Rice ubiquitin‐conjugating enzyme OsUBC26 is essential for immunity to the blast fungus Magnaporthe oryzae

Abstract The functions of ubiquitin‐conjugating enzymes (E2) in plant immunity are not well understood. In this study, OsUBC26, a rice ubiquitin‐conjugating enzyme, was characterized in the defence against Magnaporthe oryzae. The expression of OsUBC26 was induced by M. oryzae inoculation and methyl jasmonate treatment. Both RNA interference lines and CRISPR/Cas9 null mutants of OsUBC26 reduced rice resistance to M. oryzae. WRKY45 was down‐regulated in OsUBC26 null mutants. In vitro E2 activity assay indicated that OsUBC26 is an active ubiquitin‐conjugating enzyme. Yeast two‐hybrid assays using OsUBC26 as bait identified the RING‐type E3 ligase UCIP2 as an interacting protein. Coimmunoprecipitation assays confirmed the interaction between OsUBC26 and UCIP2. The CRISPR/Cas9 mutants of UCIP2 also showed compromised resistance to M. oryzae. Yeast two‐hybrid screening using UCIP2 as bait revealed that APIP6 is a binding partner of UCIP2. Moreover, OsUBC26 working with APIP6 ubiquitinateds AvrPiz‐t, an avirulence effector of M. oryzae, and OsUBC26 null mutation impaired the proteasome degradation of AvrPiz‐t in rice cells. In summary, OsUBC26 plays important roles in rice disease resistance by regulating WRKY45 expression and working with E3 ligases such as APIP6 to counteract the effector protein AvrPiz‐t from M. oryzae.

. For example, rice Spl11 is an E3 ligase and functions as a negative regulator of programmed cell death and disease resistance (Zeng et al., 2004). Rice APIP6 and APIP10 are RING-type E3 ligases that interact with AvrPiz-t effector from Magnaporthe oryzae and promote the ubiquitination and degradation of AvrPiz-t, therefore positively regulating rice immunity to blast fungus (Park et al., 2012(Park et al., , 2016. Rice Xa21-binding protein 3 (XB3) is another RING-type E3 ligase that is required in Xa21-mediated resistance to Xanthomonas oryzae pv. oryzae .
In addition to ubiquitin E3 ligases, a number of plant E2 genes have been identified and functionally characterized in diverse biological processes. The overexpression of VrUBC1, AhUBC2, and GmUBC2 enhanced the drought tolerance of transgenic Arabidopsis thaliana (Chung et al., 2013;Wan et al., 2011;Zhou et al., 2010).
Ectopic expression of wild rice, Oryza grandiglumis, ubiquitinconjugating enzyme 1 (OgUBC1) in Arabidopsis increases resistance to Botrytis cinerea infection (Jeon et al., 2012). Triticum aestivum ubiquitin-conjugating enzyme 4 (TaU4) is a negative regulator of wheat defence against Septoria, and virus-induced gene silencing of TaU4 results in delayed disease symptom development (Millyard et al., 2016). A group of E2 enzymes in tomato is essential for plant immunity, and gene silencing of the homologs in tobacco reduced pattern-triggered immunity (Zhou et al., 2017). Although these reports indicate that E2 enzymes are involved in plant defence, their specific mechanisms still need to be revealed.
There are 48 members of E2s in rice plants that have been classified into 15 groups according to Arabidopsis and human E2s (Bae & Kim, 2014). However, none of them have been characterized in rice immunity. In this study, we characterize a group VII E2 enzyme, OsUBC26, in rice immunity by using reverse genetic and biochemical methods. CRISPR/Cas9 mutants of OsUBC26 significantly compromised rice resistance to M. oryzae. We also screened a rice cDNA library and found that RING-type E3 ligase UCIP2 is a working partner of OsUBC26. In addition, OsUBC26 could work with APIP6 and ubiquitinate AvrPiz-t in vitro. Our work revealed a ubiquitin-conjugating enzyme in rice that is essential for rice immunity.

| OsUBC26 expression was induced by M. oryzae infection and methyl jasmonate treatment
Our previous study indicated that the expression of OsUBC26 can be induced by M. oryzae, X. oryzae pv. oryzae, and Rhizoctonia solani infection (Zhao et al., 2008). To further investigate the expression profile of OsUBC26, quantitative reverse transcription PCR (RT-qPCR) was used to quantify the expression level of OsUBC26 at different time points during rice blast infection. As illustrated in Figure 1a, OsUBC26 responds to rice blast fungus from a very early stage, as the expression level of OsUBC26 increased quickly after the challenge by M. oryzae with a peak at 6 hr postinoculation (hpi) and then gradually decreased. As salicylic acid (SA) and jasmonic acid (JA) are two major phytohormones involved in disease resistance, SA and methyl jasmonate (Me-JA) were used to treat rice seedlings. We found that SA had no obvious effect on the expression of OsUBC26 (data not shown). On the other hand, the expression of OsUBC26 increased to the highest level at 12 hours post-treatment (hpt) and was approximately 13 times higher than the expression level observed at 0 hpt after the Me-JA treatment (Figure 1b). This result suggests that OsUBC26 selectively participates in the Me-JA-mediated defence response of rice.

| OsUBC26 is important for rice immunity against M. oryzae infection
To characterize the role of OsUBC26 in rice resistance to M. oryzae OsUBC26 RNAi lines were generated in the background of Kitaake.
Transcription analysis showed that the expression of OsUBC26 in the RNAi lines was approximately half that of wild-type plants  Figure S1).
Homozygous mutants were selected in the T 3 generation. The punch inoculation method was used for disease resistance evaluation (Ono et al., 2001;Park et al., 2012). The result showed that the lesion size and relative fungal biomass were significantly larger and greater in OsUBC26 null mutants compared to wild-type Nipponbare

| Null mutation of OsUBC26 down-regulated the expression of WRKY45
To reveal why CRISPR/Cas9 null mutants of OsUBC26 compromised rice resistance to M. oryzae, RNA-Seq of the OsUBC26 null mutant and its wild-type Nipponbare was conducted. A total of 237 genes were up-regulated and 215 genes were down-regulated in the OsUBC26 null mutant (Tables S1 and S2). The lower expression of WRKY45 in OsUBC26 null mutant attracted our attention.
WRKY45 is a WRKY transcription factor that positively regulates rice resistance to the blast fungus and its function is affected by nuclear ubiquitin proteasome degradation (Matsushita et al., 2013;Shimono et al., 2007Shimono et al., , 2012. The down-regulation of WRKY45 in the OsUBC26 null mutants was further confirmed by RT-qPCR ( Figure 3).

| OsUBC26 is an active ubiquitinconjugating enzyme
In the presence of ATP, ubiquitin is activated by ubiquitin activating enzyme (E1), then transferred to the active site cysteine in E2 to form a thioester bond between ubiquitin and E2 (Smalle & Vierstra, 2004;Varshavsky, 1997;Wenzel et al., 2011;Ye & Rape, 2009). This thioester bond is sensitive to the thiol-reducing agent dithiothreitol (DTT) and this characteristic can be used for E2 activity assay (Kraft et al., 2005). In an effort to prove that OsUBC26 is an active ubiquitin-conjugating enzyme, histidine-tagged Arabidopsis E1 com- This experiment demonstrates that OsUBC26 is an active ubiquitinconjugating enzyme.

| OsUBC26 interacts with UCIP2
To further reveal the working partner of OsUBC26, a yeast twohybrid screening was conducted using OsUBC26 as bait. Several E3 ligases that interact with OsUBC26 were identified from a rice cDNA library in the yeast two-hybrid assay (Table 1). Among them, Values are the means of three replicates and error bars represent SEM. Significant difference, **p < 0.01

F I G U R E 4
In vitro E2 activity assay of OsUBC26. E2 charging assays of His:OsUBC26 under nonreduced and reduced conditions. Ubiquitin bound to OsUBC26 under nonreduced conditions but not reduced conditions RING-type E3 ligase UCIP2 (OsUBC26 Interact Protein 2) represented 17 of 44 of the identified interacting clones. We cloned the full-length coding sequence of UCIP2 and confirmed the interaction between OsUBC26 and UCIP2 in yeast two-hybrid assay ( Figure 5a).
In addition, glutathione S-transferase (GST)-tagged OsUBC26 and maltose-binding protein (MBP)-tagged UCIP2 were prepared in a prokaryotic expression system. A GST pull-down assay showed that GST:OsUBC26 interacted with MBP:UCIP2 ( Figure 5b). Thus, OsUBC26 may have preference for UCIP2 and mediated the ubiquitination degradation of UCIP2-recruited substrates. Interestingly, the protein ID of UCIP2 (LOC_Os01g47740) is the same as OsZFP1, which has been reported as a working partner of OsDjA6 (Zhong et al., 2018). OsDjA6, a chaperone DnaJ protein, negatively regulates WRKY45 expression and rice immunity to M. oryzae (Zhong et al., 2018). Thus, OsUBC26 as a conjugating enzyme works with E3 ligase UCIP2/OsZFP1 and chaperone DnaJ protein OsDjA6 to modulate rice innate immunity through regulation of WRKY45 expression.

M. oryzae
To reveal the role of UCIP2 in rice immunity to M. oryzae, the CRISPR/Cas9 null mutants of UCIP2 were generated. A homozygous mutant was selected in the T 2 generation with an 11 bp deletion of the target 1 sequence ( Figure S2). The 11 bp deletion causes a premature stop codon in UCIP2. Punch inoculation of the UCIP2 null mutant showed compromised resistance to M. oryzae with larger lesion size and greater relative fungal biomass compared to Nipponbare ( Figure 6).

| OsUBC26 could work with APIP6 and ubiquitinate AvrPiz-t
During rice-pathogen interaction, E3 ligase activity is widely used by the host to reprogramme the cell to counteract pathogens (Ning et al.,

Gene Code
Gene_locus Gene_description  (Park et al., 2012). In an effort to find substrate proteins of UCIP2, we screened the rice cDNA library using UCIP2 as bait and found that APIP6 is a binding partner of UCIP2. An in vitro ubiquitination assay showed that in the presence of Ub, AtE1, OsUBC26, and both APIP6 and UCIP2 were ubiquitinated ( Figure S3), which suggests that OsUBC26 may also work with APIP6. We therefore checked whether OsUBC26 could work with APIP6 and ubiquitinate AvrPiz-t by using a bacterial-based synthetic approach (Han et al., 2017). This assay demonstrated that ubiquitination of AvrPiz-t required the presence of AtE1, OsUBC26, APIP6, and ubiquitin (Figure 7). This indicates that OsUBC26 could work with APIP6 and ubiquitinate AvrPiz-t in vitro.

| OsUBC26 null mutation impairs the proteasome degradation of AvrPiz-t
As OsUBC26 could work with APIP6 and mediate the ubiquitination of AvrPiz-t, we wanted to know whether OsUBC26-APIP6 mediates the degradation of AvrPiz-t. Cell-free degradation was conducted and the results showed that the proteasome degradation of AvrPiz-t was significantly impaired in the OsUBC26 null mutant group compared to the wild-type Nipponbare group (Figure 8). When MG132, a specific proteasome inhibitor, was added to the Nipponbare group the degradation of AvrPiz-t was blocked. This experiment indicated that the degradation of AvrPiz-t in rice cells was via the 26S proteasome system and OsUBC26 was involved in the ubiquitination and degradation of AvrPiz-t.

F I G U R E 5
OsUBC26 interacts with UCIP2. (a) OsUBC26 interacts with UCIP2 in yeast two-hybrid assay. The full-length coding sequence of UCIP2 was cloned and fused to the GAL4-activation domain. Cotransformation with BD-UBC26 further confirmed the interaction between OsUBC26 and UCIP2. (b) OsUBC26 interacts with UCIP2 in glutathione S-transferase (GST) pull-down assay. GST-tagged OsUBC26 was expressed in Escherichia coli and conjugated to glutathione magarose beads, which were used for maltose-binding protein (MBP):UCIP2 pull-down assay

| Plant materials and growth conditions
Rice seeds were soaked into water for 48 hr at room temperature and the water was changed every 12 hr. The seeds were pregerminated in an incubator at 37 °C for 1 day. After germination, the seeds were planted in small pots and kept in a greenhouse at 28 °C light/23 °C dark and 70% relative humidity under a photoperiod of 14 hr light/ 10 hr dark for growth.

| Spraying inoculation and Me-JA treatment
M. oryzae isolate Guy11 was pregrown on rice bran medium for 10 days in the dark. The aerial hyphae were flattened using a sterilized glass slide and the growth plates were incubated under light to encourage a uniform crop of conidia. The concentration of conidia was adjusted to 1-2 × 10 5 spores/ml with 0.02% Tween-20. Three- week-old rice seedlings were used for spray inoculation. After spraying, rice seedlings were kept in dark and high humidity conditions for 24 hr. The seedlings were then moved to a solar greenhouse with high humidity and grown for 7 days. ImageJ software was used to quantify the lesion area. For Me-JA treatment, 3-week-old rice seedlings were sprayed with 100 µM Me-JA, then kept in an incubator at 28 °C with high humidity. Samples were taken at time points of 0, 1, 3, 6, 12, and 24 hpt for RT-qPCR analysis.

| Punch inoculation and disease resistance evaluations
For punch inoculation, M. oryzae isolate FJ86-CT was used. The concentration of the fungal spores was adjusted to 5 × 10 5 spores/ ml with 0.02% Tween-20 and 7 µl of spore suspension was used for punch inoculation as described by Zhong et al. (2018). After punch inoculation, the seedlings were kept in a growth chamber at 28 °C light/25 °C dark and 90% relative humidity under a photoperiod of 12 hr light/12 hr dark. After 7-9 days' growth, the diseased leaves were sampled for relative fungal biomass quantitation using the DNA-based quantitative PCR method as described by Park et al. (2012).

| RNA isolation and RT-qPCR
RNA was extracted by using a total RNA extraction kit (Promega) following the manufacturer's instructions. The RNA concentration was determined by NanoDrop 2000 machine (Eppendorf). A PrimeScript

RT-PCR Kit (Takara) was used for reverse transcription and TB Green
Premix Ex Taq (Takara) was used for qPCR analysis following the manufacturer's instructions. Primers are shown in Table S3.

| RNAi
For building of the OsUBC26 RNAi line, the pTCK-303 plasmid and primer pair UBC26-RNAi-F/UBC26-RNAi-R was used. Sense sequence ligation using SpeI/SacI enzyme digestion sites and subsequent KpnI/BamHI enzyme digestion sites was used to obtain an insert for antisense sequence ligation. Agrobacterium tumefaciens LBA4404 was used for the transformation of rice cultivar Kitaake callus. All the transgenic seedlings were confirmed by PCR and βglucuronidase (GUS) staining analysis. The T 2 generation of RNAi lines was used for the RT-qPCR and inoculation experiments.

| Gene editing
To generate the gene null transgenic plants, two targets were designed using the CRISPR-GE toolkit (Xie et al., 2017). The two targets were ligated to promoters and inserted into the pYLCRISPR/Cas9-MTmono vector (Ma et al., 2015). A. tumefaciens EHA105 was used for the transformation of rice cultivar Nipponbare callus. The gene editing result was further checked by sequencing of the amplified targeted sequence. Homozygotes were selected for phenotypic analysis.

F I G U R E 8
OsUB26 null mutation impairs proteasome degradation of AvrPiz-t in rice cells. Total protein was extracted from OsUBC26 null mutant and Nipponbare with or without MG132. Purified proteins GST-AvrPiz-t from Escherichia coli BL21 (DE3) were incubated in rice total protein supernatant with the addition of 1 mM ATP and cultured at 28 °C for different times. Actin was used as a loading control. Band intensity was calculated by ImageJ software

| Yeast two-hybrid screening
Coding region sequences of OsUBC26 and UCIP2 were constructed in pDBLeu vector as bait. 3-aminotriazole (3AT) sensitivity was de- plates to select positive clones.

| Cell-free degradation
Cell-free degradation was performed as described by Kong et al. (2015). Briefly, total protein from 3-week-old rice seedlings was extracted with native extraction buffer (50 mM Tris-MES, pH 8.0, 0.5 M sucrose, 1 mM MgCl 2 , 10 mM EDTA, pH 8.0, 5 mM dithiothreitol [DTT], and protease inhibitor cocktail) with or without 100 µM MG132. Insoluble debris was pelleted twice by centrifugation at 13,000 rpm for 10 min at 4 °C. For construction of GST-AvrPiz-t plasmid, pGEX-KG was digested with SmaI and SalI, and AvrPiz-t coding sequence was inserted in it using the recombination-type-cloning method (ClonExpress II One Step Cloning Kit; Vazyme). Purified proteins GST-AvrPiz-t from E. coli BL21 (DE3) were incubated in rice total protein supernatant with addition of 1 mM ATP and cultured at 28 °C for different times. Anti-GST antibody was used to detect the GST-AvrPiz-t protein level by immunoblotting analysis.

DATA AVA I L A B I L I T Y S TAT E M E N T
The data that support the findings of this study are available from the corresponding author upon reasonable request.