Introduction

Plant disease resistance can be classified as being either genetic or molecular. At the level of genetics, it can be divided into qualitative (complete) resistance conferred by major disease resistance (MR) genes and quantitative (incomplete) resistance mediated by quantitative trait loci (QTLs) or multiple defense-related genes (Kou and Wang 2010; Zhang and Wang 2013; Ke et al. 2017). At the molecular level, disease resistance can be explained by a two-tiered innate immune system consisting of pattern-triggered immunity (PTI) and effector-triggered immunity (ETI) (Jones and Dangl 2006). In general, the PTI initiated by plasma membrane-localized pattern recognition receptors (PRRs) is a weak or quantitative resistance, and ETI initiated by cytoplasm-localized nucleotide-binding–leucine-rich repeat (NB-LRR) disease resistance (R) proteins is a race-specific high level of resistance or qualitative resistance (Jones and Dangl 2006; Dodds and Rathjen 2010). However, strong PTI and weak ETI also exist (Thomma et al. 2011; Zhang and Wang 2013).

Bacterial blight, caused by Xanthomonas oryzae pv. oryzae (Xoo), is the most devastating bacterial disease in rice worldwide. One of the most efficient and cost-effective ways to control Xoo infection is to develop rice with broad-spectrum and durable disease resistance (Kou and Wang 2010). Rice qualitative resistance to Xoo can be conferred by MR gene Xa21. Xa21 encodes plasma membrane-localized LRR receptor kinase, which belongs to the largest class of PRRs, and confers a broad-spectrum and strong PTI to Xoo (Song et al. 1995; Monaghan and Zipfel 2012; Pruitt et al. 2015). Another rice MR gene Xa3/Xa26 conferring Xoo resistance encodes a similar protein (Sun et al. 2004; Xiang et al. 2006), suggesting that it may also confer PTI to Xoo. Furthermore, Xa3/Xa26 has a Xoo resistance spectrum that differs from Xa21, and it confers a broad-spectrum and durable disease resistance at both seedling and adult stages (Zhao et al. 2009; Gao et al. 2010; Li et al. 2012). This gene has played an important role in rice production in China for a long time (Gao et al. 2010), but the mechanism of Xa3/Xa26-mediated resistance to Xoo remains to be elucidated. Exploring the mechanism of Xa3/Xa26-mediated resistance may provide breeders a better way to use this gene.

Previous study has revealed that Xa3/Xa26 has a dosage effect: the higher its expression, the more resistant the rice plant (Cao et al. 2007). Asian cultivated rice consists of two major groups (subspecies), indica and japonica. Xa3/Xa26, which was first characterized in indica cultivar Minghui 63, functions better with a japonica background than with an indica background in resistance to Xoo. This resistance is associated with a higher level of Xa3/Xa26 expression throughout the growth stages based on the study of transgenic plants with a japonica background carrying transgene Xa3/Xa26 driven by its native promoter (Sun et al. 2004; Cao et al. 2007). However, it is not clear whether this higher level of Xa3/Xa26 expression represents its natural expression level, for Cauliflower mosaic virus 35S promoter in the transformation construct aiming to regulate marker gene may have a residual effect on the expression of the target transgene (Yoo et al. 2005; Zheng et al. 2007; Singer et al. 2010; Yang et al. 2011).

Phytohormones play important roles in plant–pathogen interactions (Fu and Dong 2013). Salicylic acid (SA) and/or jasmonic acid (JA) have been reported to be associated with some types of defense-responsive gene-regulated resistance to Xoo (Qiu et al. 2007; Yuan et al. 2007; Tao et al. 2009; Xiao et al. 2009; Shen et al. 2010, 2011; Deng et al. 2012; Ke et al. 2014; Hu et al. 2017). However, indole-3-acetic acid (IAA), the main type of auxin in plants, and abscisic acid (ABA) promote rice susceptibility to Xoo (Ding et al. 2008; Fu et al. 2011; Xu et al. 2013). Phytoalexins, which are low-molecular-weight compounds with antimicrobial activity, are a group of plant disease resistance products produced after pathogen infection (Van Etten et al. 1994). Previous study revealed that Xa3/Xa26-mediated resistance to Xoo is associated with increases in SA and JA levels and the flavonoid phytoalexin content but not the common terpenoid phytoalexin content in comparison to a susceptible cultivar with an indica rice background (Liu et al. 2012). However, it is unknown whether defense-related phytohormones and phytoalexins contribute to genetic background-influenced function of Xa3/Xa26 against Xoo.

To address these uncertainties, we generated both japonica and indica near-isogenic lines carrying Xa3/Xa26 by continuous backcrossing and used resistant and susceptible rice lines with the same genetic background to compare the levels of resistance and the concentrations of phytohormones and metabolites putatively involved in plant–pathogen interactions. Our results suggest that the japonica background does facilitate the resistance function of Xa3/Xa26, which is associated with increased expression of Xa3/Xa26, compared with the indica background. Xa3/Xa26-mediated resistance to Xoo with a japonica background is more strongly associated with JA than SA and a dramatic increase of both terpenoid and flavonoid phytoalexins, but there is no obvious association with reinforcement of the cell wall. In addition, genetic background-influenced function of Xa3/Xa26 is associated with the WRKY45 locus, and this locus can also slightly reduce rice susceptibility to Xoo strain that is compatible to Xa3/Xa26.

Materials and methods

Rice materials

We generated two sets of rice near-isogenic lines containing MR gene Xa3/Xa26 plus allelic gene WRKY45-1 or WRRY45-2 by using indica rice cultivar (Oryza sativa ssp. indica) Minghui 63 and japonica rice cultivar (O. sativa ssp. japonica) Mudanjiang 8 as donor and also recurrent parents. Minghui 63, moderate resistance to Xoo, was used as a donor parent of Xa3/Xa26 and WRKY45-2 genes, and Mudanjiang 8, susceptible to Xoo, was used as a donor parent of WRKY45-1. MD1 and MD2 were japonica Mudanjiang 8 near-isogenic lines containing Xa3/Xa26, which was created by backcrossing (BC) eight times, using Minghui 63 as a donor parent and Mudanjiang 8 as a recurrent parent. To start, Mudanjiang 8 was crossed with Minghui 63 to produce an F2 population. In this population, plants that appeared similar to Mudanjiang 8 and carried Xa3/Xa26 and WRKY45-1 or WRKY45-2 were chosen as male parents to backcross with Mudanjiang 8 to produce BC1F1 plants. The BC1 plants were consecutively backcrossed with Mudanjiang 8 seven times to generate BC8F1 plants. MD1 and MD2 were BC8F3 lines; MD1-1 and MD1-2 carried Xa3/Xa26 and WRKY45-1, and MD2-1 and MD2-2 carried Xa3/Xa26 and WRKY45-2. MH1 and MH2 were indica Minghui 63 near-isogenic lines, which were created by the same backcross method, using Mudanjiang 8 as a donor parent and Minghui 63 as a recurrent patent. MH1 and MH2 were BC6F3 lines; MH1-1 and MH1-2 carried Xa3/Xa26 and WRKY45-1, and MH2-1 and MH2-2 carried Xa3/Xa26 and WRKY45-2. Three pairs of primers were used to examine the existence of Xa3/Xa26, WRKY45-1, or WRKY45-2 in near-isogenic lines (Supplemental Table 1).

Transgenic line Rb49, which has high resistance to Xoo, was generated by transforming Xa3/Xa26 driven by its native promoter into japonica rice cultivar Mudanjiang 8 (Sun et al. 2004; Xiang et al. 2006). The indica cultivar Minghui 63 carrying Xa3/Xa26 has moderate resistance to Xoo (Sun et al. 2004; Liu et al. 2011). The Wase Aikoku 3 is a japonica rice cultivar in which Xa3/Xa26 has been first identified (Ezuka et al. 1975). The IRBB3 is an indica near-isogenic rice line carrying Xa3/Xa26 gene (Ogawa et al. 1988; Xiang et al. 2006).

Disease evaluation

The leaves of rice plants were inoculated with Xoo strains PXO61 and PXO99 by the leaf-clipping method at the seedling (six-leaf) or the booting (panicle development) stage (Chen et al. 2002). Disease was scored by measuring the lesion length (cm) at 14 days after inoculation.

Hormone treatment

Rice plants were grown in the greenhouse until the six-leaf stage. Benzothiadiazole (BTH) (250 μM BTH [B10900, Sigma] in 0.05% [v/v] methanol plus 0.05% [v/v] Tween 20), JA (250 μM JA [J2500, Sigma] in 0.05% (v/v) methanol plus 0.05% [v/v] Tween 20), or BTH plus JA (250 μM JA and 250 μM BTH in 0.05% [v/v] methanol plus 0.05% [v/v] Tween 20) was foliar sprayed until runoff. A solution containing 0.05% (v/v) methanol and 0.05% (v/v) Tween 20 was used as the mock-treatment control. The sprayed plants remained sealed in separated transparent plastic shades for 24 h before and after treatment, respectively. The plants were then inoculated with Xoo strain PXO61.

Quantification of phytohormone and metabolite

Approximately 3-cm-long fragments next to the bacterial inoculation sites were collected from flag leaves at the booting stage. Samples were prepared and quantified using the ultrafast liquid chromatography–electrospray ionization tandem mass spectrometry system as previously reported (Liu et al. 2012).

Gene expression analysis

Quantitative reverse transcription (qRT)-PCR analysis was conducted as described previously (Qiu et al. 2007). PCR primers are listed in Supplementary Table 1. The expression level of the rice actin gene was used as an internal control. The expression level relative to control is presented. Each analysis was repeated biologically at least twice with similar results, and each biological repeat had three technical replicates. Only one repeat is presented.

Statistical analysis

Statistical analysis between two samples was performed using Student’s t test in Excel (Microsoft, Redmond, WA). Statistical analysis among multiple samples was performed by one-way ANOVA using Tukey’s multiple comparison test in software R (The R project for Statistical Computing; https://www.r-project.org). The correlation analysis was also performed using R software.

Results

Analysis of the effect of japonica background on Xa3/Xa26-mediated resistance to Xoo

To find out if a higher expression level of Xa3/Xa26 is necessary for improved resistance with a japonica background, we transferred Xa3/Xa26 from the indica cultivar Minghui 63 into the susceptible japonica cultivar Mudanjiang 8 by continuous backcrossing, using Mudanjiang 8 as a recurrent parent. The resulting near-isogenic line MD1 had similar plant height, morphology, and growth period as the Mudanjiang 8 (Fig. 1a). The transgenic rice line Rb49, which carries a single copy of transgene Xa3/Xa26 driven by its native promoter with the genetic background of Mudanjiang 8 (Sun et al. 2004; Cao et al. 2007), also has morphology and a growth period similar to Mudanjiang 8 (Fig. 1a). Consistent with previous results (Sun et al. 2004; Xiang et al. 2006; Cao et al. 2007), Minghui 63 had moderate resistance to Xoo strain PXO61 with an average lesion length 16-fold longer than that of Rb49 (Fig. 1b). MD1 had high resistance to PXO61, but its average lesion length was approximately threefold longer than that of Rb49. Meanwhile, Rb49 showed the highest expression level of Xa3/Xa26 among the three rice lines, and Minghui 63 had the lowest (Fig. 1b). The Xa3/Xa26 expression level in MD1 was also significantly higher (P < 0.01) than that in Minghui 63 but obviously lower than that of Rb49 (Fig. 1b). These results suggest that a japonica background facilitates the resistance function of Xa3/Xa26, which is associated with significantly increased (P < 0.01) expression of Xa3/Xa26. However, in comparison with MD1, the expression level of Xa3/Xa26 in transgenic line Rb49 does not represent its natural expression level and the 35S promoter in the transformation construct may also induce Xa3/Xa26 expression.

Fig. 1
figure 1

The phenotype of rice lines carrying Xa3/Xa26 with a japonica and an indica backgrounds. a The morphology of different rice lines selectively carrying the major disease resistance gene Xa3/Xa26 and the defense-responsive allelic gene WRKY45-1 or WRKY45-2. Susceptible Mudanjiang 8 is the recurrent parent of near-isogenic lines MD1 and MD2. Rb49 is a transgenic line carrying transgene Xa3/Xa26 driven by its native promoter. Moderately resistant Minghui 63 is the recurrent parent of near-isogenic lines MH1 and MH2. b The response of different rice lines to Xoo strain PXO61 infection and the expression of Xa3/Xa26. Plants were inoculated with Xoo at the booting stage and disease was scored at 14 days after inoculation. The RNA samples were got before Xoo infection at the booting stage. Bars represent mean (8 to 13 leaves from three to five plants for lesion length and three replicates for gene expression) ± standard deviation (SD). The asterisks above bars indicate a significant difference between Minghui 63 and other rice lines at P < 0.01

Full size image

Previous studies report that multiple QTLs affect the genetic background-controlled disease resistance conferred by Xa3/Xa26 (Zhou et al. 2009). The peak region of a major QTL, which explained 17% of the phenotypic variation of resistance to Xoo strain PXO61, co-localizes with the defense-responsive WRKY45 locus; this resistance QTL is contributed by the allele from susceptible Mudanjiang 8 (Zhou et al. 2009; Kou et al. 2010). The WRKY45 locus has two alleles, WRKY45-1 and WRKY45-2, which encode WRKY-type transcription factors that differ by 10 amino acids, and WRKY45-1 but not WRKY45-2 can generate small interfering RNA (siRNA) from its intron (Tao et al. 2009; Zhang et al. 2016). Three japonica rice varieties (including susceptible Mudanjiang 8) carry the WRKY45-1 allele and three indica rice varieties (including moderately resistant Minghui 63) carry the WRKY45-2 allele; both WRKY45-1 and WRKY45-2 function in the Xa3/Xa26-initated defense signaling pathway (Tao et al. 2009; Zhou et al. 2009; Kou et al. 2010). The MD1 line carries WRKY45-1 in addition to Xa3/Xa26 (Fig. 1a). To examine whether the WRKY45 locus was associated with genetic background-influenced Xa3/Xa26 function during Xoo infection, we also generated near-isogenic line MD2, which carried WRKY45-2 in addition to Xa3/Xa26 with the genetic background of Mudanjiang 8 (Fig. 1a). Another set of near-isogenic lines MH1 and MH2 carried WRKY45-1 and WRKY45-2, respectively, with the genetic background of Minghui 63 carrying Xa3/Xa26 (Fig. 1a). The rice lines MD1 and MD2 had significantly more resistance to Xoo than the rice lines MH1, MH2, and Minghui 63 (Fig. 2a). Compared with susceptible Mudanjiang 8, both MD1 and MD2 lines were resistant to Xoo (Fig. 2a), but MD1 plants were significantly more resistant (P < 0.01) than MD2 plants. A similar result was also observed in near-isogenic lines with Minghui 63 background. Both Minghui 63 and MH2 plants showed a similar level of moderate resistance to Xoo, but MH1 plants had a significantly higher level (P < 0.01) of resistance to Xoo than MH2 plants.

Fig. 2
figure 2

The resistance levels of different rice lines to Xoo. Plants were inoculated with Xoo strain PXO61 for 14 days at the booting stage. Bars represent mean (15 to 20 plants for a and at least 10 leaves from two or three plants for b) ± SD. Different letters above the bars indicate significant differences at P < 0.01. a The resistance levels of different Mudanjiang 8 and Minghui 63 near-isogenic lines. b The resistance levels of three different japonica and two indica lines. The asterisks above the bars indicate a significant difference between Minghui 63 and other cultivars or between two cultivars linked by lines at **P < 0.01

Full size image

We have proved that japonica Mudanjiang 8 benefits Xa3/Xa26-mediated resistance against Xoo compared with indica Minghui 63-mediated resistance against Xoo (Figs. 1b and 2a). To ensure this situation also exists in other cultivars, we checked a japonica Wase Aikoku 3 cultivar and an indica IRBB3 line. Sequence analysis of genetic identified Xa3/Xa26 in Wase Aikoku 3 (Ezuka et al. 1975) showed this gene (including 3.4 kb genome region and 2 kb promoter region) having identical sequence (GenBank accession numbers MG641894 and MG641895 for Wase Aikoku 3) to the Xa3/Xa26 gene in IRBB3 and Minghui 63 (Sun et al. 2004; Xiang et al. 2006). The Wase Aikoku 3-mediated resistance to Xoo was similar to japonica Rb49, but higher than IRBB3 and Minghui 63 (Fig. 2b). Furthermore, Xa3/Xa26 expression level in Wase Aikoku 3 was significantly (P < 0.01) higher than that in IRBB3 and Minghui 63 (Fig. 2b). Interestingly, we found indica line IRBB3 carried WRKY45-1 (Fig. 2b). Consistent with this result, IRBB3-mediated resistance to Xoo was significantly (P < 0.01) higher than Minghui 63-mediated resistance (Fig. 2b). These results reproved that japonica Wase Aikoku 3 also can facilitate the resistance function of Xa3/Xa26 to Xoo compared with indica rice line IRBB3 and Minghui 63. Furthermore, WRKY45 alleles may also influence Xa3/Xa26 function in other rice lines, such as indica IRBB3 and Minghui 63.

In addition to having higher levels of Xa3/Xa26 expression, MD1 and MD2 had higher levels of WRKY45 expression (either WRKY45-1 or WRKY45-2) than Minghui 63, MH1, and MH2 (Fig. 3). However, the expression levels of Xa3/Xa26 between MD1 and MD2 and between MH1 and MH2 showed no obvious differences, but the expression levels of WRKY45-1 in MD1 and MH1 were significantly higher (P < 0.01) than those of WRKY45-2 in MD2 and MH2 both before and after Xoo infection (Fig. 3). Consistent with these results, the expression levels of Xa3/Xa26 between indica IRBB3 and Minghui 63 showed no obvious differences; however, expression level of WRKY45-1 in IRBB3 was significantly higher (P < 0.01) than that of WRKY45-2 in Minghui 63 (Fig. 2b). As transcription activators (Shimono et al. 2007; Cheng et al. 2015), neither WRKY45-1 nor WRKY45-2 interacted with the Xa3/Xa26 promoter in yeast cells, although this promoter harbors seven W and W-like boxes, putatively for the binding of WRKY transcription factors. Rice ST1 gene, encoding a LRR receptor kinase-like protein, is an important component in WRKY45-mediated base defense in susceptible rice (Zhang et al. 2016). In order to know whether the ST1 expression was associated with different genetic background, we checked its expression in rice lines carrying Xa3/Xa26. MD1 and MD2 had higher levels of ST1 expression than MH1 and MH2 in general (Fig. 3). However, the expression levels of ST1 between MD1 and MD2 and between MH1 and MH2 showed no obvious differences. In addition, in the rice lines carrying Xa3/Xa26, the expression of ST1 was significantly correlated (r = 0.512, n = 36, P < 0.01) with the expression of WRKY45 (Fig. 3).

Fig. 3
figure 3

The expression of Xa3/Xa26, WRKY45, and ST1 in different rice lines with Minghui 63 or Mudanjiang 8 background before (ck) and after Xoo strain PXO61 infection. Different letters above bars indicate a significant difference at the same time point at P < 0.01 analyzed by following two methods. (1) The significant difference of Xa3/Xa26 and ST1 expression at the same time point was analyzed using 10 rice lines including five indica lines and five japonica lines. (2) The significant difference of WRKY45 expression in five indica lines and five japonica lines at the same time point was separately analyzed, because the high expression level of WRKY45 in japonica lines will cover the expression difference in five indica lines

Full size image

In order to know whether WRKY45 locus only influences the resistance function of Xa3/Xa26 or not, we checked the near-isogenic lines with another Xoo strain PXO99, which is compatible with Xa3/Xa26 (Sun et al. 2004; Cao et al. 2007). In both Minghui 63 and Mudanjiang 8 backgrounds, near-isogenic lines carrying WRKY45-1 were significantly less susceptible to PXO99 (P < 0.01) than lines carrying WRKY45-2, although these near-isogenic lines were still susceptible to PXO99 (Supplemental Fig. 1).

All these results suggest that the japonica background also facilitated the expression of WRKY45 and ST1, which is associated with increased Xa3/Xa26-mediated resistance to Xoo. However, expression of Xa3/Xa26 facilitated by a japonica background may not be directly related to WRKY45 activation. Xa3/Xa26 in combination with WRKY45-1 can mediate a higher level of resistance than Xa3/Xa26 in combination with WRKY45-2 with either an indica or a japonica background. The result also suggests that WRKY45-1 can slightly promote rice resistance to Xoo strain (such as PXO99) that is compatible to Xa3/Xa26. Thus, using Xa3/Xa26 in combination with WRKY45-1 in a japonica background is the best way to improve rice resistance to Xoo.

Analysis of phytohormone response in Xa3/Xa26-mediated resistance to Xoo in japonica background

To explore the mechanism of how the japonica background facilitates Xa3/Xa26-mediated resistance to Xoo, more analyses were carried out as follows. SA and JA in combination or alone have been reported to be associated with rice resistance to Xoo mediated by different defense-related genes (Qiu et al. 2007; Shen et al. 2010; Hu et al. 2017). To clarify which hormone defense signaling contributes to Xa3/Xa26-mediated resistance with a japonica background, we quantified the levels of hormones commonly involved in plant–pathogen interactions in resistance rice lines MD1 and Rb49, which carry WRKY45-1 but differing levels of Xa3/Xa26 transcripts, and susceptible line Mudanjiang 8. Consistent with previous reports (Ding et al. 2008; Fu et al. 2011; Liu et al. 2012; Ke et al. 2014), rice basal SA level was much higher than the levels of JA, ABA, and IAA (Fig. 4a). No obvious difference of the SA level before and after Xoo infection was detected in any of the rice lines. However, the SA levels in resistance lines were significantly higher than in the susceptible line at 10 and 72 h after Xoo infection. JA and JA-isoleucine (Ile) are collectively known as jasmonates. The jasmonate levels in the three rice lines showed a similar pattern in response to Xoo infection, except that the JA level was approximately threefold higher than the JA-Ile level (Fig. 4a). The jasmonate levels were significantly induced (P < 0.01) in both resistant and susceptible lines. However, their levels were higher in resistant lines than in the susceptible line during early infection (2 h), lower in resistant lines than in the susceptible line at 10 and 24 h after infection, and higher again in resistant lines than in the susceptible line at 48 and 72 h. Furthermore, the jasmonate levels were much higher in Rb49 than in MD1 at 72 h after infection.

Fig. 4
figure 4

The interactions between phytohormones and Xa3/Xa26-mediated resistance. Plants were inoculated with Xoo strain PXO61. a The concentrations of endogenous phytohormones before and after Xoo infection in different rice lines with a Mudanjiang 8 background. The asterisks above the bars indicate a significant difference between wild type and MDl or Rb49 at the same time point at **P < 0.01 or *P < 0.05. The letter a or b above the bar indicates a significant difference between before (ck) and after Xoo infection in the same rice line at P < 0.01 or P < 0.05. b Effects of BTH and JA on rice response to Xoo infection. Plants were inoculated with Xoo after 24 h of treatment with BTH, JA, or BTH plus JA. Bars represent mean (12 to 51 plants) ± SD. Different letters above bars indicate significant differences at P < 0.01

Full size image

ABA and IAA increase rice susceptibility to Xoo (Ding et al. 2008; Fu et al. 2011; Xu et al. 2013). To examine whether Xa3/Xa26-mediated resistance was associated with suppressed accumulation of ABA and IAA, we qualified the levels of the two hormones in the same samples used for qualifying SA and jasmonates. In response to Xoo infection, both ABA and IAA levels showed a similar pattern in the three rice lines (Fig. 4a). However, the ABA level was significant lower (P < 0.01) in the two resistant lines than in the susceptible line at 48 and 72 h after Xoo infection. The IAA levels between the resistant and susceptible rice lines showed no obvious difference after Xoo infection (Fig. 4a). In summary, these results suggest that SA and jasmonates may be involved in rice resistance at different time points after Xoo infection of plants with the japonica background, and this resistance may be also associated with suppressed accumulation of ABA.

Analysis of the effects of SA and JA on Xa3/Xa26-mediated resistance

As a functional analog of SA, BTH can activate systemic acquired resistance in both monocots and dicots (Görlach et al. 1996; Lawton et al. 1996; Lee et al. 2013). As an efficient inducer of resistance, BTH has been widely used in rice instead of SA (Fitzgerald et al. 2004; Shimono et al. 2007, 2012; Bai et al. 2011; Ueno et al. 2015). To examine whether SA and JA play roles in Xa3/Xa26-initiated resistance, we analyzed the effects of BTH and exogenous JA application on rice resistance to Xoo. Consistent with previous findings (Shimono et al. 2012; Ke et al. 2014), BTH and JA pretreatments significantly reduced (P < 0.01) the susceptibility to Xoo in susceptible Mudanjiang 8, compared to a mock control (Fig. 4b). However, JA treatment showed a stronger effect than BTH treatment on Mudanjiang 8, and JA-treated Mudanjiang 8 had a significantly shorter lesion length (P < 0.01). Furthermore, pretreatment with both BTH and JA did not have additive effect on Mudanjiang 8 to Xoo infection. The BTH had no effect on resistant lines MD1 and Rb49, and BTH-treated plants had similar lesion lengths as mock-treated plants after Xoo infection (Fig. 4b). JA had a significant effect (P < 0.01) on MD1, and the average lesion length of the JA-treated MD1 plants was approximately half of that of mock-treated plants. However, JA showed no effect on Rb49 (Fig. 4b). This finding may be due to Rb49 having high resistance to Xoo and already containing a high level of endogenous jasmonates compared with MD1 (Fig. 4a). These results suggest that jasmonate involved defense signaling may play an important role in Xa3/Xa26-mediated resistance.

Analysis of the factors directly against pathogen invasion

Upon pathogen infection, a plant displays various defense responses. The cell wall is reinforced to slow down pathogen invasion or avoid second invasion, and the plant produces toxic chemicals such as phytoalexins to kill the invading pathogens (Agrios 2005). To study which final direct factors are involved in Xa3/Xa26-mediated resistance against Xoo invasion, we analyzed the concentration of phytoalexins in rice plants responding to Xoo. The terpenoid phytoalexin momilactone A, the flavonoid phytoalexin sakuranetin, and the precursor of sakuranetin, naringenin, are common phytoalexins associated with rice resistance to Xoo, although they are not always present (Padmavati et al. 1997; Qiu et al. 2008; Liu et al. 2012; Ke et al. 2014; Duan et al. 2016; Hu et al. 2017). The momilactone A levels in the susceptible Mudanjiang 8 and resistant MD1 and Rb49 were much higher than sakruranetin and naringenin levels both before and after Xoo infection (Fig. 5). Xoo infection significantly induced the accumulation of molitactone A, sakruranctin, and naringenin in all the three lines. However, the levels of the three phytoalexins were 3-, 31-, and 44-fold higher in MD1 than in Mudanjiang 8, and 3-, 74-, and 65-fold higher in Rb49 than in Mudanjiang 8 at 5 days after Xoo infection (Fig. 5).

Fig. 5
figure 5

The concentrations of phytoalexins before and after Xoo strain PXO61 infection in different rice lines with a Mudanjiang 8 background. The asterisks above the bars indicate a significant difference between Mudanjiang 8 and MDl or Rb49 at the same time point at **P < 0.01 or *P < 0.05. The letter a or b above the bar indicates a significant difference between before (ck) and after Xoo infection in the same rice line at P < 0.01 or P < 0.05

Full size image

Expansins are proteins that loosen the cell wall (Cosgrove 2005). Rice EXPA1, EXPA5, and EXPA10 genes encode α-expansins, and activating these genes increases rice susceptibility to Xoo. Resistance to Xoo mediated by rice genes GH3-8, GH3-2, and Xa4 is associated with suppressed expression of EXPA1, EXPA5, and EXPA10 (Ding et al. 2008; Fu et al. 2011; Hu et al. 2017). Cellulose is a major element of the cell wall, and it is synthesized by enzymes of the cellulose synthase (CESA) family (Somerville 2006). Rice CesA4, CesA7, and CesA9 genes are responsible for the secondary cell wall synthesis (Zhang and Zhou 2011). Rice MR gene Xa4-mediated resistance to Xoo is associated with increased expression of CesA4, CesA7, and CesA9 (Hu et al. 2017). To study whether cell wall reinforcement contributes to Xa3/Xa26-mediated resistance, we analyzed the expression of the above-mentioned genes in response to Xoo infection. In general, the expression of EXPA1, EXPA5, and EXPA10 in resistant MD1 and Rb49 lines showed similar or even higher levels than in susceptible Mudanjiang 8 (Supplemental Fig. 2). Only the expression of CesA4 and CesA9 was significantly induced in both resistant lines compared to susceptible Mudanjiang 8 at 2 h after Xoo infection among the time points examined. In addition, CesA4, CesA7, and CesA9 showed significantly higher expression levels in Rb49 at 10 h after infection than in MD1 and Mudanjiang 8 (Supplemental Fig. 2). At the other time points examined, CesA4, CesA7, and CesA9 showed similar or even lower expression levels in resistant lines than in the susceptible line (Supplemental Fig. 2). These results suggest that an accumulation of phytoalexins, but not an obvious strengthening of cell wall, contributes to Xa3/Xa26-mediated resistance to Xoo in plants with a japonica background.

Discussion

Although several genes functioning in the Xa3/Xa26-initiated defense signaling pathway have been identified, which reveals some characteristics of Xa3/Xa26-mediated disease resistance (Qiu et al. 2007; Tao et al. 2009; Xiao et al. 2009; Deng et al. 2012; Xiao et al. 2013), the mechanism of Xa3/Xa26-mediated resistance to Xoo remains to be elucidated. Based on previous studies, the current work offers insight in several ways.

First, we validated that a japonica background can enhance Xa3/Xa26-mediated resistance to Xoo by promoting Xa3/Xa26 expression compared with an indica background. In a previous study, we examined the response of japonica transgenic lines carrying Xa3/Xa26 driven by its native promoter to Xoo infection and found that Xa3/Xa26 has a dosage effect; the rice japonica background facilitated Xa3/Xa26-mediated resistance, which was associated with markedly increased expression of Xa3/Xa26, compared with the Xa3/Xa26 donor of the indica cultivar Minghui 63 (Cao et al. 2007). However, it is uncertain whether the expression level of Xa3/Xa26 in the transgenic lines represents a natural expression level. In this study, we compared near-isogenic lines carrying Xa3/Xa26 in both japonica cultivar Mudanjiang 8 and indica Minghui 63 backgrounds and showed that Mudanjiang 8 background facilitates Xa3/Xa26 expression and promotes rice resistance to Xoo. This result is consistent with a previous report that Xa3/Xa26 is an important resistance gene in japonica cultivars planted in China (Xu et al. 2004). In addition, the present results suggest that the WRKY45 locus appears to be associated with Xa3/Xa26 expression facilitated by Mudanjiang 8 background. The WRKY45 locus functions in the Xa3/Xa26-initiated defense pathway (Tao et al. 2009; Zhou et al. 2009), and both WRKY45-1 (named WRKY45 by Shimono et al. (2007)) and WRKY45-2 are transcription activators (Shimono et al. 2007; Cheng et al. 2015). The two WRKY45 alleles showed higher expression levels in plants with Mudanjiang 8 background than in those with Minghui 63 background as the expression of Xa3/Xa26 (Fig. 3). However, the near-isogenic lines carrying Xa3/Xa26 and WRKY45-1 had a higher resistance level than those carrying Xa3/Xa26 and WRKY45-2 with both Mudanjiang 8 and Minghui 63 backgrounds. The resistance was associated with the WRKY45-1 expression level being higher than the WRKY45-2 expression level but not obviously associated with the Xa3/Xa26 expression level (Fig. 3). These results suggest that WRKY45 may not directly regulate Xa3/Xa26 expression. This hypothesis is also supported by the finding that neither WRKY45-1 nor WRKY45-2 interacted with Xa3/Xa26 promoter in yeast cells.

Although the present results are mainly from analyzing two sets of near-isogenic lines, one japonica Mudanjiang 8 background and one indica Minghui 63 background, previous reports support that other japonica and indica cultivars’ backgrounds also influence Xa3/Xa26 function. Xa3/Xa26 driven by its native promoter was also transformed into another two japonica cultivars (carrying WRKY45-1), Zhonghua 11 and 02428 (Cao et al. 2007). These japonica Xa3/Xa26-transgenic lines with Zhonghua 11 or 02428 background had higher level of resistance to Xoo than that of indica Minghui 63, and this high level of resistance was associated with increased expression of Xa3/Xa26. Furthermore, indica rice line IRBB3 also carries Xa3/Xa26 (Xiang et al. 2006). These Xa3/Xa26-transgenic lines with Zhonghua 11 or 02428 background showed higher resistance to Xoo than that of IRBB3 (Cao et al. 2007). We also examined a japonica Wase Aikoku 3 cultivar and an indica IRBB3 line both carrying Xa3/Xa26. The Wase Aikoku 3-mediated resistance to Xoo was significantly higher (P < 0.01) than IRBB3 and also Minghui 63 (Fig. 2b). Furthermore, Xa3/Xa26 expression level in Wase Aikoku 3 was significantly higher (P < 0.01) than that in IRBB3 and Minghui 63 (Fig. 2b). All the results suggest that different japonica cultivars may suit higher efficient use of Xa3/Xa26 against Xoo.

In susceptible rice that does not carry Xa3/Xa26, activation of the WRKY45-1 gene leads to the rice plants being more susceptible to Xoo. This activation is associated with suppression of the ST1 gene, while activation of WRKY45-2 gene reduces rice susceptibility to Xoo, which is associated with increase of ST1 expression (Tao et al. 2009; Zhang et al. 2016). However, in the same susceptible background, activation of the modified WRKY45-1 gene, in which siRNA-generating insertion in the intron is omitted, results in reduced susceptibility to Xoo associated with increase of ST1 (Zhang et al. 2016). This result is due to the siRNA being able to suppress ST1 by RNA-directed DNA methylation. However, in the rice lines carrying Xa3/Xa26, the expression of ST1 was significantly correlated (P < 0.01) with expression of WRKY45 (Fig. 3). Thus, further study is required to determine whether WRKY45-1 may function differently in susceptible and resistant rice backgrounds.

Second, jasmonate signaling plays an important role in Xa3/Xa26-mediated resistance. Among 40 named MR genes against Xoo, only Xa3/Xa26 and Xa4 have been reported to confer durable resistance (Leach et al. 2001; Li et al. 2012; Hu et al. 2017; Ke et al. 2017). Xa4 encodes a cell wall-associated kinase, and reinforcing the cell wall contributes to Xa4-mediated resistance (Hu et al. 2017). This resistance is associated with increased accumulation of JA-Ile but not of JA. Xa3/Xa26-mediated resistance is likely through PTI as MR gene Xa21-mediated resistance to Xoo (Zhao et al. 2009; Ke et al. 2017). In contrast, Xa3/Xa26-mediated resistance is associated with increased accumulation of both JA and JA-Ile. Thus, MR gene-mediated durable resistance to Xoo can be regulated by different signaling pathways.

Last, the direct factor against Xoo invasion in Xa3/Xa26-mediated resistance may mainly be the accumulation of phytoalexins in plants with a japonica background. Xa4-initated defense response to Xoo infection is associated with both an accumulation of the phytoalexins molilactone A and sakruranetin and a strengthening of the cell wall through promotion of the expression of cellulose synthesis genes CesAs and suppression of the expression of cell wall expansin genes EXPAs (Hu et al. 2017). Xa3/Xa26-mediated resistance to Xoo is also associated with accumulation of molilactone A and sakruranetin in japonica rice lines (Fig. 5), but it is only associated with accumulation of sakruranetin not molilactone A in an indica rice line compared to susceptible controls (Liu et al. 2012). Previous studies have revealed that JA-Ile is indispensable for sakuranetin production and required in part for momilactone production in rice (Riemann et al. 2013; Shimizu et al. 2013). In this study, JA-Ile and phytoalexin levels in transgenic line Rb49, which overexpresses Xa3/Xa26, are significantly higher (P < 0.05) than those in near-isogenic line MD1 after Xoo infection, respectively. These higher levels may contribute to Rb49 having a higher resistance level than MD1. All these results suggest that accumulation of both molilactone A and sakruranetin may explain Xa3/Xa26 functioning better in a japonica background than in an indica background. In addition, the dosage effect of Xa3/Xa26 may be associated with accumulation of more phytoalexins in the rice–Xoo interaction.

Rice resistance to Xoo regulated by other genes is also reported to be associated with suppressed expression of EXPAs (Ding et al. 2008; Fu et al. 2011). However, the expression of EXPAs was not obviously suppressed in Xa3/Xa26-mediated resistance compared to the susceptible control (Supplemental Fig. 2). Furthermore, activation of CesAs was only detected during early infection of the resistance rice lines. These results suggest that reinforcement of the cell wall may not be the major contributor against Xoo infection in Xa3/Xa26-mediated resistance.

In summary, the present results indicate that the durable resistance gene Xa3/Xa26 is more appropriate for rice improvement with a japonica background by traditional breeding methods. The function of Xa3/Xa26 is better in combination with the WRKY45-1 allele with either a japonica or indica background.