High ultraviolet-B sensitivity due to lower CPD photolyase activity is needed for biotic stress response to the rice blast fungus, Magnaporthe oryzae

Sensitivity to ultraviolet-B (UVB, 280–315 nm) radiation varies widely among rice (Oryza sativa) cultivars due to differences in the activity of cyclobutane pyrimidines dimer (CPD) photolyase. Interestingly, cultivars with high UVB sensitivity and low CPD photolyase activity have been domesticated in tropical areas with high UVB radiation. Here, we investigated how differences in CPD photolyase activity affect plant resistance to the rice blast fungus, Magnaporthe oryzae, which is one of the other major stresses. We used Asian and African rice cultivars and transgenic lines with different CPD photolyase activities to evaluate the interaction effects of CPD photolyase activity on resistance to M. oryzae. In UVB-resistant rice plants overexpressing CPD photolyase, 12 h of low-dose UVB (0.4 W m−2) pretreatment enhanced sensitivity to M. oryzae. In contrast, UVB-sensitive rice (transgenic rice with antisense CPD photolyase, A-S; and rice cultivars with low CPD photolyase activity) showed resistance to M. oryzae. Several defense-related genes were upregulated in UVB-sensitive rice compared to UVB-resistant rice. UVB-pretreated A-S plants showed decreased multicellular infection and robust accumulation of reactive oxygen species. High UVB-induced CPD accumulation promoted defense responses and cross-protection mechanisms against rice blast disease. This may indicate a trade-off between high UVB sensitivity and biotic stress tolerance in tropical rice cultivars.


Introduction
As sessile organisms, plants are inevitably exposed to multiple stresses-including ultraviolet-B (UVB, 280-315 nm) radiation and pathogens-that reduce their yield and productivity [1,2]. Rice (Oryza sativa) is one of the most important staple food grains globally and is extensively cultivated worldwide, including in Asia and Africa. However, the UVB sensitivities of rice cultivars vary widely between regions with different climates [3][4][5]. Plants can be classified into four categories based on their UVB sensitivity: (1) UVBresistant, (2) UVB-sensitive, (3) UVB-hypersensitive, and (4) UVB-super-hypersensitive [5]. Cultivars belonging to the O. sativa subspecies, japonica, are classified into the UVB-resistant or UVB-sensitive groups, whereas those belonging to indica are classified into the UVB-sensitive or UVB-hypersensitive groups [6]. Some African rice cultivars belonging to tropical O. sativa (TOS), tropical O. glaberrima (TOG), and tropical O. barthii (TOB) are classified as UVB-super-hypersensitive, whereas others are classified as either UVB-hypersensitive or UVB-sensitive. However, no African cultivars are classified as UVB-resistant. These differences in UVB sensitivity depend on the enzymatic activity of cyclobutane pyrimidine dimer (CPD) photolyase. CPD photolyase repairs CPDs, which are one of UVB-induced DNA lesion, and its activity is determined by the amino acids sequence of the CPD photolyase gene [4,5,7,8]. In a previous study, we demonstrated that UVB-induced CPDs are one of the principal causes of plant's UVB-induced growth inhibition in tiller number, height and fresh weight, and that increasing CPD photolyase activity can significantly alleviate UVB-induced growth inhibition [9,10].
A comparison of the UVB sensitivity of rice cultivars with their corresponding region of cultivation reveals that highly UVB-sensitive cultivars (UVB-hypersensitive or UVB-super-hypersensitive cultivars) with low CPD photolyase activity have been domesticated in tropical areas, such as South Asia or the African continent. This is surprising because these locations experience relatively high UVB radiation, and are expected to be more suitable for UVB-resistant cultivars [3,5,6]. However, given the long history of rice cultivation, where rice cultivars have been selected and cultivated to suit the local climate, there may be a crosslink between CPD photolyase activity and other tropical environmental stresses, and high UVB-sensitivity may be beneficial for survival in tropical areas. Indeed, UVB radiation activates pathways related to plant defensive mechanism and increases resistance to fungal diseases by inducing defense-related genes, such as pathogenesis-related (PRs) genes or the genes that induce transcription of PRs. In a series of defense responses, defense-related genesincluding β-1,3-glucanase, chitinase, phenylalanine ammonia lyase, stilbene synthase, and chalcone synthase-are activated and play crucial role for defense in the leaves of several plant species [11]. In addition, UV-induced CPDs and  photoproducts activate the transcription of the β-1,3-glucanase gene [12]. In Arabidopsis, PRs boost plant defenses against pathogenic microorganisms [13] such as phytopathogenic oomycete Hyaloperonospora parasitica [14] and necrotrophic fungus Botrytis cinerera [15]. Therefore, highly UVB-sensitive rice with low CPD photolyase activity are expected to be more resistant to disease than UVB-resistant rice with high CPD photolyase activity. However, it is unclear whether low CPD photolyase activity is advantageous for the survival of highly UVB-sensitive tropical rice cultivars in the presence of other stresses, such as fungal diseases.
The rice blast disease is caused by the rice blast fungus (the ascomycete, Magnaporthe oryzae), which is one of the most destructive pathogens in the world [16]. To find out how rice cultivars with low CPD photolyase activity can be cultivated in tropical areas, we evaluated the interaction effects of different CPD photolyase activities on the resistance mechanism of rice against M. oryzae. For our experiments, we used highly UVB-sensitive Asian and African rice cultivars, as well as CPD photolyase-overexpressing and -deficient transgenic rice. African rice); and TOG-OxPHR (a PHR transgenic line of TOG12380 with higher CPD photolyase activity) [18]. TOS13649, TOG12380, and TOB7307 were provided by the Rice Biodiversity Center for Africa (http:// eserv ices. afric arice. org/ argis/ index. php; Cotonou, Benin). The rice plants were grown (12-h photoperiod, day/night temperatures of 28 °C/22 °C, respectively) in a growth chamber under visible radiation as described before [5]. Photosynthetically active radiation (PAR) was recorded with a data logger (LI-1000; Li-Cor Inc., Lincoln, NE, U.S.A.) and an L1-190SA sensor (Li-Cor Inc.). The PAR was adjusted to approximately 350 µmol photon m -2 s -1 at the top of the plants. When necessary, a low physiological dose (0.4 W m −2 ) of UVB radiation was irradiated with visible radiation. The UVB radiation was supplied from three UVB bulbs (FL20SE; Toshiba, Tokyo, Japan) filtered through a UV29 glass filter (Toshiba Glass Co; http:// www. jgc. co. jp), which reduces 290 nm UVB radiation by 50% [5]. The UVB intensity was measured with a data logger (LI-1400) and an SD-104B sensor (Li-COR Inc.). Biologically effective UVB radiation (UVB BE ) was calculated using the plant action spectrum of Caldwell (1971) [19] normalized to unity at 300 nm. When the plants were grown with supplementary UVB radiation at 0.4 W/m −2 , the UVB BE was 4.9 kJ m −2 day −1 .

Plant material, growth conditions, and UVB treatment
To measure transcription levels, because photo repair activity is high from the third to fourth fully expanded leaves [20], seedlings were grown for 16 d under visible radiation in a large growth cabinet until the third leaves had expanded fully. Subsequently, the seedlings were grown under visible radiation with (UVB-pretreated samples) or without supplementary UVB radiation filtered through a UV29 glass filter (0.4 W m −2 ) for 12 h. Fully expanded third leaves were harvested and used in the experiments.
For the punch inoculation assay, seedlings were grown for 6-8 weeks under visible radiation in a large growth cabinet until the fifth leaves had emerged, which is suitable for punch assay, and then the seedlings were grown under visible radiation with or without supplementary UVB radiation filtered through a UV29 glass filter (0.4 W m −2 ) for 12 h. The young fifth leaves were harvested from the seedlings for assay.
For the microscopic examination and 3, 3′-diaminobenzidine (DAB) staining of leaf sheaths, seedlings were grown for 3-4 weeks under visible radiation in a large growth cabinet, and then the seedlings were grown under visible radiation with or without supplementary UVB radiation filtered through a UV29 glass filter (0.4 W m −2 ) for 12 h. The leaf sheaths were harvested and prepared for assay.
To measure paraquat sensitivity, seedlings were grown for 16 days under visible radiation in a large growth cabinet until the third leaves had expanded fully. The third leaves were harvested and used for the assay. Prior to paraquat treatment, the UVB-pretreated samples were harvested from the seedlings exposed to supplementary UVB radiation filtered through a UV29 glass filter (0.4 W m −2 ) for 12 h.

Quantitative real-time PCR (qPCR)
Total RNA was extracted from whole seedlings using RNeasy Plant Mini Kit (Qiagen, Hilden, Germany) and treated with DNase I (Qiagen). Reverse transcription was performed with an oligo (dT) primer and a random 6-mer mixture using a Prime Script RT Reagent Kit (Takara Bio Inc., Japan) in a 10 μl reaction mixture. A 1 μl sample of the cDNA was used as a template for reverse transcription (RT)-PCR and qPCR. The actin gene was used as an internal control. The primers used in this experiment are listed in Table S1: the defense-related genes included OgJiPR10, which plays an important role in biotic and abiotic stresses [21]; disease resistance protein RPM1-like (RPM1), one of the disease resistance (R) proteins [22]; OsPAL4, which contributes to broad-spectrum disease resistance [23]; OsCCR17, a key enzyme in the defense-related production of phenolic compounds and lignification [24]; dehydrogenase (CAD6) [22], which is involved in lignin biosynthesis; and OgWRKY45, a transcription factor which enhances the resistance of rice to M. oryzae [25,26]. Real-time PCR was performed using SYBR Green to monitor double-stranded DNA synthesis (CFX96, Bio-Rad Laboratories, CA, USA), and cDNA levels were determined with the delta CT method.
The resistance of 6-to 8-week-old rice plants to M. oryzae was evaluated using a previously described punch inoculation method [27] with slight modifications. Each rice leaf was lightly punctured at a spot, and 10 µl of conidia suspension (5 × 10 5 conidia ml −1 ) was added to the wound. The leaves were kept in a culture dish with sterile water containing 0.023% 6 benzylaminopurine to maintain moisture. Inoculated plants leaves were kept in the dark for 12 h and then transferred to a growth chamber under a 12/12-h (light/ dark) photoperiod. Lesions were photographed 5-10 days post inoculation (dpi), and lesion length was measured in mm using a ruler.
The fungal biomass (that is, the DNA content of M. oryzae) in infected rice leaf tissues was quantified using a previously described method [29,30] with some modifications. Briefly, DNA was extracted from a small piece of infected rice tissue (3 × 1 cm 2 ) using the standard cetyltrimethylammonium bromide (CTAB) extraction protocol. The DNA was treated with RNase A to remove RNA. DNA-based qPCR was performed using the SYBR Green (CFX96, Bio-Rad) PCR detection system. Relative fungal growth was determined using the threshold cycle value (C T ) of M. oryzae Pot2 DNA against the C T of rice genomic ubiquitin DNA [31].

Visualization of infectious growth and fungal penetration
The excised rice leaf sheaths of three rice varieties (Sa-WT, S-C, and A-S) were prepared as described previously [28], with some modifications. Briefly, leaf sheaths were obtained from intermediate-aged leaves of 3-to 4-week-old plants and cut into strips (length, ∼9 cm). Fungal conidia were harvested at a concentration of 1 × 10 5 conidia ml −1 . A 10 µl volume of the inoculum was injected into the hollow space enclosed by the sides of the leaf sheaths above the mid-vein. Inoculated sheaths were supported horizontally in a Petri dish containing wet filter paper, such that the conidia settled on the mid-vein regions. When ready for microscopy, the sheaths were hand-trimmed to remove the chlorophyllenriched plant parts and expose the epidermal layer above the mid-vein. The lower mid-vein cells were then removed to produce sections 3-4 cell layers thick. The remaining epidermal layers of the mid-vein (3-4 layers thick) were utilized for microscopic observation.

Cytological staining assay
Conidia harvested from oatmeal agar cultures were resuspended in water to a concentration of 1 × 10 5 conidia ml −1 . This suspension was used for the examination of fungal infection and for the cytological assay. Five-week-old seedlings of three rice cultivars (Sa-WT, S-C, and A-S) were used for the infection assays. To observe H 2 O 2 deposition in rice tissue, we performed DAB (Sigma-Aldrich, St. Louis, MO, USA) staining as follows. Rice leaf sheath samples were incubated in a DAB solution (1 mg ml −1 , pH 3.8) for 8 h in dark conditions. The samples were de-stained for 40 min in a de-staining solution (60% ethanol + 20% acetic acid + 20% glycerol; 3:1:1) and then examined by light microscopy [32].
To observe penetration peg formation within rice leaf tissue, aniline blue staining was performed as previously described [33]. Briefly, rice leaves were sampled 72 h after inoculation and incubated in 1 M KOH at 70 °C for 20 min. The samples were washed three times with water and received a final wash with 0.067 M K 2 HPO 4 (pH 9.0). Following this, the samples were stained with 0.05% aniline blue and examined with fluorescence microscopy (Axio Imager D1; Carl Zeiss Imaging Systems, Germany) using the DAPI filter (excitation, 365 nm; emission, 420 nm) (http:// www. olymp us-global. com).

Paraquat susceptibility test
The paraquat susceptibility test was performed as described previously [34], with slight modifications. Briefly, plants were either pretreated with or without UVB radiation filtered through a UV29 glass filter (0.4 W m −2 ) for 12 h, and the leaf discs were excised from the center of fully expended leaf blades harvested from independent rice plants. The discs were immediately dipped into different solutions (4 ml) in 12-well plastic plates. Solutions of Tween 20 (0.1%) and paraquat (1,1′-dimethyl-4,4′-bipyridinium) at different concentrations (0, 10, 50, and 100 μM) were used in this assay. Our preliminary experiments revealed that a 50 μM concentration of paraquat was optimal for detecting the differences in paraquat tolerance in all cultivars examined in this study. Note that this concentration is higher than that suggested in a previous study [34]. The disks were illuminated with lowintensity light for 20 h and photographed immediately after treatment. The degree of paraquat-induced damage in a leaf disc shown in Fig. 5 was calculated from the percentage of the area of the white discolored region per the area of each leaf disc using Image J software (https:// wsr. imagej. net/ distr os/ osx/ ij153-osx-java8. zip).

UVB radiation enhances susceptibility to M. oryzae among UVB-resistant rice plants with high CPD photolyase activity
To determine whether there is a negative correlation between UVB sensitivity and disease sensitivity among rice cultivars, we measured the basal disease resistance of rice cultivars with punch inoculation experiments [31]. The low-dose UVB-pretreated leaves of Sasanishiki (UVB-resistant rice) showed the highest susceptibility to M. oryzae. The lesion length (caused by punch inoculation) and the M. oryzae DNA contents of Sasanishiki were higher than those of other rice cultivars (Fig. S1). In contrast, the highly UVB-sensitive rice cultivars did not show increased susceptibility to M. oryzae strain Ai79-149 after UVB pretreatment (Fig. S1). Based on these results, we developed transgenic rice lines to evaluate the effect of CPD photolyase activity on the susceptibility of rice plants to Ai79-149. These transgenic lines share a genetic background with Sasanishiki or TOG12380, but exhibit differences in CPD photolyase activity. The transgenic lines were developed by inserting the CPD photolyase gene in the sense (S-C line) or antisense (A-S line) orientation into wild-type UVB-resistant Sasanishiki (Sa-WT), and in the sense (TOG-OxPHR) into wild-type UVB-sensitive TOG12380 (TOG-WT) [9,18]. Without UVB pretreatment prior to M. oryzae inoculation, there was no significant difference in the susceptibility of S-C, Sa-WT, and A-S to M. oryzae (as evaluated by comparing the lesion length and M. oryzae DNA content of infected leaves; Figs. S2a-c). Similar results were observed in TOG-OxPHR and TOG-WT, although there was difference in baseline susceptibility to Ai79-149 between the background lines of Sa-WT and TOG12380 (Figs. S2d-f). After 12 h of UVB pretreatment followed by M. oryzae inoculation, the lesion was longest in S-C, followed by Sa-WT, followed by A-S (Fig. 1a, b). In African rice, the lesion was longer in TOG-OxPHR than in TOG-WT (Fig. 1a, d). To confirm the observed differences in basal resistance, we measured the M. oryzae DNA content of infected leaves and found that the M. oryzae DNA contents of PHR transgenic lines corresponded with their respective lesion length (Fig. 1c, e). These results strongly suggest that rice plants with low CPD photolyase activity are resistant to M. oryzae infection when exposed to UVB radiation.

UVB radiation reduces the reactive oxygen species (ROS)-triggered defense responses of UVB-resistant plants (S-C and Sa-WT)
Next, we examined the mechanisms underlying resistance to rice blast fungus in rice plants with low CPD photolyase activity. ROS accumulation at the infection site-typically in response to an oxidative burst-is an important and early defense response [35,36]. As such, the increased susceptibility of UVB-pretreated S-C and Sa-WT to M. oryzae may be related to the suppressed activation of defense mechanisms in these plants. To test this hypothesis, we examined host-derived ROS by staining the leaves of S-C, A-S, and Sa-WT plants with 3, 3′-diaminobenzidine (DAB) at 48 and 72 h post inoculation (hpi). When the plants were inoculated with sterile water (without M. oryzae), DAB staining was not detected on plants with or without UVB pretreatment (Fig. 2). After inoculation with M. oryzae and without UVB pretreatment, we detected intense DAB staining and infectious hyphae in all rice lines. The DAB stain appeared as a reddish-brown precipitate around the appressoria, and infected cells were observed in all lines at 48 and 72 hpi (Fig. 2a), indicating that ROS production may be similar across lines. When UVB-pretreated leaves were inoculated with M. oryzae, a faint reddish-brown precipitate around the appressoria and infected cells were detected in the S-C and Sa-WT cells within 48 and 72 hpi (Fig. 2b). In contrast, the A-S cells exhibited an intense reddish-brown precipitate in the infected cells (Fig. 2b). This suggested that ROS accumulation may occur earlier or be more active in the A-S line

UVB-induced defense responses reduce the multicellular invasion of secondary infectious hyphae in A-S plants with low CPD photolyase activity
The strong ROS production in UVB-pretreated A-S plants may affect the early stages of the development of M. oryzae infection in rice cells. To investigate this, we used a leaf sheath assay to examine infectious hyphae within the host cells of S-C, Sa-WT, and A-S. Without UVB pretreatment, the infectious hyphae spread widely in all lines and occupied cells adjacent to the primary infected cells by 24 and 48 hpi (Fig. 3a). After UVB pretreatment followed by M. oryzae infection, the infectious hyphae spread widely in the neighboring cells in S-C and Sa-WT, but not in A-S (Fig. 3a). When plants were pretreated with or without UVB and inoculated with sterile water (without M. oryzae conidia), there was no fungal biomass detected by aniline blue staining (Fig.  S3). In UVB-pretreated leaves with M. oryzae infection, aniline blue staining revealed a rapid growth of fungal biomass in all lines during the first 72 hpi (Fig. 3b). UVB pretreatment for 12 h reduced the amount of secondary infectious hyphae in the A-S line, and the pathogens were well-contained within the initial infection site compared with those in the S-C and Sa-WT lines (Fig. 3b). To confirm these results, we counted 100 appressoria in each line. Depending on the growth of the infectious hyphae, we categorized the infection sites from each appressorium as single-cell infection, multi-cell infection, uninfected, and unspecified. Without UVB pretreatment, the number of multi-cell infection site was similar across all lines (60, 64, and 65 in S-C, Sa-WT, and A-S, respectively; Fig. 3c). However, when UVB-pretreated plants were inoculated with M. oryzae, the number of multicellular infection was considerably lower in the A-S line than in the S-C or Sa-WT lines (58, 53, and 25 in S-C, Sa-WT, and A-S, respectively; Fig. 3c). The number of multicellular infection was similar between S-C samples with or without UVB treatment, representing the lowest level of disease resistance exhibited by this line (Fig. 3c). These results indicated that the strong plant defense response of the UVB-pretreated A-S line (Fig. 2b) reduced the spread of infectious hyphae of M. oryzae (Fig. 3).

UVB radiation strongly activates PR genes in CPD photorepair-deficient plants without pathogen inoculation
The activation of PR proteins is often believed to be the molecular basis for the activation of defense response [37,38]. In addition, H 2 O 2 has been suggested to have a signaling role in defense response [39]. Based on the high amount H 2 O 2 in UVB-pretreated A-S plants, we speculated that UVB pretreatment activates defense response in A-S plants, resulting in reduced numbers of multicellular appressoria (Figs. 2b and 3). We also hypothesized that compared to UVB-resistant plants, highly UVB-sensitive plants with low CPD photolyase activity may experience a strong activation of defense-related genes. To verify this possibility, we measured changes in the transcription levels of defense-related genes in UVB-pretreated S-C, A-S, and  TOG-OxPHR plants (Fig. 4) and various UVB-sensitive rice cultivars (Fig. S4) without M. oryzae inoculation. Without UVB pretreatment, the expression levels of all the aforementioned defense-related genes were extremely low in the transgenic plants and rice cultivars (Fig. 4, S4). There were no significant differences in the expression levels of these genes among the transgenic plants and rice cultivars, with the exception of RPM1 in S-C (Fig. 4b), OgWRKY45 in A-S and TOG-OxPHR (Fig. 4f), and OsPAL4 and CAD6 in Surjamkhi (Fig. S4). However, after 12 h of UVB pretreatment, all the defense-related genes examined in this study were more strongly expressed in the A-S line than in the S-C and Sa-WT lines. The expression levels of defense-related genes were lower in the transgenic S-C and TOG-OxPHR plants than in the respective WT plants and were unchanged or slightly higher compared with those in plants without UVB pretreatment. Similar trends were observed in the expression patterns of OgJiPR10, OsPAL4, OsCCR17, and OgWRKY45 among various UVB-sensitive rice cultivars with different CPD photolyase activities, including UVB-sensitive (TOS13649), UVB-hypersensitive (Surjamkhi), and UVBsuper-hypersensitive (TOB7307) plants (Fig. S4). The CAD6 and OsPAL4 expression levels of Surjamkhi without UVB pretreatment were higher than those of other cultivars, and UVB pretreatment did not induce higher levels of CAD6 expression in these cultivars. Importantly, as UVB sensitivity increased, the expression levels of defense-related genes in non-inoculated plants also tended to increase, although UVB pretreatment also activated the defense-related genes of UVB-resistant (S-C, TOG-OxPHR) and UVB-sensitive (TOS13649) plants ( Fig. 4 and S4). These results suggested that UVB pretreatment strongly activates defense response in highly UVB-sensitive plants with low CPD photolyase activity (A-S, TOG-WT, Surjamkhi, and TOB7307) prior to infection occurrence, thus reducing the number of multicellular appressoria in the A-S line (Fig. 3).

High UVB sensitivity increases resistance to paraquat-induced oxidative damage
Elevated UVB radiation can enhance AOS-scavenging enzymes [40]. The high fold change in OgJiPR10 activity in this study (Fig. 4a) suggested the enhancement of oxidative resistance to both biotic and abiotic stresses [21] in the A-S transgenic line, UVB-hypersensitive (Surjamkhi) plants, and UVB-super-hypersensitive (TOB7307) plants (Fig. S4). Thus, we speculated that UVB pretreatment may increase the antioxidant capacity of plants, thus minimizing the oxidative damage caused by biotic and abiotic stresses. To verify this, we pretreated the S-C, A-S, Sa-WT, TOS13649, Surjamkhi, and TOB7307 rice cultivars with or without UVB (0.4 W m −2 ). The excised leaf discs were subjected to a low concentration (10-100 μM) of paraquat (methyl viologen) (Fig. 5) [34,41]. In transgenic lines of Sasanishiki (S-C, A-S and Sa-WT) without UVB pretreatment, there were no differences in sensitivity to oxidative damage caused by paraquat treatment (Fig. 5). After 12 h of UVB pretreatment, the lower oxidative damage was caused by 50 μM paraquat in A-S line compared with that exhibited by the Sa-WT and S-C line (Fig. 5). Similarly, the UVB-sensitive TOS13649 cultivar also showed sensitive to oxidative damage (by 50 μM paraquat) compared with that exhibited by Surjamkhi and TOB7307 cultivars (Fig. 5). These results indicate that after UVB pretreatment, the increased antioxidant capacity of highly UVB-sensitive rice (the A-S line, Surjamkhi, and TOB7307) may minimize the oxidative damage caused by biotic and abiotic stresses.

Discussion
UV radiation has been shown to boost the plant defense system against several pathogenic fungi [13], B. cinerera [15], and H. parasitica [14]. However, little is known about the relationship between CPD photolyase activity and resistance to the rice blast fungus, and very few studies have investigated this association. In this study, we used transgenic O. sativa and O. glaberrima plants and rice cultivars with different CPD photolyase activities to show that 12 h of UVB pretreatment (0.4 W m −2 ) induces higher resistance to the M. oryzae, in highly were assayed by qRT-PCR. Transcript levels of the PR genes were quantified relative to that of the reference gene (actin) using the 2 −ΔΔCT method. UVB pretreatment strongly activates the defense-related genes of highly UVB-sensitive plants. Primer sequences are listed in Table S1.
Values are the mean ± SD. N = 9 with three biological replicates. Different letters and asterisks indicate significant differences, as determined by the Tukey-Kramer test (P < 0.05). Blue, violet, bright green, sky blue, and orange represent S-C, Sa-WT, A-S, TOG-OxPHR, and TOG-WT, respectively UVB-sensitive rice with lower CPD photolyase activity (the A-S line and TOG-WT plants; Fig. 1). We demonstrated that UVB pretreatment strongly induces defense response in highly UVB-sensitive plants and confers disease resistance via the high activation of defense-related genes, PR, lignin, and flavonoid-related genes (Fig. 4,  S4). In contrast, the increased susceptibility of transgenic S-C and TOG-OxPHR plants with higher CPD photolyase activity ( Fig. 1) was due to the lower activation of defense response in these plants after UVB pretreatment. Similar to the A-S line, the Surjamkhi and TOB7307 cultivars also exhibited high activation of defense-related genes (Fig. 4a,  S4). These results indicate that in UVB-pretreated leaves, the degree of defense response induction by M. oryzae inoculation depends on the degree of UVB sensitivity; that is, it depends on the amount of accumulated CPDs. In this study, UVB pretreatment (0.4 W m −2 ) was performed for 12 h (UVB BE ; 4.9 kJ m −2 day −1 ). This is a low dose that simulates the amount of UVB in natural sunlight [18]. The average annual UVB BE at ground level is 8.0 kJ m −2 day −1 in the Sonoran Desert (USA) [42] and 8.5 kJ m −2 day −1 in Cape town (South Africa) [43]. The 4.9 kJ m −2 day −1 UVB intensity used in this study is known to induce CPD in rice leaves [18]. We did not observe any visible symptoms of UVB-induced leaf damage in rice leaves without infection (Fig. 1a). However, this UVB intensity was sufficient to induce the expression of defenserelated genes in rice leaves pretreated with UVB without infection (Fig. 4, S4).
The OgJiPR10 gene was highly activated in the A-S line and in the highly UVB-sensitive cultivars, Surjamkhi, TOB7307, and TOG12380 (Fig. 4a, S4). This may be helping the plants cope with the increased UVB-induced oxidative stress or represent the plants' enhanced capability to resist multiple stresses. This is consistent with the findings of a previous study [21] which showed that a JIOsPR10overexpressing line exhibited increased host resistance to fungal, drought, salt, and oxidative stress. The A-S line may also have developed mechanisms to cope with high ROS production by developing resistance to oxidative damage (Fig. 2b). To elucidate this mechanism, UVB-pretreated rice plants were treated with paraquat. UVB pretreatment increased resistance to paraquat-induced oxidative damage in the A-S line, Surjamkhi, and TOB7307, but not in the S-C line, Sa-WT, or TOS13649 (Fig. 5). This suggests that the higher accumulation of CPDs [9,18] is important for enhancing plant resistance to paraquat-induced oxidative damage. These results are in agreement with those of Fujibe et al. (2004) [40], who reported that UVB treatment promoted the production of ROS-scavenging enzymes. Our results are also consistent with those of Wang et al. (2007) [44], who showed that cross-resistance between oxidative stress and UVB reduced the efficacy of paraquat. Similarly, Hideg et al. (2013) demonstrated that suitable UVB radiation can induce defensive mechanisms and reduce oxidative damage in plants [11]. Thus, high levels of UVB-induced photoproducts-presumably CPDs-enhance the antioxidant capacity of plants with low CPD photolyase activity (Fig. 5). The high expression levels of OsPAL4, CAD6, and OsCCR17 in A-S, Surjamkhi, TOB7307, and TOG-WT High UVB sensitivity increases resistance to paraquat-induced oxidative damage. Comparison of paraquat sensitivity in rice strains with different CPD photolyase activity, including Sasanishiki (Sa-WT), S-C, and A-S and Surjamkhi, TOS13649 and TOB7307. Plants were either pretreated with or without UVB (0.4 W m −2 ) for 12 h, and the excised leaf discs were subjected to a low concentration of methyl viologen (paraquat) for 24 h. a Photographs of some of the leaf discs are shown. Numbers above each photograph indicate the paraquat concentration (μM). Three independent experiments were performed and yielded similar results. Representative photographs are shown from one of the independent experiments. b The percentage of paraquat-induced oxidative damage in a leaf disc of rice strains pretreated with or without UVB. The degree of paraquat-induced damage in a leaf disc shown in Fig. 5 was calculated from the percentage of the area of the white discolored region per the area of each leaf disc using ImageJ software. Values are the mean ± SD. N = 3-6 with three biological replicates. Different letters and asterisks indicate significant differences, as determined by the Tukey-Kramer test (P < 0.05) (Fig. 4, S4) suggested an increase in the production of flavonoids (that confer broad-spectrum disease resistance) [23] and phenolic compounds, as well as an increase in lignification [22,24]. Moreover, UVB radiation can also produce secondary metabolites such as cinnamate esters, flavonoids, tannin, and lignin [45]. Thus, the number of multicellular appressoria may have been lower in the A-S line (Fig. 3c) because the high levels of flavonoids and lignin created an unfavorable environment for M. oryzae. In infected plants, lignin deposition in the cell walls can arrest pathogen growth [46]. Similarly, high UV-induced phenolic compounds can help create an unfavorable environment for powdery mildew, as powdery mildew growing on the leaves needs to penetrate through the epidermal cells [47].
WRKY45 is essential for the benzothiadiazole-induced and nucleotide-binding leucine-rich repeat (NLR) proteinmediated immunity of plants against M. oryzae, and it promotes disease resistance in rice [25,26]. Thus, the high expression levels of this gene in A-S, Surjamkhi, TOB7307, and TOG-WT (Fig. 4f, S4) suggested an enhanced resistance to rice blast disease prior to infection occurrence. Plants exhibit innate immunity to pathogenic microbes via at least two pathways: pattern-triggered immunity (PTI) and effector-triggered immunity (ETI) [48,49]. PTI is a plant's first active response to microbial perception. In rice, PTI is activated by cell wall derivatives of the fungus and can induce plant defense-related gene expression [50]. H 2 O 2 acts as a second messenger for the induction of defenserelated genes. It is crucial for the PTI response in plants [50]. Thus, the reduced number of multicellular appressoria in the A-S line (Fig. 3a, b) may be due to higher ROS-related defense responses in the A-S line than in the S-C and Sa-WT (Fig. 2b). H 2 O 2 was not detected in plants that had not been inoculated with M. oryzae (regardless of UVB treatment; Fig. 2), suggesting that the high H 2 O 2 levels in A-S were mainly due to enhanced disease resistance, and not a direct result of UVB treatment. OgJiPR10 and OgWRKY45, the marker genes of jasmonic acid (JA) and salicylic acid (SA), respectively, show increased fold change in JA-and SAinduced plant defense [51,52]. This is similar to the trends observed in UVB-pretreated A-S, Surjamkhi, TOB7307, and TOG-WT plants ( Fig. 4 and S4). These results show that the strong accumulation of H 2 O 2 may account for the increased PTI response, as revealed by the high expression levels of PR genes in these plants. Accordingly, the high susceptibility of UVB-pretreated S-C and TOG-OxPHR plants to M. oryzae at 9 dpi (Fig. 1) may be due to low defense response activation ( Fig. 4 and S4), ROS accumulation (Fig. 2b), and antioxidant capacity (Fig. 5) in these plants, possibly because of low levels of CPDs [9,18].
Our results showed that lower CPD photolyase activity-that is, higher CPD accumulation-induced disease resistance. UVB radiation has been reported to induce various defense-related genes (including PR genes). Moreover, ROS accumulation is important for the induction of these responses by environmental stresses. However, UVBinduced DNA damage, CPDs, and (6-4) photoproducts have been shown to activate the transcription of the β-1,3glucanase gene [12] and promote resistance to pathogens [14]. In this study, we demonstrated that the accumulation of CPDs activated the induction of H 2 O 2 production and promoted defense-related genes. However, it is unclear whether CPD directly activated these responses, or whether they were induced by the intracellularly accumulated ROS as a result of CPD accumulation.
The primary aim of our study was to examine how highly UV-sensitive rice varieties continue to be cultivated in areas with relatively high UV levels. We suggest that high sensitivity to UVB may offer advantages to plants being cultivated in these environments and that there exists a trade-off between UVB sensitivity and disease sensitivity. We demonstrated that highly UVB-sensitive rice acquired disease resistance. Therefore, in areas with high UVB radiation, UVB-sensitive rice may have evolved to be highly resistant to broad-spectrum plant diseases. However, rice plants grown in these areas are exposed to various environmental stresses, including high temperature, drought or salt stress, as well as disease stress. In fact, exposure to UVB radiation improves drought tolerance in silver birch seedlings [13]. In addition, our results suggest that UVB-sensitive rice exhibits resistance to paraquat-induced stress, thus showing a multistress response.
Our results indicate that UVB-sensitive rice with low CPD photolyase activity would be useful for cultivation in areas with higher UVB radiation, such as Southeast Asia and Africa. Investigating the trade-off between low CPD photolyase activity and stress response to multiple stresses (such as high temperature and drought) may help researchers develop more varieties of multi-stress-resistant rice plants.

Conclusion
There is a lack of specific evidence regarding UVB sensitivity and increased resistance to fungal pathogens [53]. Our results improve our fundamental understanding of cross-tolerance between UVB radiation and resistance to M. oryzae. Moreover, our findings may help design multi-stress-resistant plants that can be cultivated without the excessive use of synthetic pesticides, which have negative impacts on human health and ecosystems. Crops with combined resistance to UVB and rice blast disease would have improved yield. This would help meet the food demand of a growing global population, especially in tropical areas (such as Africa) where plants experience multiple stress simultaneously.