The rice plants used in this study were Oryza sativa japonica (Jap). The Arabidopsis thaliana plants used were wild-type Columbia (Col).

To create 35S promoter-driven Myc-tagged OsWAK112 or OsWAK112^K678E or OsWAK112^D794A constructs, the OsWAK112 coding sequence was amplified by PCR with the primers OsWAK112-F and OsWAK112-R (Supplementary Table 1) to construct OsWAK112/pENTR using the commercial donor vector pENTR™/SD/D-TOPO™ (K242020; Invitrogen, Carlsbad, CA, United States). Next, OsWAK112^K678E and OsWAK112^D794A pENTR were generated using the Fast Mutagenesis System (FM111; TransGen Biotech, Beijing, China) with OsWAK112-K678E-F and OsWAK112-K678E-R or OsWAK112-D794A-F and OsWAK112-D794A-R as the overlapping mutagenic primers (Supplementary Table 1) and OsWAK112/pENTR as the template. Sequences encoding both native and mutated versions of OsWAK112 were cloned into the binary vector pG7MH1 (Wang et al., 2013) using the Gateway LR Clonase™ II Enzyme Mix (11,791,020; Invitrogen).

To generate OsWAK112 RNAi transgenic plants, a 581-bp OsWAK112 fragment was amplified with specific primers OsWAK112RNAi-F and OsWAK112RNAi-R (Supplementary Table 1) using the LA Taq with GC Buffer kit (Takara, DRR20AG). The OsWAK112 RNAi fragment was inserted into the RNAi vector pTCK303 (Wang et al., 2004) through its KpnI/SpeI and BamHI/SacI sites.

For the native promoter-driven GUS construct proOsWAK112::GUS, the promoter sequence of OsWAK112 (with 2,108 base pairs upstream of the ATG) was amplified and inserted into the binary vector pCAMBI1300::GUS using HindIII and BamHI.

For our bimolecular fluorescence complementation (BiFC) assays, the coding sequences of OsSAMS1/2/3 were amplified by PCR using the primers OsSAMS1/2/3-F and OsSAMS1/2/3-R (Supplementary Table 1) to construct OsSAMS1/2/3/pENTR. Then, the coding sequences of OsWAK112 and OsSAMS1/2/3 were individually introduced into 35S::X-NYFP and 35S::X-CCFP (Wang et al., 2013) using OsWAK112/pENTR and OsSAMS1/2/3/pENTR as templates.

For the co-immunoprecipitation (Co-IP) assays, the coding sequences of OsSAMS1, OsSAMS2, OsSAMS3, and the OsWAK112 kinase domain (OsWAK112KD) were cloned into pMDC83 and pG7MH1 to produce ectopic expression of OsSAMS1/2/3-GFP and OsWAK112KD-Myc, respectively.

The above-mentioned binary expression constructs were introduced into Agrobacterium tumefaciens strain EHA105 and then transformed into Arabidopsis or rice for stable transformation or co-infiltrated into 4-week-old Nicotiana benthamiana leaves for transient assays.

For prokaryotic protein expression, the coding sequence of the OsWAK112KD was amplified by PCR using the primers OsWAK112-KD-F and OsWAK112-KD-R (Supplementary Table 1) and inserted into pGEX-4T-1 to produce glutathione S-transferase (GST)-OsWAK112KD. Next, GST-OsWAK112^K678E, GST-OsWAK112^K791R, and GST-OsWAK112^D794A were generated using the Fast Mutagenesis System (FM111; TransGen Biotech) with OsWAK112-K678E-F/R, OsWAK112-K791R-F/R, or OsWAK112-D794A-F/R as the overlapping mutagenic primers (Supplementary Table 1) and GST-OsWAK112KD as the template. For myelin basic protein (MBP)-OsSAMS1/2/3, the coding sequences of OsSAMS1/2/3 were cloned into the destination vector gc-PML2C (Xin et al., 2021) using OsSAMS1/2/3/pENTR as the donor vector. The above-mentioned constructs were transformed into Escherichia coli strain BL21 for protein purification.

All rice plants were grown in a greenhouse with a 12-h/12-h light-dark cycle at 28°C (100μmolm⁻² s⁻¹ intensity) or in a paddy field under natural conditions (May to October in Hebei, China) for general growth and seed production. The Arabidopsis seedlings were grown either in a growth chamber under continuous light (100μmolm⁻² s⁻¹ intensity) or in a culture room under 16h of light/8h of darkness (90μmolm⁻² s⁻¹ intensity) at 22°C for seed production.

For salt sensitivity analyses of the rice in hydroponic culture, seeds were germinated in water for 2days, grown in a hydroponic culture solution (Ren et al., 2005) for 10days, and treated with a 200mM NaCl-containing culture solution for another 4days. After 8days of recovery, the survive rate was determined. To assess the salt response in solid medium, seeds were germinated on 0.5×Murashige and Skoog (MS) solid medium for 2days then transplanted to 0.5×MS solid medium without (0mM) or with 150mM or 200mM NaCl for 20days at 28°C in a light chamber under a 14-h/10-h light-dark cycle.

Arabidopsis seeds were sown on 0.5×MS medium with or without 125mM NaCl, kept for 2days in the dark at 4°C, grown under 16h of light/8h of darkness at 22°C for 10days, and then observed to determine the phenotype.

Seedlings that were still green and continuing to produce new leaves were registered as survivors. All experiments were repeated at least three times.

Total RNA from 7-day-old rice seedlings was isolated using TRIzol reagent (15596–026; Invitrogen) following the manufacturer’s instructions. cDNA was synthesized from 0.5 to 1μg of total RNA with a RevertAid First Strand cDNA Synthesis Kit (K1622; Thermo Fisher Scientific, Waltham, MA, United States) using oligo (dT)₁₆ primer. To analyze OsWAK112, quantitative semi-quantitative reverse transcription (RT-qPCR) was performed using the SYBR Premix Ex Taq Reagent (RR420A; Takara, Otsu, Japan); the Os18S rRNA gene was used as an internal control. The primers used are listed in Supplementary Table 1.

The roots of 7-day-old rice plants treated with or without 150mM NaCl for 30min were stained with 5-(and-6)-chloromethyl-2ʹ,7ʹ-dichlorodihydrofluorescein diacetate, acetyl ester (CM-H₂DCFDA; Thermo Fisher Scientific). The fluorescence intensity was determined with an Axio Imager (M2; Carl Zeiss, Jena, Germany); all pictures were analyzed using ImageJ. The fluorescent intensity of Jap at 0mM NaCl is set to 1.

To investigate the subcellular localization of OsWAK112, the coding sequence of OsWAK112 from OsWAK112/pENTR was introduced into binary vector pMDC83 to construct the vector 35S:: OsWAK112-GFP, and 35S::GFP was used as a negative control vector. Both constructs were transferred to A. tumefaciens strain GV3101 and then transiently transformed into tobacco (N. benthamiana) leaves or transformed into onion (Allium cepa) epidermal cells by particle bombardment using the Bio-Rad PDS-1000/He system according to the manufacturer’s protocol. After 40 to 48h infiltration, localization of the protein was examined using a Zeiss LSM 510 confocal microscopy system with a 488nm laser for excitation from 500 to 515nm for GFP emission. 0.9M mannitol was used to induce plasmolysis.

Recombinant GST-OsWAK112KD, GST-WAK112KD^K678E, GST-WAK112KD^K791R, and GST-WAK112KD^D794A were affinity purified using glutathione agarose beads (GE Healthcare, Chicago, IL, United States) and subjected to an in vitro kinase assay, which included incubation with 1μL of cold ATP (1mM) in kinase reaction buffer (50mM HEPES, 10mM MgCl₂, 5mM MnCl₂, and 1mM ATP, 10μCi of [γ-32P]ATP) at 30°C for 3h. The mixture was separated by 10% SDS-PAGE and detected with a Typhoon 9,410 imager (GE Healthcare).

For our BiFC assays, OsWAK112-NYFP and OsSAMSs-CCFP or YCE (control) were co-expressed in N. benthamiana leaves. Fluorescence from the BiFC signals was observed under a confocal microscope (Meta710-LCM; Carl Zeiss).

For the overlay assays, gel blots (polyvinylidene fluoride [PVDF] filters) containing GST-WAK112KD and GST were incubated with 2μg of MBP-OsSAMS1, MBP-OsSAMS2, and MBP-OsSAMS3, followed by anti-MBP-HRP (E8038S; New England BioLabs, Ipswich, MA, United States).

In our Co-IP assays, N. benthamiana leaves transiently co-expressing OsSAMSs-GFP and OsWAK112KD-Myc, or expressing OsSAMSs-GFP or OsWAK112KD-Myc alone, were harvested and extracted with IP buffer (50mM Tris-HCl, pH 7.5, 150mM NaCl, 2mM EDTA, and 0.1% NP-40) and then incubated with anti-GFP agarose at 4°C for 2h. Next, the proteins were washed with IP buffer five times and then boiled at 100°C in water for 5min with 2×loading buffer. The proteins were then separated by SDS-PAGE. Anti-Myc (019M4760V; Sigma-Aldrich, St. Louis, MO, United States) and anti-GFP antibodies (HT801-01; TransGen Biotech) were used to detect OsWAK112KD and OsSAMS1/2/3, respectively.

For ethylene measurement, 7-day-old wild-type (Jap) and OsWAK112-overexpressing (OsWAK112OE) rice seedlings treated with or without 200mM NaCl for 24h were placed into 15-mL vials containing 3ml of 0.5×MS medium and closed with a rubber cap on the top; the vials were then allowed to sit for 24h (16-h light/8-h dark) at 28°C. A total of 1mL of the air from the headspace of each vial was taken to inject into a gas chromatograph to measure the ethylene content.

For SAMS measurement, total protein from 7-day-old Jap and OsWAK112OE seedlings treated with or without 200mM NaCl for 24h was extracted and used for SAMS content measurement with a SAMS ELISA Kit (SY-P09712; Shanghai Win-Win Biotechnology Co., Ltd., Shanghai, China). The assay was performed according the manufacturer’s protocol.

To test whether OsWAK112 affects OsSAMSs protein stability, OsSAMS1-GFP alone or OsSAMS1 together with OsWAK112-Myc was filtrated in tobacco leave. After 40 to 48h, tobacco leave was separated into two parts, treated in liquid 1/2 MS medium with or without 200mM NaCl for 2h, and then, the total protein was extracted for Western blotting.

In this study, we examined the involvement of OsWAK112 in plant salt responses. OsWAK112 bound directly to and inhibited OsSAMSs to suppress the induction of ethylene production so as to negatively regulate salt tolerance.

WAKs are protein kinases with important roles not only in plant growth and development but also in plant responses to exogenous stresses (Hou et al., 2005; Bot et al., 2019; Meco et al., 2020); however, their role in salt stress responses is poorly understood. Published microarray data revealed that the expression levels of several WAKs are regulated by salt (Mishra et al., 2018). Herein, the expression pattern of one of them, OsWAK112 (Os10g10130; Mishra et al., 2018), was examined under high-salt conditions. Its transcript level decreased gradually as the treatment time increased; at 24h, it was reduced to nearly half of the control level (Supplementary Figure S1), indicating that OsWAK112 is a potential salt-related regulator in rice.

To determine the biological function of OsWAK112 in salt responses, ectopic expression lines using a 35S promoter-driven OsWAK112 coding sequence with a Myc tag were generated in a wild-type Jap background. Two independent OsWAK112OE lines named OE7 and OE8 were used in our subsequent experiments because they exhibited high transcript levels (approximately 2.6- or 3.2-fold that in Jap; Figure 1A). Under normal growth conditions, the OsWAK112OE plants exhibited no clear difference from Jap plants (Figure 1B, before treatment). However, after being treated with 200mM NaCl for 4days and subsequently allowed to recover in regular hydroponic culture for another 8days, the OsWAK112OE plants were drier and yellower than the Jap plants (Figure 1B), in accordance with their lower survival rates (50% for OE7 and 43% for OE8) compared to Jap (63%) (Figure 1C), indicating that OsWAK112 induces poor salt resistance in rice.

To confirm the function of OsWAK112 in plant salt tolerance, the salt responses of OE7 and OE8 plants were examined on solid medium. After 2 days of germination on regular medium, seeds were transplanted into medium containing 0, 150, or 200mM NaCl. The OE7 and OE8 seedlings did not show any difference from Jap seedlings when grown on regular medium (Figure 2A, 0mM NaCl); however, they showed greater growth inhibition by salt than Jap plants depending on the NaCl concentration (Figure 2A). The shoot length and shoot fresh weight of OE7 and OE8 were both lower than those of Jap (Figures 2B,C).

To completely investigate the function of OsWAK112, the RNAi of OsWAK112 transgenic rice plants was also generated by introducing the vector pTCK303. Two independent OsWAK112 RNAi lines (RNAi1 and RNAi3, which showed around 4 and 7% of OsWAK112 mRNA level) were assessed under salt stress conditions. Without NaCl treatment, the OsWAK112 RNAi plants were indistinguishable from wild-type japonica (Jap; Supplementary Figure S2). After treatment with 200mM NaCl for 4 days and subsequent growth by regular hydroponic culture for 8 days, the RNAi plants showed significantly higher survival rates compared with Jap (Supplementary Figure S2).

Taken together, these findings (Figures 1, 2 and Supplementary Figure S2) demonstrate that OsWAK112 plays a negative role in salt tolerance.

Salt stress induces the rapid accumulation of ROS, and an imbalance in ROS levels often triggers serious membrane damage in plants (You and Chan, 2015; Yang and Guo, 2018). To examine whether the increased salt sensitivity of OsWAK112OE was related to the ROS level, the quantity of H₂O₂, a major and stable ROS, was evaluated using CM-H₂DCFDA with 7-day-old roots from Jap and OsWAK112OE lines with or without 200mM NaCl treatment. This probe can be transported into cells, where its acetate groups are passively cleaved by intracellular esterases, producing the fluorescent compound dichlorodihydrofluorescein (DCF; Wang et al., 2021). As shown in Supplementary Figure S3A, without salt treatment, the H₂O₂ levels in the OsWAK112OE lines were slightly lower than that in Jap. After salt treatment, the H₂O₂ levels in Jap and in the two OsWAK112OE lines were dramatically increased, indicating the induction of H₂O₂ production by salt as described previously (Li et al., 2014). However, the H₂O₂ levels in the OsWAK112OE lines were significantly higher (110 and 103% higher than in OE7 and OE8 [individual controls], respectively) than in Jap (46% higher than its control; Supplementary Figures S3A,B), revealing that the OsWAK112OE plants accumulated more ROS during salt treatment.

To investigate the tissue-specific expression pattern of OsWAK112, the 2,108-base pair promoter sequence was inserted into a pCAMBIA1300 vector harboring the gene encoding GUS. This proOsWAK112::GUS construct was introduced into rice plants by A. tumefaciens-mediated callus transformation. Histochemical staining revealed OsWAK112 expression in coleoptiles (Supplementary Figure S4A), ligules (Supplementary Figure S4B), leaf blades (Supplementary Figure S4C), stems (Supplementary Figures S4D,E), roots (Supplementary Figure S4F), and the radicles of germinated seeds (Supplementary Figure S4G) during vegetative growth. During reproductive growth, OsWAK112 was observed in paleae (Supplementary Figure S4H), stamens, and stigmas (Supplementary Figure S4I). These results show that OsWAK112 is expressed in most plant tissues.

Online sequence analyses have shown that OsWAK112 possesses a typical receptor kinase structure, including an extracellular domain, a transmembrane domain, and an intracellular kinase domain.¹ Thus, it is speculated to be a plasma membrane protein (Figure 3A); however, the sub-location of OsWAK112 is unclear. Thus, the full-length coding sequence of OsWAK112 fused with the sequence encoding green fluorescence protein (GFP) was generated and expressed transiently in tobacco (N. benthamiana) and onion (A. cepa) epidermal cells. As shown in Figure 3, GFP alone was expressed ubiquitously in the intracellular region of the tobacco epidermal cells (Figure 3B, GFP) and onion epidermal cells (Figure 3C, GFP), whereas OsWAK112-GFP was detected only in the region of the plasma membrane (Figures 3B,C, OsWAK112-GFP). To verify whether OsWAK112 is associated with the cell wall, onion epidermal cells were treated with 0.9 M mannitol to induce plasmolysis. After plasmolysis, OsWAK112-GFP was detected in the plasma membrane region and in the cell wall region (Figure 3C, OsWAK112-GFP plasmolyzed). In contrast, the control protein (GFP) was detected only in the intracellular region (Figure 3C, GFP plasmolyzed). These data indicate that OsWAK112 is localized to the plasma membrane and that it may connect tightly to the cell wall. Moreover, these data suggest that OsWAK112 is a wall-associated protein in plant cells.

To investigate the function of OsWAK112 further, we performed a sequence alignment and homology analysis between OsWAK112 and multiple RLKs in Arabidopsis using GenBank from the National Center for Biotechnology Information.² The sequence of OsWAK112 showed 40–60% similarity with those of WAK and cytosolic receptor kinase family members in Arabidopsis, and most of the conserved sequences were in the kinase domain (Supplementary Figure S5).

The sequence alignment of the kinase domain from OsWAK112 with those from several kinases in Arabidopsis is shown in Figure 4A. A typical ATP-binding motif and kinase active site were located in the intracellular domain of OsWAK112 (Figure 4A), indicating that it is an active kinase. Although it is predicted that OsWAK112 is linked to the cell wall via its cytosolic kinase domain (Figure 4A), whether WAK112 has kinase activity is unknown. To test this, the lysine (Lys-678) in the ATP-binding region and lysine (Lys-791) and aspartic acid (Asp794) in the activation loop of the kinase domain (KD) were mutated to generate WAK112KD^K678E, WAK112KD^K791R, and WAK112KD^D794A, respectively (Figure 4B). The wild-type and mutated forms of the WAK112 KD were produced as fusions with GST. An in vitro kinase assay showed that the OsWAK112KD produced strong autophosphorylation and phosphorylation of MBP (Figure 4B, lanes 1 and 6). WAK112KD^K791R also showed strong autophosphorylation (Figure 4B, lane 3), but the mutated forms (GST-OsWAK112^K678E and GST-OsWAK112^D794A; Figure 4B, lanes 2 and 4) and GST alone (Figure 4B, lane 5) did not exhibit autophosphorylation activity, indicating that OsWAK112 is an active RLK and that both Lys-678 and Asp-794 are necessary for its activity.

To assess the contribution of the kinase activity of OsWAK112 to plant salt sensitivity, we generated transgenic Arabidopsis expressing OsWAK112, OsWAK112^K678E, and OsWAK112^D794A as Myc fusions under the control of the 35S promoter. Western blotting showed that wild-type and mutated forms of OsWAK112 were well expressed in Arabidopsis (Supplementary Figure S6). On regular medium, little difference was noted among the seedlings (Figures 5A,C, 0mM NaCl). On 0.5×MS medium containing 125mM NaCl, the WAK112-overexpressing seedlings (OE10 and OE11) exhibited more strongly bleached cotyledons (Figure 5A) and lower survival rates than wild-type Col seedlings (57% for OE10 and 50% for OE11, compared with 63% for Col; Figure 5B). This is consistent with the effects of overexpressed OsWAK112 on salt hypersensitivity in rice (Figures 1, 2). However, the Arabidopsis 35S::OsWAK112^K678E (named OE12 and OE13) and 35S::OsWAK112^D794A (named OE14 and OE15) seedlings showed less damage in response to salt exposure, with around 70–86% survival (higher than that of Col; Figures 5C,D). These results suggest that the kinase activity of OsWAK112 affects salt resistance as a negative regulator.

Our work indicated that OsWAK112 plays a negative role in salt tolerance. As a wall-associated protein, it is likely that OsWAK112 confers salt response by directly binding to specific target proteins. Then, we sought to identify its interacting proteins to gain insight into the roles of OsWAK112 in salt response.

According to a structural analysis conducted using WebLab ViewerLite (Accelrys, San Diego, CA), several salt-associated proteins harbor important binding elements for OsWAK112, including GRP1, GRP2, GRP3, OsRP-1, PSBP1, and SAMSs (without OsACS and OsACO) as WAK112 candidate interacting proteins. Firstly, we performed BiFC assay to test whether OsWAK112 interacted with these proteins. The sequences encoding GRP1, GRP2, GRP3, OsRP-1, OsPSBP1, SAMS1/2/3, and OsWAK112 were fused with those sequences encoding the N-terminal half and C-terminal half of YFP to generate GRPs-CCFP, OsRP-1-CCFP, OsPSBP1-CCFP, OsSAMSs-CCFP, and OsWAK112-NYFP, respectively. The constructs were then co-infiltrated into N. benthamiana leaves. When the OsGRP1, OsGRP2, OsGRP3, OsRP-1, and OsPSB1 fusion with CCFP co-filtrated in OsWAK112 fusion with NYFP into N. benthamiana leaves, no fluorescence signal was observed (data not shown). However, co-expression of OsSAMS1/2/3-CCFP and OsWAK112-NYFP resulted in a strong YFP signal, while the negative control (OsWAK112-NYFP+YCE) produced almost no signal (Figure 6A), demonstrating the specific interaction of OsWAK112 with OsSAMSs in vivo. The direct interaction between OsWAK112 and OsSAMS1/2/3 was then examined in overlay assays. Purified MBP-OsSAMS1, MBP-OsSAMS2, and MBP-OsSAMS3 were used as primary antibodies and incubated with PVDF membranes containing GST and GST-OsWAK112KD; MBP-HRP antibodies were used as secondary antibodies. MBP-OsSAMS1, MBP-OsSAMS2, and MBP-SAMS3 were detected at positions corresponding to the size of GST-OsWAK112KD, but no band was detected at the position of the GST tag, suggesting a direct interaction between OsWAK112 and OsSAMS1/2/3 in vitro (Figure 6B). In addition, the overlay assay showed that D794A mutation in OsWAK112 did not affect its interaction with OsSAMS1 (Supplementary Figure S7), indicating the conserved aspartic acid or kinase activity was not required for this interaction. Furthermore, their interaction was further tested using Co-IP assays. The sequences encoding OsSAMS1/2/3 and OsWAK112KD were fused with GFP and Myc to generate OsSAMSs-GFP and OsWAK112KD-Myc, respectively. The constructs were then co-infiltrated into N. benthamiana leaves. Co-IP was achieved using N. benthamiana leaves transiently expressing OsSAMSs-GFP and OsWAK112KD-Myc. OsWAK112KD-Myc could be co-precipitated only when OsSAMSs-GFP was present (Figure 6C), confirming the interaction between OsWAK112 and OsSAMS1, OsSAMS2, or OsSAMS3 in vivo.

Together, these results suggest that OsWAK112 interacts physically with OsSAMS1/2/3 in vitro and in vivo. Next, the biological function of the OsWAK112–OsSAMS1 interaction was investigated in tobacco plants. A co-infiltration assay showed that OsSAMS1-GFP was well expressed in tobacco. Compared with mock treatment, 2h of treatment with 200mM NaCl decreased the levels of OsSAMS1 when co-expressed with OsWAK112, indicating that OsWAK112 affects OsSAMS1 stability under saline conditions (Supplementary Figure S8). Together, these data suggest that OsWAK112 interacts with OsSAMS1/2/3 and promotes OsSAMS1 degradation in the presence of salt.

To investigate the influence of OsWAK112 on OsSAMS contents in transgenic plants, the SAMS contents were measured in OE7, OE8, and Jap plants. The data showed that SAMS contents were not significantly different between Jap and OE7/8 under normal conditions, but the SAMS levels were lower in OE7 and OE8 than in Jap following salt treatment (Figure 7A).

As a major abiotic stress, salt can trigger the production of stress-related hormones and initiate endogenous hormone signaling, resulting in plant resistance reactions (Zhao et al., 2021b). Ethylene is a stress hormone closely related with salt stress (Riyazuddin et al., 2020). In plants, ethylene is biosynthesized from S-adenosyl-L-methionine (SAM; Zhao et al., 2021a), which is synthesized from l-methionine and ATP catalyzed by SAMSs. To explore the mechanism underlying the function of OsWAK112 in plant salt responses, the ethylene content was measured in Jap and OsWAK112OE plants. Under normal growth conditions, the ethylene contents did not differ much between Jap and the OsWAK112OE lines. After salt treatment, the ethylene contents increased in all of the plants; however, they were lower in OE7 and OE8 than in Jap (Figure 7B), while they were higher in RNAi1 and RNAi3 than in Jap (Supplementary Figure S9). Together, these data suggest that OsWAK112 interacts with OsSAMS1/2/3 and promotes OsSAMS1 degradation, resulting in decreased ethylene production in the presence of salt.

To our knowledge, OsWAK112 has only been shown to play a negative role in biotic stress tolerance in rice (Delteil et al., 2016); there are no reports of its function in response to abiotic stresses. The time-dependent suppression of OsWAK112 by NaCl treatment (Supplementary Figure S1 ) suggests that OsWAK112 is closely related to plant responses to high salinity. When OsWAK112 was overexpressed in rice, two independent lines (OE7 and OE8) showed hypersensitivity to salt stress. Further, the higher OsWAK112 expression level and the lower the survival rate in OE8 and OE7 (Figures 1, 2), and the less OsWAK112 expression level and the higher survival rate in RNAi lines compared with Jap (Supplementary Figure S2), indicate that the OsWAK112 expression level correlates negatively with plant salt resistance. These findings from morphological and physiological measurements indicate that OsWAK112 is a negative regulator of salt resistance reactions in rice.

Salinity stress induces ROS accumulation. Low levels of ROS usually act as signal molecules in salt signaling networks, but excessive accumulation of ROS is highly toxic to proteins, lipids, and nucleic acids, resulting in cell death (You and Chan, 2015). H₂O₂ is an especially stable ROS in plant cells. We found that more H₂O₂ accumulated in OsWAK112OE (OE7/8) plants under salt stress and that the variation in their H₂O₂ levels was consistent with that of their OsWAK112 expression levels (Supplementary Figure S3). Excessive accumulation of H₂O₂ accompanied by strong OsWAK112 expression under conditions of salt stress may damage plant cells, resulting in reduced survival.

As a kinase, how does OsWAK112 coordinately regulate plant responses to salt stress? It would be a worthwhile question to investigate.

Each WAK protein contains an extracellular calcium-binding domain, a transmembrane domain, and a cytoplasmic serine/threonine kinase domain (Kohorn and Kohorn, 2012). The extracellular domain senses the signal, while the cytosolic protein kinase domain translates the signal to downstream components via phosphorylation. OsWAK112 contains an arginine residue in its kinase catalytic motif (RDxxxxN) and is classified as a WAK-RD kinase (de Oliveira et al., 2014). Several RLKs have been shown to be involved in salt responses depending on their kinase activity. For example, SIT1, a lectin RLK, negatively regulates salt responses in rice, and it kinase activity plays an important role in its salt response (Li et al., 2014). Another receptor-like cytoplasmic kinase, STRK1, phosphorylates and stimulates the activity of catalase C to adjust H₂O₂ concentrations and improve salt tolerance (Zhou et al., 2018). Herein, we used an in vitro kinase assay to show that OsWAK112 has autophosphorylation activity and strong substrate activity (Figure 4). We found that Lys-791 near the DFG motif of kinase domain VII produced the same autophosphorylation signal as wild-type OsWAK112 in vitro (Figure 4); therefore, it appears that Lys-791 is not essential for the kinase activity of the enzyme in vitro. That may be because Lys-791 is not conserved in OsWAK112 compared with other known kinases. Our results show that OsWAK112 exhibited kinase activity depending on its conserved residues Lys-678 and Asp-794 (Figure 4). Lys-678 is in the AXK motif of kinase domain II for interaction with the phosphates of ATP. Asp-794 in the DFG motif of kinase domain VII chelates a Mg²⁺ ion to orient the γ-phosphate of ATP for transfer (Hanks and Hunter, 1995). When Lys-678 or Asp-794 was mutated, OsWAK112 lost its kinase activity (Figure 4). Exogenous expression of OsWAK112 in Arabidopsis made the transgenic plants hypersensitive to salt (Figure 5), indicating that OsWAK112 functions as a negative regulator of salt responses in both Arabidopsis and rice. However, exogenous overexpression of OsWAK112^K678E or OsWAK112^D794A did not increase the sensitivity of Arabidopsis to salt (Figure 5). A plausible explanation is that kinase-dead forms of OsWAK112 could still sense the salt stimulus and bind to downstream substrates, but the signal could not be transduced due to the loss of kinase activity (Figure 4), resulting in a dominant-negative phenotype. For example, in plants, when ERECTA lack the cytoplasmic kinase domain it acts as dominant-negative receptors by blocking the normal activity of the endogenous counterparts (Shpak et al., 2003). And in plant hormone brassinosteroids signaling pathway, the receptor kinase BRI1 phosphorylates the substrate BSK3 on S215. When S215 mutated into A, S215A had a dominant-negative effect: Plants overexpressing the S215A version of OsBSK3 were smaller and dwarfed than the nontransgenic controls when grown in the light under the same conditions (Zhang et al., 2016).

Ethylene homeostasis is tightly controlled to maintain its dual functions in growth inhibition and growth stimulation. Under normal conditions, the ethylene concentration is low and only increases dramatically at defined developmental stages, such as fruit ripening (Zhao et al., 2021a). A variety of environmental stimuli, including salt stress, can also induce ethylene production (Nascimento et al., 2018; Riyazuddin et al., 2020), in which the ACC synthase-induced conversion of SAM to ACC is a key step. SAM is synthesized by SAMSs, which play important roles in regulating plant–environment interactions. Studies show that SAMSs can be induced by multiple stress treatments and that they play important roles in regulating plant tolerance to environmental changes, especially soil salinity and drought. For example, in sugar beet, BvM14-SAMS2 is induced by salt, and its ectopic expression in Arabidopsis plants increased salt and oxidation resistance (Ma et al., 2017). In tobacco, overexpression of SlSAMS1 affects ethylene emission and improves drought and salt stress tolerance (Zhang et al., 2020). We found that OsWAK112 interacted directly with OsSAMS1/2/3 in overlay, BiFC, and Co-IP assays (Figure 6). Among these three SAMSs in rice, OsSAMS1 plays a predominant role and is expressed abundantly in every tissue (Li et al., 2011). We found that OsWAK112 promoted OsSAMS1 degradation under high-salt conditions (Supplementary Figure S8). There was a lower OsSAMS content and a reduced ethylene content in our OsWAK112OE plants, following salt treatment (Figure 7), while a higher ethylene existed in OsWAK112 RNAi plants content under salt stress. Consistent with our study, OsARD1, a metal-binding metalloenzyme involved in the methionine salvage pathway, enhances ethylene biosynthesis and reduces the salt sensitivity of rice (Liang et al., 2019). These data suggest that OsWAK112 is an active kinase that interacts with OsSAMS1/2/3, promoting OsSAMS1 degradation, and reducing ethylene production under saline conditions.

Soil salinity is a serious threat to crop growth and agricultural productivity. Our findings demonstrate that salt sensitivity in rice is mediated by OsWAK112 depending on its kinase activity. These findings provide important new information for engineering salt-tolerant crops.