Aquaporin OsPIP2;2 links the H₂O₂ signal and a membrane-anchored transcription factor to promote plant defense

Abstract

To overcome pathogen infection, plants deploy a highly efficient innate immune system, which often uses hydrogen peroxide (H₂O₂), a versatile reactive oxygen species, to activate downstream defense responses. H₂O₂ is a potential substrate of aquaporins (AQPs), the membrane channels that facilitate the transport of small compounds across plasma membranes or organelle membranes. To date, however, the functional relationship between AQPs and H₂O₂ in plant immunity is largely undissected. Here, we report that the rice (Oryza sativa) AQP OsPIP2;2 transports pathogen-induced apoplastic H₂O₂ into the cytoplasm to intensify rice resistance against various pathogens. OsPIP2;2-transported H₂O₂ is required for microbial molecular pattern flg22 to activate the MAPK cascade and to induce the downstream defense responses. In response to flg22, OsPIP2;2 is phosphorylated at the serine residue S125, and therefore gains the ability to transport H₂O₂. Phosphorylated OsPIP2;2 also triggers the translocation of OsmaMYB, a membrane-anchored MYB transcription factor, into the plant cell nucleus to impart flg22-induced defense responses against pathogen infection. On the contrary, if OsPIP2;2 is not phosphorylated, OsmaMYB remains associated with the plasma membrane, and plant defense responses are no longer induced. These results suggest that OsPIP2;2 positively regulates plant innate immunity by mediating H₂O₂ transport into the plant cell and mediating the translocation of OsmaMYB from plasma membrane to nucleus.

A rice AQPs acts as a positive immune regulator by transporting H₂O₂ and modulating a membrane-anchored transcription factor.

Introduction

Reactive oxygen species (ROS) play an important role in plant defenses against pathogen infections, and a major site for ROS production during plant–pathogen interactions is the apoplast (Qi et al., 2017; Zhang et al., 2018; Yang et al., 2020). The apoplastic ROS, especially superoxide (O₂⁻), is mainly produced by the NADPH oxidases (NOXs) located in the plasma membrane (Ryder et al., 2019). Upon perception of microbial molecular patterns such as bacterial flagellin, the plasma membrane-localized pattern-recognition receptor (PRR) complex activates the central immune regulator BOTRYTIS-INDUCED KINASE1 (BIK1). The activated BIK1 is released from the PRR complex and directly phosphorylates NOX to activate apoplastic O₂⁻ production (Kimura et al., 2020). The O₂⁻ production is followed by its dismutation to hydrogen peroxide (H₂O₂), which is aided by superoxide dismutase (SOD) present in the apoplast (Baluška and Vivanco, 2009; Sies et al., 2017). H₂O₂ is a relatively stable component of ROS compared with its precursor. H₂O₂ can be transported into the cytoplasm by particular members of the aquaporin (AQP) super family (Sies et al., 2017). The cytoplasmic H₂O₂ acts as a second messenger to mediate multiple defense responses, such as the activation of mitogen-activated protein kinases (MAPKs) and callose deposition (Petrov and Van Breusegem, 2012; Tian et al., 2016; Zhang et al., 2018; Yang et al., 2019).

AQPs were initially reported as water (H₂O) channel proteins (Preston and Agre, 1991). More and more studies have revealed that the function of AQPs extends beyond H₂O transport. The plasma membrane intrinsic protein (PIP) subfamily members have been shown to transport carbon dioxide (CO₂) or H₂O₂ (Nakhoul et al., 1998; Uehlein et al., 2003; Kaldenhoff and Fischer, 2006; Ludewig and Dynowski, 2009; Bienert and Chaumont, 2014; Zhang et al., 2019a). In mice, AQP3 regulates innate immunity at mucosal surfaces by transporting extracellular H₂O₂ into the cells (Thiagarajah et al., 2017). Human AQP8 facilitates the diffusional transport of H₂O₂ across the mitochondrial membrane to regulate cell viability (Marchissio et al., 2012). In plants, AtPIP1;4 contributes to photosynthesis by mediating CO₂ transport (Li et al., 2015), and also promotes plant disease resistance by transporting H₂O₂ from apoplast to cytoplasm (Tian et al., 2016). The cytoplasmic H₂O₂ transported by AtPIP1;4 activates pattern-triggered immunity (PTI), such as callose deposition and upregulation of defense-related genes (Tian et al., 2016). Interestingly, the enhanced disease resistance by AtPIP1;4 expression in Arabidopsis thaliana (Arabidopsis) is not due to the higher H₂O₂ accumulation, since NOX and AtPIP1;4 act independently in H₂O₂ production and transportation (Tian et al., 2016). The role of another AQP AtPIP2;1 is required for cytoplasmic H₂O₂ accumulation after flg22 or ABA treatment, thus promoting stomatal closure (Rodrigues et al., 2017). In addition to AtPIP1;4 and AtPIP2;1, four other AtPIP2 isoforms (AtPIP2;2, AtPIP2;4, AtPIP2;5, and AtPIP2;7) mediate H₂O₂ transport in yeast and decrease the viability of yeast (Bienert and Chaumont, 2014). However, it remains unclear if these PIPs also transport H₂O₂ across plant plasma membranes. It is also largely unknown if PIPs in crops act in a similar way to modulate plant immunity.

The activity of AQPs is modulated by posttranslational modifications (Maurel et al., 2015). Mass spectrometry analyses have revealed that AQPs undergo numerous posttranslational modifications, including phosphorylation, methylation, NH₂-terminal acetylation, and ubiquitination (Maurel et al., 2015). Among these modifications, phosphorylation plays a critical role in AQP gating and trafficking. Phosphorylation of serine (S) residues has been identified in the N-terminal region of Arabidopsis PIP2 and maize (Zea mays) PIP1 (Maurel et al., 2009). Phosphorylation of multiple serine and threonine residues has also been identified in the C-terminal tail of Arabidopsis PIP2s (Maurel et al., 2009). An S203A mutation in ZmPIP2;1 reduces its H₂O transport activity. The gating activity of spinach (Spinacia oleracea) SoPIP2;1 is activated by the phosphorylation of conserved S197 (Zhang et al., 2019a). Phosphorylation of AtPIP2;1 at S121 promotes H₂O₂ transport into guard cells to trigger stomatal closure after the treatment of flg22 or ABA (Rodrigues et al., 2017), indicating a role for AQP phosphorylation in response to biotic and abiotic stresses.

In rose (Rosa hybrida), the AQP RhPIP2;1 associates with a membrane-anchored MYB transcriptional factor (maMYB) RhPTM, and the phosphorylation of RhPIP2;1 at S273 promotes the nuclear accumulation of RhPTM C terminus under drought condition. Silencing RhPTM promoted plant growth but decreased the survival rate under drought stress (Zhang et al., 2019b). The role of maMYB in development has also been elucidated (Slabaugh et al., 2011). However, it is still largely unknown if maMYBs play a role during plant–pathogen interactions.

In this study, we identified an AQP OsPIP2;2 from rice, which is strongly induced by pathogen infection and exhibits H₂O₂ transport activity. We demonstrated that OsPIP2;2 positively regulates plant resistance to bacterial and fungal pathogens partially by transporting apoplastic H₂O₂ into cytoplasm. In addition, we identified a transcriptional factor OsmaMYB that associates with OsPIP2;2. We demonstrated that OsmaMYB contributes to plant defense and its translocating into the nucleus is regulated by the phosphorylation of OsPIP2;2 at S125. Together, our results indicate that OsPIP2;2 functions as a positive immune regulator by linking H₂O₂ transport and nuclear accumulation of OsmaMYB, providing a scientific basis for engineering enhanced plant resistance by editing OsPIP2;2 in the future.

Results

OsPIP2;2 acts as a major H₂O₂-transporting facilitator

We analyzed the expression levels of 11 OsPIP genes in wild-type Nipponbare (NPB) at 8 h after Xanthomonas oryzae pv. oryzae (Xoo) strain PXO99^A inoculation or 1 h after 2 mM H₂O₂ treatment. The expression of most OsPIPs was upregulated by PXO99^A or H₂O₂ treatment, among which the expression levels of OsPIP1;3, OsPIP2;2, OsPIP2;3, and OsPIP2;6 were more obviously induced (Supplemental Figure S1, A and B). Under normal growth conditions, the expression degree of OsPIP2;2 was much higher than that of OsPIP1;3, OsPIP2;3, or OsPIP2;6 (Supplemental Figure S1, C). Given that the total H₂O₂ content was significantly increased at 8 h postinoculation with PXO99^A (Supplemental Figure S1, D), these OsPIPs may link H₂O₂ signal to plant defense response.

To investigate the H₂O₂ transport activity of OsPIPs, 11 OsPIPs were cloned and transformed individually into yeast cells. Yeast cells were grown on the SD-Ura medium plate containing different concentrations of H₂O₂. The results demonstrated that yeast cells transformed with OsPIP1;3, OsPIP2;2, OsPIP2;3, or OsPIP2;6 were more sensitive to H₂O₂ compared with the empty vector control and other OsPIPs (Figure 1, A), which was caused by toxicity of the transported H₂O₂ from apoplast to cytoplasm. Western blot assay confirmed that each OsPIP was normally expressed in yeast cells (Figure 1, A). This result indicated that OsPIP1;3, OsPIP2;2, OsPIP2;3, and OsPIP2;6 promote yeast sensitive to H₂O₂ by mediating H₂O₂ transport. To further demonstrate the role of these OsPIPs in H₂O₂ transport, the H₂O₂ probes 2′,7′-dichlorofluorescin diacetate (H₂DCF-DA) and Amplex Red (AR) were employed to evaluate cytoplasmic H₂O₂, and Amplex Ultra Red (AUR) probe which is impermeable to plasma membrane was used to detect the apoplastic H₂O₂. The combination of these probes can indicate the transport efficiency of H₂O₂. The H₂O₂ in the apoplast and cytoplasm was quantified by a microplate reader. The result indicates that OsPIP1;3, OsPIP2;2, and OsPIP2;3, but not OsPIP2;6, can efficiently transport H₂O₂ into cytoplasm by using the H₂DCF-DA probe, among which OsPIP2;2 exhibited the highest H₂O₂ transport activity (Figure 1, B and Supplemental Figure S2, A, D, and G). Similar results were obtained by using another two probes AR and AUR (Figure 1, C and D and Supplemental Figure S2, B, C, E, F, H, and I). In addition, OsPIP2;2 exhibits the highest expression level among these three OsPIPs under normal growth conditions (Supplemental Figure S1, C), and its expression was significantly induced by PXO99^A inoculation or H₂O₂ treatment compared to other OsPIPs (Supplemental Figure S1, A and B). Therefore, we focused on OsPIP2;2 for further analyses. To visualize H₂O₂ transport, laser confocal microscopy was conducted with H₂DCF-DA and AR probes. As expected, H₂O₂ can be efficiently transported into yeast cells by OsPIP2;2 expression as indicated by fluorescent signal (Figure 1, E and F).

Figure 1.

OsPIP2;2 mediates H₂O₂ transport in yeast. A, Yeast colonies grown on SD-Ura medium with different concentrations of H₂O₂. The expression levels of OsPIPs in yeast cells were evaluated by western blot using an antibody against His tag, and Coomassie brilliant blue (CBB) staining was used to show protein loading. Photographs were taken at 3 d postincubation. In this figure and in yeast-involved figures provided hereafter, “Control” indicates the empty yeast binary vector without any insert of tested genes, in contrast to the recombinant vector carrying each of the OsPIPs. B, C, and D, Chronological changes of H₂O₂ content in yeast cells after treatment with 0 or 300 μM H₂O₂ detected by H₂DCF-DA, AR, and AUR probes respectively. The symbol “+” indicates supplied with 300 μM H₂O₂, and “−” indicates supplied with H₂O as control. Data are shown as means ± SEM (n = 8). E and F, H₂DCF-DA (E) and AR (F) probing of yeast cells 45 min after treatment with 300 μM H₂O₂.

We generated OsPIP2;2 mutant lines (Ospip2;2) using CRISPR/Cas9 technology and OsPIP2;2 overexpression (OsPIP2;2OE) lines (Supplemental Figure S3) to investigate if OsPIP2;2 is required for H₂O₂ transport and defense response in rice. Two representative homozygous mutant lines (Ospip2;2#41, Ospip2;2#279) and two independent OsPIP2;2OE lines (OsPIP2;2OE#1, OsPIP2;2OE#2) were obtained. There is a T insertion in the mutant line Ospip2;2#41, causing a frameshift in OsPIP2;2. A point mutation and three continuous nucleotide deletion (or probably three discontinuous nucleotide deletion) were identified in another mutant Ospip2;2#279 (Supplemental Figure S3, B), so that glutamic acid at position 6 (E6) was mutated into glycine (G) and alanine at position 7 (A7) was deleted in OsPIP2;2 (Supplemental Figure S3, B). The mutation of AtPIP2;1 at E6 interferes with its subcellular localization and H₂O transported activity (Sorieul et al., 2011). The OsPIP2;2 overexpression lines were firstly confirmed by genomic polymerase chain reaction (PCR) (Supplemental Figure S3, C). Then the increased transcriptional expression of OsPIP2;2 was confirmed by quantitative reverse transcription PCR (qRT-PCR), and OsPIP2;2 protein accumulation was detected by western blot in OsPIP2;2OE lines (Supplemental Figure S3, D and E). Proper H₂O₂ production is required for cell elongation (Tian et al., 2018). To evaluate if OsPIP2;2 is involved in H₂O₂ transport in planta, transgenic plants were grown on 1/2 Murashige and Skoog (MS) medium containing 0.5 mM H₂O₂. The plumule elongation was significantly promoted in NPB upon H₂O₂ treatment, and the effect was more obvious in OsPIP2;2-transgenic rice (Supplemental Figure S4). However, the Ospip2;2 lines were insensitive to H₂O₂ treatment (Supplemental Figure S4), suggesting that OsPIP2;2 can efficiently transport H₂O₂ into cytoplasm to promote cell elongation. The Ospip2;2 and OsPIP2;2OE lines grow similarly to the wild-type in greenhouse (Supplemental Figure S5), suggesting that OsPIP2;2 does not obviously affect plant growth and development.

OsPIP2;2 contributes to rice resistance against bacterial blight, leaf streak, and rice blast

It was widely considered that H₂O₂ acts as a signal molecule to regulate plant defense (Tian et al., 2016; Yang et al., 2020). Since OsPIP2;2 can transport H₂O₂, we propose that it modulates plant immunity. To test this hypothesis, we inoculated Ospip2;2, OsPIP2;2OE lines with rice bacterial pathogen Xoo strain PXO99^A or X. oryzae pv. oryzicola (Xoc) strain RS105. The OsPIP2;2OE lines were more resistant than the NPB, whereas the Ospip2;2 lines were more susceptible than NPB, as indicated by the lesion length and the bacterial titers after inoculation with PXO99^A or RS105 (Figure 2, A–F). To further investigate if OsPIP2;2 enhances rice resistance to fungal pathogens, the transgenic plants were inoculated with Magnaporthe oryzae strain HB3 by spraying a conidia suspension. The OsPIP2;2OE lines showed enhanced resistance to HB3 with scattered lesions and decreased fungal biomass compared to NPB (Figure 2, G and H). In contrast, the Ospip2;2 exhibited enhanced susceptibility to HB3 with increased fungal biomass (Figure 2, H). Together, these results indicate that OsPIP2;2 confers broad-spectrum resistance to multiple pathogens.

Figure 2.

OsPIP2;2 confers resistance to X. oryzae pv. oryzae (Xoo) strain PXO99^A, X. oryzae pv. oryzicola (Xoc) strain RS105, and M. oryzae strain HB3. A, Bacterial blight symptoms in leaves 15 d postinoculation with PXO99^A using the clipped-end method. In this figure and in plant-involved figures provided hereafter, “NPB” refers to the rice variety NPB, “OsPIP2;2OE” means an OsPIP2;2-overexpressing rice (NPB) line coded with a number during characterization, and “Ospip2;2” indicates an OsPIP2;2-defected NPB line generated by CRISPR/Cas9 and coded with a number. B, Lesion length in (A). Data are shown as means ± SEM (n = 20). C, Titers of PXO99^A 3 d postinoculation. Data are shown as means ± SEM (n = 4). D, Bacterial leaf stripe symptoms in leaves 5 d after RS105 infiltration. E, Lesion length in (D). Data are shown as means ± SEMs (n = 20). F, Titers of RS105 3 d postinoculation. Data are shown as means ± SEM (n = 3). G, Rice blast symptoms in leaves 5 d after spray inoculation with HB3 conidia. H, Relative fungal biomass in (G). The fungal biomass was determined using qPCR of the M. oryzae Pot2 gene against the rice OsActin gene. Data are shown as means ± SEM (n = 3). Lowercase letters indicate significant differences at by one-way ANOVA and Duncan’s multiple range tests (P ≤ 0.01).

OsPIP2;2 transports apoplastic H₂O₂ into cytoplasm and is required for PTI

To evaluate if OsPIP2;2 promotes total H₂O₂ accumulation, we measured H₂O₂ content in rice after PXO99^A inoculation. No obvious difference was observed in total H₂O₂ content among the wild-type NPB, OsPIP2;2OE, and Ospip2;2, as indicated by 3,3-diaminobenzidine (DAB) staining and AR quantification after PXO99^A inoculation (Supplemental Figure S6, A and B). The H₂O₂ production was significantly inhibited by the treatment with diphenylene iodonium (DPI), which reduces the H₂O₂ levels in plants, further confirming the H₂O₂ accumulation induced by PXO99^A inoculation (Supplemental Figure S6, A and B). The activity of SOD, which catalyzes the dismutation of superoxide anion to H₂O₂, was similar among the wild-type NPB, OsPIP2;2OE, and Ospip2;2 (Supplemental Figure S6, C). In addition, the expression of H₂O₂-related genes was also similar among these different lines (Supplemental Figure S6, D). Similar results were obtained when these lines were challenged with RS105 or HB3 (Supplemental Figures S7, S8). Together, the results above suggest that OsPIP2;2 does not enhance plant resistance by promoting total H₂O₂ accumulation.

To elucidate if OsPIP2;2 transports H₂O₂ during plant defense, we analyzed the kinetic changes of AR fluorescent intensity in wild-type NPB, OsPIP2;2OE and Ospip2;2 over 40 min after flg22 treatment. The content of cytoplasmic H₂O₂ was considerably increased in OsPIP2;2OE compared with the wild-type or Ospip2;2 (Figure 3, A). This result was confirmed by confocal laser scanning microscopy using H₂DCF-DA probe (Figure 3, B). These results suggest that OsPIP2;2 transports apoplastic H₂O₂ into the cytoplasm in planta. Callose deposition is linked to H₂O₂ signaling, and has been considered as a typical PTI defense response (Tian et al., 2016; Yang et al., 2019). The callose deposition was increased in OsPIP2;2OE but reduced in Ospip2;2 compared with that in the wild-type after flg22 treatment (Figure 3, C and D).

Figure 3.

OsPIP2;2 links PAMP-induced apoplastic H₂O₂ to the PTI pathway. A, Chronological changes in the H₂O₂-probing AR fluorescence densities in leaves of 2-week-old rice seedlings after flg22 treatment (means ± SEM, n = 8). OE#1, OsPIP2;2OE#1; OE#2, OsPIP2;2OE#2; KO#41, Ospip2;2#41; KO#279, Ospip2;2#279. B, Cell imaging of the DCF fluorescence by laser confocal microscopy. Leaves were preloaded with H₂DCF-DA 45 min before treatment with 300 μM H₂O₂. C, flg22-induced callose deposition in rice. The leaves of 2-week-old rice seedlings were treated with 10 μM flg22 for 16 h and then stained by aniline blue. B and C, The scale bar representing 100 μm applies to all the images. D, Quantification of callose deposition in leaves in C by ImageJ. Data were shown as means ± SEM (n = 3). E, MAPK activation was detected after flg22 treatment by western blot. The expected identities of the respective bands are marked on the right. The experiment was performed twice with similar results. F–H, Expression levels of pathogenesis-related genes OsPR1a and OsPR10 expression levels in plants with PXO99^A, RS105, or HB3 24 h postinoculation. Data are shown as means ± SEM (n = 4). Lowercase letters indicate significant differences by one-way ANOVA and Duncan’s multiple range tests (at P ≤ 0.01).

The MAPK activation, another typical defense response during PTI, was enhanced in OsPIP2;2OE but compromised in Ospip2;2 compared with that in the wild-type (Figure 3, E). Consistently, the expression of defense-related genes was also upregulated in OsPIP2;2OE, but reduced in Ospip2;2 compared with that in the wild-type when plants were challenged with different pathogens (Figure 3, F–H). Collectively, these results suggest that OsPIP2;2 positively regulates PTI by linking H₂O₂ transport to trigger plant defense response.

The S125 phosphorylation is required for the H₂O₂ transport activity of OsPIP2;2

The S121 phosphorylation of AtPIP2;1 is required for flg22- or ABA-induced H₂O₂ influx in Arabidopsis guard cells (Rodrigues et al., 2017). Multiple protein sequence alignment of AtPIP2;1, OsPIP2;2, ZmPIP2;5, and SoPIP2;1 indicated that the phosphorylation residue is conserved, and the corresponding residue in OsPIP2;2 is S125 (Supplemental Figure S9). To determine the role of S125 phosphorylation in H₂O₂ transport activity of OsPIP2;2, we generated a phosphorylation deficient mutant (OsPIP2;2 S125A) and a phosphomimetic mutant (OsPIP2;2 S125D). We transformed wild-type OsPIP2;2 and these two phosphorylation mutants into yeast strain NMY51 individually, and evaluated yeast growth on the SD-Ura plate supplemented with H₂O₂. The yeast transformants expressing OsPIP2;2 S125D were more sensitive to H₂O₂, whereas the yeast transformants expressing OsPIP2;2 S125A were insensitive to H₂O₂ compared to the wild-type OsPIP2;2 (Figure 4, A). This result indicates that the S125 phosphorylation of OsPIP2;2 is required for its H₂O₂ transport activity. In good agreement with this result, we further found that the OsPIP2;2 phosphomimetic mutant can efficiently transport exogenously supplied H₂O₂ into yeast cells, whereas the phosphorylation deficient mutant cannot (Figure 4, B). Quantification of the cytoplasmic H₂O₂ further confirmed the role of S125 phosphorylation in OsPIP2;2’s H₂O₂ transport function (Figure 4, C). To demonstrate the role of S125 phosphorylation in H₂O₂ transport activity of OsPIP2;2 in planta, we transiently expressed wild-type OsPIP2;2 or its two mutants in tobacco (Nicotiana benthamiana). We confirmed that the OsPIP2;2 phosphomimetic mutant can transport H₂O₂ efficiently into plant cells, whereas the phosphorylation deficient mutant cannot (Figure 4, D). Similar protein levels were observed among OsPIP2;2, OsPIP2;2 S125A, and OsPIP2;2 S125D as shown by western blot assay (Supplemental Figure S10, A).

Figure 4.

Phosphorylation of OsPIP2;2 at S125 is required for H₂O₂ transport. A, Yeast transformants grown in SD-Ura medium with H₂O₂ at 28°C. Photographs were collected at 3 d postincubation. B, Imaging of the DCF fluorescence in yeast cells by laser confocal microscopy using H₂DCF-DA dye after treatment with 300 μΜ H₂O₂. C, Kinetic variations of AR fluorescence densities in yeast transformants after application of 300 μΜ H₂O₂ over 40 min. Data are shown as means ± SEM (n = 8). D, Kinetic variations of AR fluorescence densities in N. benthamiana expressing OsPIP2;2 or its mutants after H₂O₂ treatment. Data are shown as means ± SEM (n = 8). E, Immunoblot analysis of OsPIP2;2 phosphorylation at S125 by using His antibody after Phos-tag gel separation. OsPIP2;2 or OsPIP2;2 S125A were expressed in leaves of N. benthamiana, and then the infiltrated leaves were treated with 1 μM flg22 for 30 min before protein extraction. Protein samples treated with (+) or without (–) calf intestinal alkaline phosphatase were separated by a Phos-tag gel. CBB staining was used to show protein loading (bottom panel). F, Subcellular localization of OsPIP2;2 and its mutants in N. benthamiana observed under lasers intensity as in (B).

To clarify if OsPIP2;2 is phosphorylated upon flg22 perception, we expressed OsPIP2;2 and OsPIP2;2 S125A in N. benthamiana leaves. The total protein was separated by a Phos-tag gel, and the phosphorylation was detected by western blot. Increased phosphorylation of OsPIP2;2 was observed upon flg22 treatment in the wild-type OsPIP2;2 but not in the OsPIP2;2 S125A mutant (Figure 4, E), suggesting that the S125 of OsPIP2;2 is phosphorylated after flg22 treatment. Since OsPIP2;2 transports apoplastic H₂O₂, we evaluated if the S125 phosphorylation of OsPIP2;2 is also induced by H₂O₂. We transformed OsPIP2;2 and OsPIP2;2 S125A into yeast strain ΔX33, respectively, and detected phosphorylation of these proteins after H₂O₂ treatment. Similar to flg22 treatment, only the wild-type OsPIP2;2 can be phosphorylated upon H₂O₂ treatment, but the OsPIP2;2 S125A cannot (Supplemental Figure S10, B). To evaluate if S125 phosphorylation affects OsPIP2;2’s subcellular localization, we expressed yellow-fluorescent protein (YFP)-tagged wild-type OsPIP2;2 and two phosphorylation mutants in N. benthamiana individually, and observed their fluorescent signal using a confocal laser scanning microscope. The wild-type OsPIP2;2 and the phosphorylation mutants exhibited similar subcellular localization, and they were likely to be localized to the plasma membrane (Figure 4, F and Supplemental Figure S10, C).

OsPIP2;2 associates with a membrane-anchored MYB-like transcription factor OsmaMYB that contributes to plant immunity

We identified potential OsPIP2;2 interactors using a split-ubiquitin membrane yeast two-hybrid (MYTH) system. Among 30 potential OsPIP2;2-interacting protein candidates obtained by screening a cDNA library, an Oryza sativa membrane-anchored MYB-like transcription factor (OsmaMYB) was identified, which is a homolog of the previously reported maMYB in rose (Zhang et al., 2019b; Supplemental Figure S11, B). Full-length cDNA of OsmaMYB was cloned and the association was confirmed in yeast two-hybrid system (Figure 5, A). To verify the OsPIP2;2–OsmaMYB association in planta, we co-expressed OsPIP2;2 and OsmaMYB in N. benthamiana, and confirmed their interaction by co-immunoprecipitation (Co-IP) and luciferase (LUC) complementation assays (Figure 5, B–D).

Figure 5.

OsPIP2;2 associates with OsmaMYB. A, Associations between OsPIP2;2 and OsmaMYB in the split-ubiquitin yeast two-hybrid system. Yeast cells co-transformed with bait and prey vectors were grown on medium supplemented with 10 mM of 3-amino-1,2,4-triazole. B, Co-IP of OsPIP2;2 and OsmaMYB after co-expression in N. benthamiana. The indicated constructs were co-expressed in N. benthamiana by A. tumefaciens. Total proteins were extracted 2 d postinfiltration, and subjected to anti-FLAG immunoprecipitations. C and D, Association between OsPIP2;2 and OsmaMYB indicated by the split-LUC complementation assay. The indicated constructs were co-expressed in N. benthamiana leaves and then LUC activities were examined in both qualitative (C) and quantitative assays (D). Cluc-CPR5 and BIK1-Nluc were used as the negative control. Cluc-XLG2 and BIK1-Nluc served as the positive control. Data are shown as means ± SEM (n = 8). Lowercase letters indicate significant differences at P ≤ 0.01, Duncan’s multiple range tests and one-way ANOVA.

To preliminarily evaluate if OsmaMYB is involved in plant immunity, OsmaMYB was expressed in N. benthamiana and the infiltrated leaves were then inoculated with DC3000 ΔhopQ1-1 48 h postinfiltration. The bacterial titer of DC3000 ΔhopQ1-1 was significantly lower in the leaves expressing OsmaMYB compared with that expressing LTI6b which is a plasma membrane protein as a control (Figure 6, A). This result indicates that OsmaMYB positively regulates plant defense against biotrophic bacterial pathogens.

Figure 6.

OsmaMYB contributes to plant immunity and is activated by phosphorylation of OsPIP2;2 at S125. A, Titers of DC3000 ΔhopQ1-1 3 d post-inoculation. Agrobacterium tumefaciens was used to transiently express OsmaMYB on one-half of an N. benthamiana leaf and LTI6b control on the other. Data are shown as means ± SEM (n = 6). Asterisks indicate significant differences compared to control by Student’s t test (**P ≤ 0.01). The experiment was performed twice with similar results. B, Subcellular localization of OsmaMYB by confocal laser microscopy. OsmaMYB-GFP fusion was transiently expressed in N. benthamiana leaves, and the infiltrated leaves were treated with H₂O or flg22. C, Immunoblot analysis of OsmaMYB. The infiltrated leaves were treated with flg22, and samples were collected at three different time points for immunoblot analysis. The OsmaMYB was detected using an anti-GFP antibody. CBB staining is used to show protein loading (bottom panel). D, Subcellular localization of OsmaMYB-GFP after co-expression with OsPIP2;2 S125A or OsPIP2;2 S125D. B and D, orange scale bar representing 100 μm applies to all the images. E, Immunoblot analysis of OsmaMYB-GFP co-expressed with the OsPIP2;2 phosphorylation mutants. F, Subcellular fractionation analysis of the distribution of OsmaMYB. Histone H3 was used as a nuclear marker, Na⁺/K⁺-ATPase was used as a membrane marker. G, Quantification of LUC activity. The indicated Nluc and Cluc constructs were co-expressed in N. benthamiana leaves for LUC complementation assay. The relative luminescence unit indicates the strength of protein–protein interaction. Data are shown as mean ± SEM (n = 8). S125A, phosphor-deficient form; S125D, phosphomimetic form. Lowercase letters indicate significant differences at P ≤ 0.01 by Duncan’s multiple range tests and one-way ANOVA. These experiments were performed at least twice with similar results.

To primarily explore how OsmaMYB modulates plant immunity, we analyzed the subcellular localization of OsmaMYB upon flg22 treatment. OsmaMYB contains a predicted nuclear localization signal near the C terminus (Supplemental Figure S11, A), suggesting the possibility of its nuclear localization. In nuclear staining with 4,6-diamidino-2-phenylindole (DAPI), no fluorescent signal of OsmaMYB-green-fluorescent protein (GFP) was observed in the nucleus after H₂O treatment, but nuclear accumulation of OsmaMYB-GFP was observed after flg22 treatment (Figure 6, B). The result of western blot demonstrated that an additional band of OsmaMYB-GFP appeared after flg22 treatment (Figure 6, C), indicating that OsmaMYB is cleaved after microbial pattern treatment. To further determine if OsmaMYB is released from the plasma membrane to regulate plant defense, we generated an OsmaMYB mutant by deleting its predicted transmembrane domain (TMD), and transiently expressed it in N. benthamiana leaves. As expected, the deletion mutant OsmaMYB^ΔTMD escaped from the plasma membrane, and was distributed in the cytoplasm and nucleus (Supplemental Figure S12, A). The infiltrated leaves were then inoculated with DC3000 ΔhopQ1-1, and enhanced resistance was observed in leaves expressing OsmaMYB^ΔTMD compared with the wild-type OsmaMYB or the LTI6b control (Supplemental Figure S12, B).

The transactivation activity of OsmaMYB was evaluated in yeast cells. The deletion mutant OsmaMYB^ΔTMD was fused to the DNA binding domain (BD) of the yeast transcription factor GAL4, and the recombinant plasmid was introduced into yeast strain AH109. The yeast cells expressing BD-OsmaMYB^ΔTMD can grow on the selective medium SD-Trp-His, whereas the cells expressing BD empty vector cannot (Supplemental Figure S12C), suggesting that OsmaMYB^ΔTMD exhibits transactivation activity.

Phosphorylation of OsPIP2;2 at S125 promotes translocation of OsmaMYB into the nucleus

Membrane-anchored transcriptional factors are maintained in an inactive state by associating with membranes, and are released from the plasma membrane and translocated into the nucleus to regulate transcription of target genes in response to certain signals (Zhang et al., 2019b). Upon flg22 treatment, the OsmaMYB C terminus accumulates in the nucleus (Figure 6, B). We also demonstrated that OsPIP2;2 associated with OsmaMYB (Figure 5). To clarify if the translocation of OsmaMYB C terminus to the nucleus is modulated by phosphorylation of OsPIP2;2 at S125, we co-expressed OsmaMYB-GFP with OsPIP2;2 S125A or OsPIP2;2 S125D in N. benthamiana leaves, and observed the subcellular localization of OsmaMYB-GFP by confocal laser scanning microscopy. No fluorescent signal of OsmaMYB-GFP was observed in the nucleus when OsPIP2;2 S125A was co-expressed, whereas nuclear accumulation of OsmaMYB-GFP was found when OsPIP2;2 S125D was present (Figure 6, D). The western blot result confirmed that OsmaMYB was cleaved when OsPIP2;2 S125D, but not OsPIP2;2 S125A, was co-expressed (Figure 6, E). To further confirm that OsmaMYB is translocated into the nucleus upon the perception of flg22 or phosphorylation of OsPIP2;2 at S125, we performed subcellular fractionation and detected OsmaMYB by western blot. OsmaMYB was only detected in the membrane fraction without flg22 treatment or when co-expressed with OsPIP2;2 S125A. However, it was detected in both membrane and nucleus upon flg22 treatment or when co-expressed with OsPIP2;2 S125D. We also noticed that only the truncated OsmaMYB can be detected in the nucleus, further indicating that OsmaMYB is cleaved and translocated into the nucleus upon flg22 perception or phosphorylation of OsPIP2;2 at S125 (Figure 6, F).

Intriguingly, OsPIP2;2 S125A associates stronger with OsmaMYB, whereas OsPIP2;2 S125D has a lower affinity with OsmaMYB compared with wild-type OsPIP2;2 as indicated by LUC complementation assay (Figure 6, G and Supplemental Figure S13, A). This result was confirmed by yeast two-hybrid assay (Supplemental Figure S13, B). These results indicate that OsmaMYB associates with OsPIP2;2 under normal growth conditions and is released from the OsmaMYB–OsPIP2;2 protein complex upon microbial pattern perception to regulate plant defense.

Discussion

AQPs were initially defined as H₂O transporting channels and subsequently were demonstrated to transport many other substrates (Ludewig and Dynowski, 2009; Bienert et al., 2014; Bienert and Chaumont, 2014). In addition to transporting H₂O, the PIP subfamily members have been shown to transport H₂O₂ (Tian et al., 2016; Rodrigues et al., 2017). H₂O₂ transport connects signals between apoplast and cytoplasm, resulting in plant resistance to pathogens (Tian et al., 2016). In this study, we systematically analyzed the expression of OsPIPs upon pathogen infection and their H₂O₂ transport activity, and demonstrated that the expression of most OsPIPs is upregulated upon pathogen infection, but only a few of OsPIPs are able to transport H₂O₂ (Figure 1 and Supplemental Figure S1, A). These results strongly suggest a role for OsPIPs in plant defense. Among these OsPIPs, the expression of OsPIP2;2 is strongly induced by pathogen infection, and exhibits a high basal expression level and H₂O₂ transport activity. Therefore, we focused on OsPIP2;2.

The OsPIP2;2 overexpression lines are more resistant to three different pathogens compared with the wild-type plants (Figure 2), suggesting that OsPIP2;2 is an ideal target for engineering broad-spectrum resistance in rice. No obvious difference was found in total H₂O₂ production among the wild-type NPB, OsPIP2;2OE and Ospip2;2 lines, suggesting that OsPIP2;2 does not modulate plant defense by promoting total H₂O₂ accumulation (Supplemental Figures S6–S8). Therefore, we proposed that OsPIP2;2 may regulate plant immunity by linking H₂O₂ transport. The H₂O₂ transported from apoplast can act as a signaling molecule to trigger defense responses (Hu et al., 2017; Sies et al., 2017). Indeed, enhanced callose deposition was observed in the OsPIP2;2 overexpression lines. It has been demonstrated in both animals and plants that H₂O₂ activates MAPK cascades (Zhang et al., 2018). In this study, MAPK cascade was enhanced in the OsPIP2;2 overexpression lines upon flg22 treatment compared to wild-type, but was significantly compromised in Ospip2;2 (Figure 3, E). The MAPK cascade activates the expression of many defense-related genes such as PR1 and PR10 in the salicylic acid signaling pathway (Medina-Puche et al., 2020). As expected, expression of PR1 and PR10 was significantly upregulated in OsPIP2;2OE after pathogen infection (Figure 3, F–H). Our results strongly support that OsPIP2;2 positively modulates plant resistance to bacterial and fungal pathogens by transporting apoplastic H₂O₂ into the cytoplasm.

It is still unclear how PIPs are regulated during plant–pathogen interactions. The activity of PIPs can be regulated by protein modifications, such as phosphorylation and dephosphorylation (Prak et al., 2008; Qing et al., 2016). The E6 residue was mutated into glycine in the mutant Ospip2;2#279 and the mutant exhibited compromised H₂O₂ transport and disease resistance, indicating that the E6 residue of OsPIP2;2 is essential for its function in plant defense. It has been shown that mutation of AtPIP2;1 at E6 interferes with its subcellular localization and H₂O transported activity (Sorieul et al., 2011). It was also indicated that the E6 residue in AtPIP2;1 can be methylated (Santoni et al., 2006). However, it is still unclear if methylation at E6 modulates AQP protein targeting. It was demonstrated that the AQP proteins retained partial substrate transport activity after the protein kinase inhibitor K252a treatment (Johansson et al., 1998; Azad et al., 2008), indicating that the AQPs exhibit basal activity at a nonphosphorylated state. The H₂O transport activity of spinach SoPIP2;1 is activated by phosphorylation at S197 (Johansson et al., 1996, 1998; Zhang et al., 2019a), and is suppressed by dephosphorylation at S115 and S274 under drought stress (Nyblom et al., 2009). Recently, AtPIP2;1 was demonstrated to transport H₂O₂ and its activity is activated by phosphorylation at S121. Brassinosteroid insensitive 1-associated receptor kinase 1 (BAK1) and open stomata 1/Snf1-related protein kinase 2.6 (SnRK2.6) were both required for flg22-induced AtPIP2;1 phosphorylation at S121 (Grondin et al., 2015; Rodrigues et al., 2017). Interestingly, the S121 is conserved among the homologs of AtPIP2;1. The corresponding residue in OsPIP2;2 is S125. We demonstrated that the OsPIP2;2 phosphorylation deficient mutant failed to transport H₂O₂, whereas the phosphomimetic mutant transported apoplastic H₂O₂ into the cytoplasm more efficiently (Figure 4, A–D), highlighting the role of S125 phosphorylation in OsPIP2;2’s activity. In Arabidopsis, both BAK1 and SnRK2.6 are required for AtPIP2;1 phosphorylation at S121 (Rodrigues et al., 2017). The S125 phosphorylation of OsPIP2;2 was detected after flg22 treatment in N. benthamiana (Figure 4, E), indicating that the co-receptor BAK1 is involved in OsPIP2;2 phosphorylation. The SnRK–PIP association has been demonstrated before (Song et al., 2019), so it is not surprising that SnRK family members are required for OsPIP2;2 phosphorylation. The phosphorylation of OsPIP2;2 at S125 is detected in yeast cells after H₂O₂ treatment, and the SnRK family exists in yeast (Supplemental Figure S10, B). Therefore, we propose that BAK1 and SnRK2.6 homologs in rice are responsible for the S125 phosphorylation of OsPIP2;2.

A membrane-anchored transcription factor OsmaMYB was identified as an interactor of OsPIP2;2 (Figure 5). We also demonstrated that several other OsPIPs with H₂O₂ transport activity also associated with OsmaMYB (Supplemental Figure S13, A and B), indicating that these OsPIPs utilize a similar mechanism to modulate plant defense. However, OsPIP2;2 associated more strongly with OsmaMYB (Supplemental Figure S13, A and B) and exhibited higher basal expression level (Supplemental Figure S1, C). It was reported that OsPIP2;2 may recruit OsPIP1;3 to the plasma membrane (Liu et al., 2019), indicating that OsPIP2;2 can modulate other OsPIPs. The evidence above together may explain why mutating OsPIP2;2 alone significantly compromised disease resistance and H₂O₂ transport (Figures 2, 3).

The role of maMYB in development has been characterized (Slabaugh et al., 2011). Recently, a membrane-anchored MYB protein, RhPTM, was reported to associate with an AQP RhPIP2;1 to regulate the trade-off between growth and stress survival in rose (Zhang et al., 2019b). Although it has been well documented that some MYB transcriptional factors contribute to plant resistance (Chinnusamy and Zhu, 2009), the role of maMYB subfamily in plant–pathogen interactions is still poorly understood. To uncover if OsmaMYB is involved in plant defense, we transiently expressed OsmaMYB in N. benthamiana, and inoculated the infiltrated leaves with bacterial pathogen DC3000 ΔhopQ1-1. The result indicates that OsmaMYB positively modulates plant immunity (Figure 6, A). To preliminarily elucidate how OsmaMYB modulates plant immunity, subcellular localization of OsmaMYB was observed after flg22 treatment. OsmaMYB was localized to the plasma membrane at a resting state (Figure 6, B), but was cleaved and translocated into the nucleus after flg22 treatment (Figure 6, B and C), indicating that it is released from the plasma membrane upon the microbial pattern perception. The truncated OsmaMYB^ΔTMD conferred more resistance to pathogen infection compared with the full-length OsmaMYB (Supplemental Figure S12, B), further suggesting that OsmaMYB needs to be released from the plasma membrane to regulate plant defense. OsmaMYB was maintained in the plasma membrane when co-expressed with the phosphorylation deficient mutant OsPIP2;2 S125A, but was processed and translocated into the nucleus when co-expressed with the phosphomimic mutant OsPIP2;2 S125D (Figure 6, D–F). This result indicates that OsmaMYB is dynamically modulated by the phosphorylation of OsPIP2;2 at S125. OsmaMYB associated more strongly with OsPIP2;2 S125A than OsPIP2;2 S125D (Figure 6, G and Supplemental Figure S13, A–C), further confirming that OsmaMYB is regulated by the S125 phosphorylation of OsPIP2;2. Under normal growth condition, we did not observe obvious phosphorylation signal by Phos-tag analysis (Figure 4, E and Supplemental Figure S10, B), indicating that OsPIP2;2 is not phosphorylated or phosphorylated at a very low level. We proposed that OsPIP2;2 retains its basal substrate transport activity under normal growth condition, and its activity is enhanced by phosphorylation and results in cleavage of OsmaMYB. Thus, we propose that OsPIP2;2 associates with OsmaMYB without treatment, and is phosphorylated at S125 upon microbial perception, causing the processing and release of OsmaMYB from the plasma membrane (Figure 7). The truncated C terminus of OsmaMYB is then translocated into the nucleus and activates the expression of defense-related genes (Figure 7). Hence, we propose that OsPIP2;2 enhances plant resistance by mediating H₂O₂ transport into the plant cells and mediating translocation of OsmaMYB from the plasma membrane to the nucleus, providing an effective mechanism to finely modulate plant immunity.

Figure 7.

Model for the modulation of plant immunity by OsPIP2;2 through H₂O₂ transport and OsmaMYB. At a resting state, OsPIP2;2 associates with the membrane-anchored OsmaMYB, and plant defense cannot be activated. Upon perception of microbial pattern flg22, the membrane-associated enzyme RBOHA/B is activated to catalize H₂O₂ generation in the apoplast. Upon flg22 perception, moreover, the AQP OsPIP2;2 is phosphorylated at S125 and therefore functions to mediate the H₂O₂ transport into the cytoplasm. The transported H₂O₂ triggers a series of defense responses including MAPK cascade, callose deposition, and upregulation of defense-related genes. Meanwhile, OsmaMYB is processed by an unknown protease, and the C terminus is released from the plasma membrane and de-associates with the OsPIP2;2 complex. Thereafter OsmaMYB is translocated into the nucleus and activates the expression of defense-related genes. In this way, OsPIP2;2 enhances plant resistance by linking H₂O₂ transport and OsmaMYB.

Materials and methods

All the gene constructs were sequenced to make sure they were correct. The primers used in this study are provided in Supplemental Table S1.

Biological materials and growth conditions

Rice NPB and N. benthamiana plants were grown in a growth chamber at 27°C and 70% relative humidity under a 16-h light and 8-h dark photoperiod. Escherichia coli strains were grown in Luria–Bertani (LB) medium containing appropriate antibiotics at 37°C. Agrobacterium tumefaciens strains were cultured in LB medium supplemented with rifampicin at 28°C. Saccharomyces cerevisiae strain NMY51 was grown on yeast extract peptone dextrose agar) medium at 28°C. The Pichia pastoris deletion strain ΔX33 containing a disrupted endogenous AQY1 gene was cultured on yeast extract peptone agar medium at 28°C. Xanthomonas oryzae pv. oryzae (Xoo) strain PXO99^A and X. oryzae pv. oryzicola (Xoc) RS105 were grown in nutrient broth (NB) or NB amended with agar at 28°C. The M. oryzae strain HB3 was cultured on potato sucrose agar at 28°C, and conidial formations were induced under light for 2–3 d after removing the surface mycelium.

RNA and DNA isolation

RNA or DNA was extracted from leaves of 2-week-old rice seedlings. Total RNA was isolated using TRIzol according to the manufacturer’s protocol (Invitrogen, Carlsbad, CA, USA). Genomic DNA was extracted following the cetyltrimethylammonium bromide-based method as described previously (Porebski et al., 1997).

qRT-PCR

One microgram of total RNA was reverse transcribed with an oligo(dT) primer using HiScript Q RT SuperMix for qRT-PCR (Cat#R123-01, Vazyme, Nanjing, China). The OsActin1 gene was used as an internal control. qRT-PCR was performed in a 20-μL reaction using ChamQ SYBR qPCR Master Mix following the manufacturer’s recommendation (Cat#Q341-02; Vazyme). qRT-PCR primers are provided in Supplemental Table S1. Reactions were performed on a QuantStudio 3 Real-Time PCR System (Applied Biosystems, Waltham, MA, USA) under the following conditions: 95°C for 30 s, then 40 cycles at 95°C for 10 s and 60°C for 30 s, followed by 95°C for 15 s and 60°C for 1 min, and then 95°C for 15 s to obtain melt curves. The relative expression level of each gene was determined using the 2^−ΔΔCt method (Livak and Schmittgen, 2001).

Bioinformatics analyses

The amino acid sequence of the OsPIP2;2 protein was aligned with the AtPIP2;1, SoPIP2;1, and ZmPIP2;5 sequences using the NCBI Blast Multiple Alignment tool. The Multiple Alignment tool was also used to align the protein sequences of OsmaMYB and its homologs (ZmmaMYB-like, TdmaMYB-like, RhPTM, SlymaMYB, and AtmaMYB). TMDs in these proteins were predicted with TMHMM Server version 2.0 (http://www.cbs.dtu.dk/services/TMHMM/) and located by Multiple Alignment, along with the known transporter filters and phosphorylation sites in the sequences. A consensus neighbor-joining tree was established between OsmaMYB and its homologs using the MEGA version 6.0 program (https://www.megasoftware.net/).

H₂O₂ transport assay

To evaluate the H₂O₂ transport activity of OsPIPs, the OsPIP gene constructs were transformed into S. cerevisiae strain NMY51 individually. The yeast transformants were cultured in SD-Ura medium overnight at 30°C, and yeast cells were collected by centrifugation. The precipitate was washed 3 times with sterile H₂O, and resuspended in sterile H₂O to a final concentration of OD₆₀₀ = 0.01. Ten microliter of the suspension was spotted onto SD-Ura solid medium containing different concentrations of H₂O₂. Three days later, the viability of the yeast cells was scored.

Visualization and quantification of H₂O₂ in yeast and plant cells were performed as described previously (Tian et al., 2016).

Plant transformation

A full-length OsPIP2;2 coding sequence (CDS) was cloned from NPB, and the fragment fused with His tag was inserted into the binary vector pCAMBIA1301. The recombinant construct was transformed into wild-type NPB to generate OsPIP2;2OE lines. To edit OsPIP2;2 using the CRISPR/Cas9 system, the target sequences were selected within the coding regions of OsPIP2;2 (Supplemental Figure S3A), and sgRNAs were designed with CRISPR MultiTargeter (http://www.multicrispr.net/index.html) and synthesized by Tsingke (Beijing). BLAST program was used to evaluate the target specificity against rice genome sequences. The above constructs were introduced into NPB by the A. tumefaciens-mediated transformation method (Biorun, Wuhan).

Plant inoculations

Virulent Xoo strain PXO99^A and Xoc strain RS105 were grown overnight at 28°C, washed twice, and resuspended to OD₆₀₀ = 0.6 in sterile H₂O for inoculation. Leaves of 2-month-old rice plants were cut with scissors and dipped in the PXO99^A suspension, and the lesion lengths were measured at 14 d postinoculation (dpi). The leaves of 2-week-old seedlings were infiltrated with RS105 suspension, and the lesion lengths were measured at 5 dpi (Cai et al., 2017; Ma et al., 2021). Bacterial titers were determined by counting colony-forming units (CFUs). To evaluate rice resistance to M. oryzae, 2-week-old seedlings were sprayed inoculated with M. oryzae strain HB3 conidia suspension at a concentration of 1 × 10⁵ conidia/mL in 0.2% v/v Tween 20 (Yang et al., 2019). The fungal biomass was examined by the qPCR assay as described previously (Wang et al., 2020).

Callose deposition assay

Leaves of 2-week-old rice seedlings were sprayed with 10 μΜ flg22 (GeneScript, Nanjing) or H₂O containing 0.025% v/v Silwet L-77. Callose deposition was detected 16 h after treatment as described previously (Yang et al., 2019). The leaf samples were observed by using a Nikon microscope under UV light (340–380 nm).

DAB staining

DAB was used for staining H₂O₂ according to a previously described method (Wang et al., 2016). The leaves of 2-week-old rice seedlings were immersed in 100-mL aqueous DAB solution (50 mg DAB, 0.5 mM NaH₂PO₄, 20 μL Tween 20) and vacuumed for 20 min. The leaf tissues were then stained in DAB solution for 8 h, and de-stained in a solution containing ethanol: acetic acid (3:1 v/v) for 8 h. The de-stained leaves were observed under a light microscope (Nikon, Tokyo, Japan).

SOD activity analysis

SOD activity was measured using the Total SOD Assay Kit (Cat#S0103, Beyotime, Shanghai) according to the manufacturer’s instructions. Briefly, 0.2 g of leaf tissues were harvested, ground into fine powder in liquid nitrogen, and resuspended in phosphate buffer. The suspension was centrifuged at 12,000 g for 5 min at 4°C. The supernatant was used to determine total SOD activity.

MAPK assay

Two-week-old rice plants were sprayed with 10 µM flg22 (Genscript, Nanjing, 95% purity) or H₂O containing 0.025% v/v Silwet L-77 for 10 min. Leaves were ground into powder in liquid nitrogen and resuspended in extraction buffer (10 mM Tris–HCl, 50 mM NaCl, 5 mM dithhiothreitol [DTT], 0.2% v/v Triton X-100, 1× protease inhibitor cocktail (Cat#CW2200S, CWBIO, Beijing), 1× halt phosphatase inhibitor cocktail [Cat#P1081; Beyotime, Shanghai, China]). The protein concentration was determined using the BCA Protein Assay Kit (Cat#CW0014S; CWBIO, Beijing, China). Equal amounts of total protein were loaded on a 12% w/v SDS-PAGE gel, and the phosphorylation of MPK3, MPK4, and MPK6 was determined by using an anti-phospho-p44/42 monoclonal antibody (Cat#AM071; Beyotime, Shanghai).

Yeast two-hybrid assay

The split-ubiquitin based-MYTH system was used in this study. Full-length CDS of target genes were cloned into the pMetCgate vector, and the CDS of OsmaMYB was inserted into the pNXgate32-3HA. The recombinant constructs were co-transformed into S. cerevisiae strain NMY51. The presence of both plasmids was confirmed by growth on a SD-Trp-Leu plate, and interactions were evaluated by plating on a SD-Trp-Leu-His-Ade (SD-WLHA) plate.

Co-IP

Agrobacteria harboring the corresponding constructs were co-infiltrated into N. benthamiana leaves. The infiltrated leaves were harvested 3 d postinfiltration, ground in liquid nitrogen, and resuspended with IP buffer (25 mM Tris–HCl pH 7.4, 150 mM NaCl, 1 mM EDTA, 1% v/v Nonidet P-40, 5% v/v glycerol, 1 mM PMSF, 20 μM MG132, 5 mM DTT, and 1× protease inhibitor cocktail). After centrifugation at 12,000 g for 15 min at 4°C, 20 µL of anti-FLAG agarose beads (Abmart, Shanghai, China) were added and incubated for 3 h at 4°C. The beads were washed 3 times with wash buffer (10 mM Tris–HCl pH 8.0, 150 mM NaCl, 0.5 mM EDTA, 0.5% v/v Triton X-100, 0.5 mM DTT, 1 mM PMSF, and 1× protease inhibitor cocktail). Immunoprecipitates were analyzed by western blot.

LUC complementation assay

The LUC complementation assay was performed as described previously (Zhou et al., 2018). Agrobacterium tumefaciens harboring the indicated Nluc and Cluc constructs were co-infiltrated into N. benthamiana leaves. The leaves were infiltrated with 0.5 mM D-luciferin (Cat#L0159; Bivision, Nanjing, China) 48-h postagroinfiltration and kept in the dark for 5 min. The images were captured by a Living Imaging system (IVIS200; Xenogen, Inc., Alameda, CA, USA). To quantify relative LUC activity, leaf discs were taken 2 d after agroinfiltration, incubated with 1 mM luciferin in a 96-well, and relative LUC activity was measured by a microplate reader.

Transactivation assay

To evaluate transactivation activity assay in yeast, the truncated version of OsmaMYB, in which the predicted TMD was deleted, was cloned into pGBKT7 (Clontech) to fuse to the DNA-BD of GAL4. The pGBKT7 empty vector was served as a negative control. The construct was transformed into yeast strain AH109 and screened on the selective medium (lacking either Trp or both Trp and His).

Phosphorylation assay

OsPIP2;2 and its mutant were cloned and inserted into pPICZb and pCAMBIA1300 vector, respectively. To detect OsPIP2;2 phosphorylation in yeast, the recombinant pPICZb plasmids were transformed into yeast strain ΔX33 individually. To detect OsPIP2;2 phosphorylation in N. benthamiana, the pCAMBIA1300 recombinant constructs were separately introduced into Agrobacteria, and the recombinant proteins were transiently expressed in N. benthamiana leaves for 48 h. The infiltrated leaves were treated with 1 μM flg22 30 min before protein extraction.

The above total protein samples were incubated with calf intestinal alkaline phosphatase (Cat#9001-78-9; Solarbio, Beijing, China) at 37°C for 30 min, and then were separated by a 12% w/v SDS-PAGE gel contained biotinylated Mn²⁺-Phos-tag complex AAL-107 (Kinoshita et al., 2007) purchased from Wako (http://www.Phos-tag.com). The OsPIP2;2 protein was further detected by western blot with His antibody.

Transient expression in tobacco

Agrobacterium tumefaciens strain GV3101 containing different constructs was cultured in liquid LB medium at 28°C in a shaking incubator for 24 h, bacterial cells were collected by centrifugation at 3,500 g for 5 min, and resuspended in infiltration buffer (10 mM MgCl₂, 100 mM MES pH 5.7, and 2 mM acetosyringone) to an OD₆₀₀ = 0.3. The bacterial suspension was infiltrated into leaves of 4-week-old N. benthamiana. After 48 h, the infiltrated areas of the leaves were excised and examined with a laser confocal scanning microscope (LSM880, Zeiss, Jena). The argon laser intensity was set to 6% and 2% for DAPI and YFP, respectively. Wave lengths of excitation filters and emission light were 405 and 410–508 nm for DAPI, and the YFP signal was captured with 514-nm emission and 519–620-nm excitation. Chanel gains were adjusted as 600 and 550 for DAPI and YFP, respectively.

Protein extraction and western blotting

To detect OsPIP2;2 or OsmaMYB protein expression, total protein was extracted from leaves with extraction buffer made of 10 mM Tris–HCl, 50 mM NaCl, 5 mM dithiothreitol (DTT), 0.5% w/v SDS, 1× protease inhibitor cocktail (Cat#CW2200S; CWBIO, Beijing). For subcellular fractionation assay, 0.5 g leaf samples were ground in liquid nitrogen, and resuspended with extraction buffer (10 mM HEPES pH 7.8, 10 mM KCl, 10 mM MgCl₂, 5 mM EDTA, 250 mM sucrose, 0.5% v/v Triton X-100, 1 mM DTT, 0.2 mM PMSF), then the extracts were filtered with Miracloth to remove cellular debris. After centrifugation at 3,000 g for 20 min at 4°C, the precipitation was washed with extraction buffer and collected as the nuclear fraction. The supernatant was centrifuged at 16,000 g for 30 min at 4°C, and the precipitation was collected as the membrane faction. Protein extracts were separated on 12% w/v SDS-PAGE gels and then transferred to a polyvinylidene difluoride membrane (Millipore, Burlington, MA, USA). The target proteins were detected by hybridization with the corresponding antibodies used at dilutions: α-His (TransGen Biotech, Beijing, China) or α-GFP (TransGen Biotech) at 1:4,000, α-FLAG (Affinity Biosciences) at 1:5,000, and α-Histone H3 (PhytoAB Inc., San Francisco, CA, USA) or α-Na⁺/K⁺-ATPase (PhytoAB Inc.) at 1:2,000. Secondary peroxidase-conjugated anti-mouse or anti-rabbit antibody (TransGen Biotech) was used at 1:5,000 dilution.

Statistical analysis

Quantitative data were analyzed using the NCBI ImageJ program and IBM SPSS version 19.0 (Shi, 2012), respectively. Significance in differences between (among) data from different plants or treatments was estimated by analysis of variance along with Student’s t tests or Duncan’s new multiple range tests.

Accession numbers

Sequence information in this study can be found in GenBank databases under the following accession numbers (proteins): LOC_Os02g44630 (OsPIP1;1), LOC_Os04g47220 (OsPIP1;2), LOC_Os02g57720 (OsPIP1;3), LOC_Os07g26690 (OsPIP2;1), LOC_Os02g41860 (OsPIP2;2), LOC_Os04g44060 (OsPIP2;3), LOC_Os07g26630 (OsPIP2;4), LOC_Os07g26660 (OsPIP2;5), OsPIP2;6 (LOC_Os04g16450), LOC_Os09g36930 (OsPIP2;7), LOC_Os03g64330 (OsPIP2;8), Os05g0438800 (OsActin1), LOC4337055 (OsPPO1), LOC432412 (OsPPO2), LOC4326969 (OsPOD1), LOC4340091 (OsSOD), LOC4342317 (OsPR1a), LOC_Os03g18850 (OsPR10), LOC4344711 (OsmaMYB), and MGG_05850 (MoPot2).

Supplemental data

The following supplemental materials are available.

Supplemental Figure S1. The expression profiles of OsPIPs in response to PXO99^A.

Supplemental Figure S2. H₂O₂ transport mediated by OsPIPs in yeast.

Supplemental Figure S3. Confirmation of the OsPIP2;2 mutation and overexpression lines.

Supplemental Figure S4. Contribution of OsPIP2;2 in H₂O₂-mediated cell elongation.

Supplemental Figure S5. Growth and development of transgenic lines in the greenhouse.

Supplemental Figure S6. Analysis of the H₂O₂ content in rice after infection with PXO99^A.

Supplemental Figure S7. Analysis of the H₂O₂ content in rice after infection with RS105.

Supplemental Figure S8. Analysis of the H₂O₂ content in rice after infection with HB3.

Supplemental Figure S9. Protein alignment of several PIPs proteins showing conserved S125 residue in OsPIP2;2.

Supplemental Figure S10. Phosphorylation of OsPIP2;2 at S125 does not change its expression level.

Supplemental Figure S11. Phylogenetic analysis of maMYBs.

Supplemental Figure S12. Promotion of plant resistance by OsmaMYB C terminus.

Supplemental Figure S13. Association between OsmaMYB and OsPIPs.

Supplemental Table S1. List of primers used in this study.

 

M.Z. and H.D. conceived and supervised this project. M.Z. performed most of the experiments, analyzed the data, and wrote the manuscript. H.S., N.L., N.W., Y.T., and L.Z. helped identify the mutant. J.P. and X.C. helped manage rice. M.Z. and H.D. analyzed the data and wrote the manuscript.

The authors responsible for distribution of materials integral to the findings presented in this article in accordance with the policy described in the Instructions for Authors (https://academic.oup.com/plphys/pages/General-Instructions) are: Meixiang Zhang (meixiangzhang@njau.edu.cn) and Hansong Dong (hsdong@njau.edu.cn).