAS PubMed Google Scholar
Cai, Z. et al. GSK3-like kinases positively modulate abscisic acid signaling through phosphorylating subgroup III SnRK2s in Arabidopsis. Proc. Natl. Acad. Sci. USA 111, 9651–9656 (2014).
Article ADS CAS PubMed PubMed Central Google Scholar
Tsukagoshi, H., Busch, W. & Benfey, P. N. Transcriptional regulation of ROS controls transition from proliferation to differentiation in the root. Cell 143, 606–616 (2010).
Article CAS PubMed Google Scholar
Contento, A. L., Xiong, Y. & Bassham, D. C. Visualization of autophagy in Arabidopsis using the fluorescent dye monodansylcadaverine and a GFP-AtATG8e fusion protein. Plant J. 42, 598–608 (2005).
Article CAS PubMed Google Scholar
Lerich, A. et al. ER Import sites and their relationship to ER exit sites: A new model for bidirectional ER-Golgi transport in higher plants. Front. Plant Sci. 3, https://doi.org/10.3389/fpls.2012.00143 (2012).
Nelson, B. K., Cai, X. & Nebenfuhr, A. A multicolored set of in vivo organelle markers for co-localization studies in Arabidopsis and other plants. Plant J. 51, 1126–1136 (2007).
Article CAS PubMed Google Scholar
Wright, Z. J. & Bartel, B. Peroxisomes form intralumenal vesicles with roles in fatty acid catabolism and protein compartmentalization in Arabidopsis. Nat. Commun. 11, 6221 (2020).
Article ADS CAS PubMed PubMed Central Google Scholar
Kim, D., Landmead, B. & Salzberg, S. L. HISAT: a fast spliced aligner with low memory requirements. Nat. Methods 12, 357–U121 (2015).
Article CAS PubMed PubMed Central Google Scholar
Liao, Y., Smyth, G. K. & Shi, W. featureCounts: an efficient general purpose program for assigning sequence reads to genomic features. Bioinformatics 30, 923–930 (2014).
Article CAS PubMed Google Scholar
Kumar, L. & E Futschik, M. Mfuzz: a software package for soft clustering of microarray data. Bioinformation 2, 5–7 (2007).
Article PubMed PubMed Central Google Scholar
Yu, G., Wang, L. G., Han, Y. & He, Q. Y. clusterProfiler: an R package for comparing biological themes among gene clusters. Omics A J. Integr. Biol. 16, 284–287 (2012).
Article CAS Google Scholar
Love, M. I., Huber, W. & Anders, S. Moderated estimation of fold change and dispersion for RNA-seq data with DESeq2. Genome Biol. 15, 550 (2014).
Article PubMed PubMed Central Google Scholar

Download references

Acknowledgements

We thank Z. Gong (China Agricultural University), M. Bai (Shandong University), J. Xu (Zhejiang University), M. Zhang (Nanjing Agriculture University) and J. Li (Northwest A&F University) for providing seeds; G. Yang (China Agricultural University), B. Liu (Chinese Academy of Agricultural Sciences), and K. Zhang (Zhejiang Normal University) for manuscript discussion. This work was supported by the National Natural Science Foundation of China (31970292 and 32170306 to C.Z., and 32241039 to C.W.), the Shandong Agricultural Breeding Project (2017LZN025 to G.Y.), and Shandong Natural Science Foundation (ZR2021MC012 to G.Y.).

Author information

Authors and Affiliations

College of Life Sciences, Shandong Agricultural University, Taian, Shandong, PR China
Junyao Lu, Ning Li, Gaojian Li, Ziang Tian, Lianping Shi, Yan Wang, Yingao Cai, Kaiyuan Zhang, Wanting Sun, Danyang Wang, Jinxin Lin, Jinguang Huang, Changai Wu, Kang Yan, Shizhong Zhang, Chengchao Zheng & Guodong Yang
National Key Laboratory of Wheat Improvement, Shandong Agricultural University, Taian, Shandong, PR China
Junyao Lu, Ning Li, Gaojian Li, Ziang Tian, Lianping Shi, Yan Wang, Yingao Cai, Kaiyuan Zhang, Wanting Sun, Danyang Wang, Jinxin Lin, Jinguang Huang, Changai Wu, Kang Yan, Shizhong Zhang, Chengchao Zheng & Guodong Yang

Authors

Junyao Lu
View author publications
Search author on:PubMed Google Scholar
Ning Li
View author publications
Search author on:PubMed Google Scholar
Gaojian Li
View author publications
Search author on:PubMed Google Scholar
Ziang Tian
View author publications
Search author on:PubMed Google Scholar
Lianping Shi
View author publications
Search author on:PubMed Google Scholar
Yan Wang
View author publications
Search author on:PubMed Google Scholar
Yingao Cai
View author publications
Search author on:PubMed Google Scholar
Kaiyuan Zhang
View author publications
Search author on:PubMed Google Scholar
Wanting Sun
View author publications
Search author on:PubMed Google Scholar
Danyang Wang
View author publications
Search author on:PubMed Google Scholar
Jinxin Lin
View author publications
Search author on:PubMed Google Scholar
Jinguang Huang
View author publications
Search author on:PubMed Google Scholar
Changai Wu
View author publications
Search author on:PubMed Google Scholar
Kang Yan
View author publications
Search author on:PubMed Google Scholar
Shizhong Zhang
View author publications
Search author on:PubMed Google Scholar
Chengchao Zheng
View author publications
Search author on:PubMed Google Scholar
Guodong Yang
View author publications
Search author on:PubMed Google Scholar

Contributions

G.Y., C.Z. and J.L. designed the study; J.L., N.L, G.L. and Z.T. performed the experiments; J.L. N.L. and G.Y. made the figures; J.L., N.L, G.L., L.S., Y.W, Z.T., Y.C., K.Z., W.S., D.W. and J.Lin analysed the data. G.Y. wrote the manuscript, J.L. and N.L. wrote the method; J.H., C.W., K.Y., S.Z. and C.Z. discussed and reviewed the paper.

Corresponding authors

Correspondence to Chengchao Zheng or Guodong Yang.

Ethics declarations

Competing interests

The authors declare no competing interests.

Peer review

Peer review information

Nature Communications thanks Harvey Millar and Pengcheng Wang, and the other anonymous, reviewers for their contribution to the peer review of this work. A peer review file is available.

Additional information

Publisher’s note Springer Nature remains neutral with regard to jurisdictional claims in published maps and institutional affiliations.

Supplementary information

Supplementary Information

Peer Review File

Reporting summary

Source data

Source Data

Rights and permissions

Open Access This article is licensed under a Creative Commons Attribution-NonCommercial-NoDerivatives 4.0 International License, which permits any non-commercial use, sharing, distribution and reproduction in any medium or format, as long as you give appropriate credit to the original author(s) and the source, provide a link to the Creative Commons licence, and indicate if you modified the licensed material. You do not have permission under this licence to share adapted material derived from this article or parts of it. The images or other third party material in this article are included in the article’s Creative Commons licence, unless indicated otherwise in a credit line to the material. If material is not included in the article’s Creative Commons licence and your intended use is not permitted by statutory regulation or exceeds the permitted use, you will need to obtain permission directly from the copyright holder. To view a copy of this licence, visithttp://creativecommons.org/licenses/by-nc-nd/4.0/.

Reprints and permissions

About this article

Cite this article

Lu, J., Li, N., Li, G. et al. N-glycosylation of SnRK2s affects NADPH maintenance in peroxisomes during prolonged ABA signalling. Nat Commun 15, 6630 (2024). https://doi.org/10.1038/s41467-024-50720-3

Download citation

Received: 19 May 2023
Accepted: 19 July 2024
Published: 05 August 2024
DOI: https://doi.org/10.1038/s41467-024-50720-3

Subjects

Abstract

Similar content being viewed by others

Introduction

Results

NPEs promote desensitisation of the transcriptional ABA response

SnRK2.2 and SnRK2.3 are N-glycosylated

SnRK2s are epistatic to NPEs in ABA signalling

SnRK2.2/2.3 redistribute to the peroxisome membrane

N-glycosylation of SnRK2s modulates H2O2 accumulation

The SnRK2-GPT1 module regulates NADPH maintenance

Discussion

Methods

Plant materials, growth conditions

Quantitative real-time PCR

Detached leaf senescence and measurement of chlorophyll content

Yeast two-hybrid analysis

Co-immunoprecipitation analysis

Firefly luciferase complementation imaging (LCI) assays

Bimolecular fluorescence complementation (BiFC) analysis

GST pull-down assay

In vitro kinase assay

N-glycosylation analysis of SnRK2s

Measurement of ROS in plants

Measurement of NADH/NAD+, NADPH/NADP+ and GSH/GSSG

Agrobacterium rhizogenes mediated hairy root transformation for Arabidopsis

Confocal laser scanning microscopy

LC-MS/MS for identification Phosphorylation of site of GPT1

Transcriptome analysis

Statistics and reproducibility

Reporting summary

Data availability

References

Acknowledgements

Author information

Authors and Affiliations

Contributions

Corresponding authors

Ethics declarations

Competing interests

Peer review

Peer review information

Additional information

Supplementary information

Source data

Rights and permissions

About this article

Cite this article

Share this article

Search

Quick links

N-glycosylation of SnRK2s modulates H₂O₂ accumulation

Measurement of NADH/NAD⁺, NADPH/NADP⁺ and GSH/GSSG