Mutation of the rice XA21 predicted nuclear localization sequence does not affect resistance to Xanthomonas oryzae pv. oryzae

Background The rice receptor kinase XA21 confers robust resistance to the bacterial pathogen Xanthomonas oryzaepv. oryzae(Xoo). We previously reported that XA21 is cleaved in transgenic plants overexpressing XA21 with a GFP tag (Ubi-XA21-GFP) and that the released C-terminal domain is localized to the nucleus. XA21 carries a predicted nuclear localization sequence (NLS) that directs the C-terminal domain to the nucleus in transient assays, whereas alanine substitutions in the NLS disrupt the nuclear localization. Methods To determine if the predicted NLS is required for XA21-mediated immunity in planta, we generated transgenic plants overexpressing an XA21 variant carrying the NLS with the same alanine substitutions (Ubi-XA21nls-GFP). Results Ubi-XA21nls-GFP plants displayed slightly longer lesion lengths, higher Xoobacterial populations after inoculation and lower levels of reactive oxygen species production compared with the Ubi-XA21-GFP control plants. However, the Ubi-XA21nls-GFP plants express lower levels of protein than that observed in Ubi-XA21-GFP. Discussion These results demonstrate that the predicted NLS is not required for XA21-mediated immunity.

We previously reported that transgenic plants overexpressing XA21 produce XA21 cleavage products (Chen et al., 2010a;Chen et al., 2010b;Park et al., 2010;Park et al., 2008;Park & Ronald, 2012;Wang et al., 2006). For example, a 110-kDa amino-terminal XA21 cleavage product was detected in Ubi-Myc-XA21 transgenic plants and a 70-kDa carboxyterminal cleavage product was observed in Ubi-XA21-CFP and Ubi-XA21-GFP plants (Park & Ronald, 2012). A predicted nuclear localization sequence (NLS) located between the XA21 transmembrane and juxtamembrane domain was identified. In rice protoplasts, the predicted NLS is able to direct transiently expressed GFP-tagged XA21 C-terminal domain to the nucleus. In contrast, a construct carrying alanine substitutions in the predicted NLS fails to direct the nuclear localization of XA21 C-terminal domain (Park & Ronald, 2012).
To assess the biological relevance of the predicted XA21 nuclear localization sequence in planta, we generated transgenic plants expressing an XA21 variant with alanine substitutions in the predicted NLS (XA21nls). The Ubi-XA21nls-GFP transgenic lines displayed partial resistance to Xoo infection, as reflected in slightly longer lesion lengths and higher Xoo bacterial populations compared with Ubi-Xa21-GFP control plants. A tyrosine-sulfated 21-amino acid derivative of RaxX (RaxX21-sY) triggered ROS production in Ubi-XA21nls-GFP plants. The expressed protein levels of XA21nls-GFP in Ubi-XA21nls-GFP lines were lower than those observed in Ubi-XA21-GFP plants. These results suggest that lower levels of expressed protein account for the slightly longer lesions observed in Ubi-XA21nls-GFP transgenic plants and indicate that the predicted NLS is not critical for XA21-mediated immunity.

Rice growth condition and leaf treatment with elicitors
Oryza sativa ssp. japonica cultivar Kitaake and a transgenic line expressing XA21 (Park & Ronald, 2012), Ubi-XA21-GFP (5B-5-4-3-2-1; the transgene driven by the maize ubiquitin promoter), were used in this study. The Kitaake genetic background lacks XA21 and is fully susceptible to the Xoo strain used in this study. Rice seeds were surfaced-sterilized with 15% bleach, rinsed with water, and soaked in water for one week. Well-germinated seedlings were transplanted into a soil mixture (80% sand, 20% peat from Redi-Gro, Sacramento, CA, USA) in an environmentally-controlled greenhouse.

Vector construction and rice transformation
The XA21nls-GFP (KRTKK678-682AATAA) construct was generated using site-directed mutagenesis. The DNA sequence containing alanine substitutions were amplified from an XA21-GFP/pENTR vector (Park & Ronald, 2012) using two overlapping primers (XA21nls-F and XA21nls-R, Table S1). The PCR product was digested with DpnI at 37 • C overnight and transformed into Escherichia coli competent cells, followed by kanamycin selection. Plasmids carrying the mutations were confirmed by standard Sanger sequencing. XA21nls-GFP was subcloned into Ubi-pCAMBIA-1300, which carries the maize ubiquitin promoter to drive expression of the transgene (Chern et al., 2005) using Gateway LR Clonase (Invitrogen, Carlsbad, CA, USA). The resulting Ubi-XA21nls-GFP construct was used to transform rice Kitaake calli by Agrobacterium-mediated method in Plant Transformation Facility at UC Davis (Hiei et al., 1994). The presence of the transgene was confirmed by PCR using gene-specific primers (Ubi-pro and XA21-seq2 in Table S1).

Bacterial infection assays
Xanthomonas oryzae pv. oryzae (Xoo) Philippines race 6 PXO99 A (referred to as PXO99) was used in this study. Xoo was cultured on peptone sucrose agar plates supplemented with 20 mg L −1 cephalexin for two days, then washed off and re-suspended in sterilized water. The concentration of bacterial suspension was adjusted to an OD 600 of 0.5 (approximately 1×10 8 cfu/mL) for inoculation.
Before inoculation, 5-week-old greenhouse-grown rice plants were transferred to a walk-in growth chamber (14-h-light/10-h-dark photo-period, 28 • C/24 • C, 80%/85% humidity) to acclimate the chamber conditions for one week. Plants were inoculated using the scissor clipping method (Song et al., 1995). For each line, 8-12 plants were inoculated and in each plant two fully-expanded leaves from 3-6 tillers were clipped using scissors with the Xoo inoculum. The lesion lengths were measured 14 days post-inoculation. In planta bacterial growth was assessed as described previously (Bahar et al., 2014). Briefly, inoculated leaves were collected at indicated time points, cut into 5-mm pieces and incubated in 10 mL sterile water with shaking at 28 • C for 1 h. The suspension was diluted accordingly and spread out on PSA plates with 20 mg L −1 cephalexin. The bacterial colonies were counted after a two-day culture at 28 • C.

Protein extraction and western blot assays
Protein extraction from rice leaves and western blot assays were performed as previously described (Park & Ronald, 2012). Briefly, total protein was extracted from 100 mg of rice leaf tissue in 200 µL of pre-chilled extraction buffer (0.15 M NaCl, 0.01 M sodium phosphate buffer pH 7.2, 2 mM EDTA, 1% Triton X-100, 10 mM DTT, 20 mM NaF, 1 mM PMSF, 1% Sigma protease cocktail) and separated in an 8% SDS-polyacrylamide gel. A mouse anti-GFP antibody (Santa Cruz Biotechnology, Santa Cruz, CA, USA) was used as the primary antibody for detection of GFP-tagged XA21 and XA21nls.

ROS assays
ROS assays were performed as previously described (Pruitt et al., 2015). Fully-expanded leaves were harvested from 4-week-old hydroponically grown rice plants, cut into 2-mm 2 pieces and floated on water overnight. Leaf pieces were treated with water, 1 µM nonsulfated 21-amino acid synthetic RaxX peptides (RaxX21-Y) or tyrosine sulfated RaxX21 peptides (RaxX21-sY). For each treatment, four biological replicates were included and in each replicate two leaf pieces were used. Chemiluminescence was recorded every 30 s for 3 h in a high-sensitivity TriStar plate reader (Berthold, Germany). in Ubi-XA21-GFP and T2 progeny derived from three independent Ubi-XA21nls-GFP T1 lines (XA21nls #1-8, #2-3 and #7-3) determined by qRT-PCR and normalized to ubiquitin reference gene. The XA21 expression level in Ubi-XA21-GFP was set to 1. Bars represent mean ± SD of three technical replicates. Different letters indicate significant differences between the groups (Tukey's honestly significant difference test, α < 0.05). The experiment was repeated three times with similar results. (C) Western blot analysis of 70 µg total protein extracts from 5-wk Ubi-XA21-GFP, Kitaake and two individual T2 plants derived from three Ubi-XA21nls-GFP T1 lines. The full-length XA21-GFP was detected with an anti-GFP antibody. Equal loading of total proteins was confirmed by Coomassie blue staining (CBB).

Generation of Ubi-XA21nls-GFP transgenic plants
The predicted NLS in XA21 is a basic amino acid-rich sequence localized between the transmembrane and juxtamembrane domains (Park & Ronald, 2012). An XA21 variant with alanine substitutions in the predicted NLS (XA21nls; Fig. 1A) was generated using site-directed mutagenesis and introduced into the rice Kitaake cultivar. Five independent Ubi-XA21nls-GFP lines were obtained, and three lines that expressed detectable full-length XA21nls-GFP protein were selected for further analysis. We observed reduced gene and protein expression levels of XA21-GFP in these transgenic lines (#1-8, #2-3 and #7-3) as compared with those in the Ubi-XA21-GFP control plants (Figs. 1B and 1C).
The positive segregants in line #1, #2 and #7 (closed columns in Fig. S1) displayed short lesion lengths, while the null segregants (open columns in Fig. S1) displayed lesions as long as the Kitaake control plants. These results demonstrate that the T1 progeny derived from the three Ubi-XA21nls-GFP lines displayed similar or slightly longer lesions than the Ubi-XA21-GFP control.
To assess Xoo growth in planta, the same plants from the first inoculation experiment were used for a bacterial growth assay. Inoculated leaves were collected 0, 4, 8 and 12 days post-inoculation (dpi). As shown in Fig. 2B, the bacterial population in Kitaake reached to 10 8 cfu/leaf at 4 dpi, whereas the populations in Ubi-XA21-GFP and Ubi-XA21nls-GFP plants reached over 10 7 cfu/leaf. At 8 dpi, the Xoo population grew in Ubi-XA21nls-GFP plants but remained unchanged in the Ubi-XA21-GFP control plants. At 12 dpi, the bacterial populations in three Ubi-XA21-GFP transgenic lines were significantly higher than the Ubi-XA21-GFP plants but lower than that in the Kitaake plants (Fig. 2B). The result of bacterial populations is consistent with the lesion lengths in Ubi-XA21nls-GFP plants, indicating that Ubi-XA21nls-GFP plants display partial resistance to Xoo as compared with the Ubi-XA21-GFP control.

RaxX21-sY triggers ROS production in Ubi-XA21nls-GFP plants
The production of ROS is a hallmark of XA21-mediated immune responses (Pruitt et al., 2015). We therefore carried out a ROS assay to test whether RaxX21-sY treatment leads to ROS production in Ubi-XA21nls-GFP plants. Detached leaves from four-week-old hydroponically-grown Ubi-XA21-GFP, Kitaake and Ubi-XA21nls-GFP plants were treated with water, 1 µM RaxX21-Y or RaxX21-sY. As shown in Fig. 3, RaxX21-sY triggered ROS production in Ubi-XA21-GFP plants but not in Kitaake. The ROS production peaked around 70 min and last for 150 min. The three Ubi-XA21nls-GFP lines displayed ROS production upon RaxX21-sY treatment, in a similar manner as the Ubi-XA21-GFP control plants. RaxX21-sY induced ROS production reached a lower level in three Ubi-XA21nls-GFP lines compared with that in Ubi-XA21-GFP plants. These results demonstrate that Ubi-XA21nls-GFP plants maintain the ROS response to RaxX21-sY.

DISCUSSION
We previously reported that XA21 is cleaved in transgenic plants overexpressing XA21 with a GFP tag (Ubi-XA21-GFP), that the released GFP tagged C-terminal domain is localized to the nucleus and that a predicted NLS directs this domain to the nucleus in transient assays (Park & Ronald, 2012). To investigate the biological relevance of these observations, here we used a genetic approach to assess the resistance of transgenic plants expressing XA21 with mutations in the predicted NLS (Ubi-XA21nls-GFP) to Xoo.
We observed that the three Ubi-XA21nls-GFP transgenic lines displayed slightly reduced resistance to Xoo, as illustrated by slightly longer lesion lengths and higher bacterial populations compared with the Ubi-XA21-GFP control plants (Fig. 2). Sulfated, but not nonsulfated, RaxX21 is able to trigger ROS production in the Ubi-XA21nls-GFP plants, to slightly lower levels than that observed in the Ubi-XA21-GFP control (Fig. 3). Considering the relatively lower XA21 transcript and protein levels of XA21 in Ubi-XA21nls-GFP plants as compared with the Ubi-XA21-GFP control plants (Figs. 1B and 1C), we hypothesize that the reduced level of XA21 protein accumulation is responsible for the slightly reduced resistance and reduced ROS levels observed in the Ubi-XA21nls-GFP plants. The differences in XA21 protein levels in Ubi-XA21-GFP vs. Ubi-XA21nls-GFP may be due to position effects of the transgene, as we and others have observed in previous studies.
Together, these results indicate that Ubi-XA21nls-GFP is able to respond to RaxX21-sY and confers resistance to Xoo, which suggest that the predicted NLS is not required for XA21-mediated immunity. This result conflicts with our previous report that disruption of XA21 nuclear localization in Ubi-XA21-GFP-NES caused enhanced susceptibility (Park & Ronald, 2012). One possible explanation for this discrepancy is that the addition of the NES to XA21 disrupts activity of XA21.
The biological role of the XA21 cleavage product remains unknown. The XA21 intracellular domain interacts with several proteins predicted to be nuclear localized, including the XB10/WRKY62 transcription factor (Seo et al., 2011). The nuclear localization of the XA21-GFP cleavage products in transient assays suggests a role in transcriptional regulation in the nucleus (Park & Ronald, 2012). However, here we demonstrate that nuclear localization of the XA21 intracellular domain is not critical for XA21-mediated immunity.
• Yuen Ting Ho performed the experiments.
• Pamela C. Ronald conceived and designed the experiments, analyzed the data, wrote the paper, prepared figures and/or tables, reviewed drafts of the paper.

Data Availability
The following information was supplied regarding data availability: