UV-B-sensitive Cheongja 3 and UV-B-resistant Buseok (IT162669), which were identified as the most sensitive and resistant soybean genotypes to supplementary UV-B irradiation, respectively (Kim et al., 2015), were used in this study. Under 6 h of UV-B treatment, Buseok had only one or two slightly yellowish leaves and still showed vigorous growth, whereas in Cheongja 3, most leaves turned yellow with red spots, ultimately leading to defoliation (Figure S1). Seeds of both soybean genotypes were planted in plastic pots (18 cm diameter, 20 cm deep) in a 1:1:1 mixture of field soil, desalinated sand, and commercial peat soil in a greenhouse at Seoul National University Experimental Farm. One plant per pot was grown under a natural photoperiod of 11.5–14.5 h per day using standard cultivation methods.

For UV-B treatment, soybean plants were divided into three groups; no UV-B treatment (0 h) as a control, 0.5- and 6-h UV-B treatments). Supplementary UV-B irradiation was conducted at V4 growth stage as previously described (Kim et al., 2015). UV-B irradiation began at 9:00 am, and the plants were exposed to prolonged UV-B stress for 0.5 and 6 h. The intensity of 0.5 h UV-B irradiation was equivalent to 11.5 kJ/m² daily soybean UV-B biological effective dose (UV-B_BE), and the 6 h treatment was equivalent to a dose of irradiation that was 12-times higher than daily UV-B_BE (Caldwell, 1971). For Illumina RNA sequencing, the uppermost trifoliate leaves were collected from UV-B-treated and non-treated plants. At each sampling time, the leaf tissues from four pots per group of each genotype were pooled together as one biological replicate, frozen immediately in liquid nitrogen, and stored at −80 until use. Thus, a total of six leaf samples (three treatments [0, 0.5, and 6 h] × two genotypes [Cheongja 3 and Buseok] × one biological replicate) were collected for RNA-seq analysis.

Total RNA was extracted from the leaf samples using TRIzol reagent (Invitrogen, Carlsbad, CA, USA) according to the manufacturer's instructions. The cDNA library for Illumina HiSeq 2000 sequencing was constructed using an mRNA-seq sample preparation kit (TruSeq® RNA Sample Prep Kit v2, Cat.8207217, Illumina Inc., San Diego, CA, USA), including mRNA purification, cDNA synthesis, end-repair of cDNA, adaptor ligation, and cDNA amplification. Six cDNA libraries were pooled at equal molar concentration for loading onto two lanes of an Illumina flow cell and then merged for data analysis. Sequencing runs were performed in paired-end mode for 10l cycles using the Illumina HiSeq 2000 platform.

The 100 bp paired-end sequence reads were mapped to the soybean reference genome (Glyma v1.1) from Phytozome (http://www.phytozome.net/soybean) using Bowtie (http://bowtie-bio.sourceforge.net/index.shtml) and TopHat (http://ccb.jhu.edu/software/tophat/index.shtml) in default modes (Langmead et al., 2009; Trapnell et al., 2009). The Cufflinks program was used to assemble gene transcripts and to normalize transcript abundance in terms of fragments per kilobase pair of transcript per million mapped reads (FPKM) (Trapnell et al., 2012). The Cuffdiff program (http://cufflinks.cbcb.umd.edu/index.html) was used to test statistically significant differences in transcript expression in seven pairs of comparisons, which consisted of three comparisons of Cheongja 3 vs. Buseok under control conditions and two UV-B treatments, and four comparisons of control vs. UV-B treatments in the two genotypes (Figure S2). Significant DEGs were detected using the following criteria: (i) absolute fold-change > 1 and (ii) q-value (false discovery rate [FDR]) <0.05 and p-value < 0.004. R studio (https://www.rstudio.com/, version 3. 3. 0) was used to run custom R scripts to perform hierarchical cluster analysis of DEG expression and to construct heat maps.

UV-B related genes from A. thaliana were surveyed using Gene Ontology (GO) (http://amigo.geneontology.org/amigo/search/ontology?q=UV-B) and The Arabidopsis Information Resource (TAIR) (https://www.Arabidopsis.org/) database. In addition, a list of A. thaliana genes shown to be involved in the UV-B response was compiled from the literature; 194 UV-B-related A. thaliana genes were ultimately obtained (Figure S2, Supplementary Table 1). To detect genes that function in networks with the UV-B related genes, the set of selected UV-B related genes was analyzed with Aranet (http://www.functionalnet.org/aranet/), a probabilistic functional gene network for A. thaliana, resulting in the identification of 4666 genes predicted to respond directly or indirectly to UV-B (Figure S2, Supplementary Table 2). A total of 15,074 soybean homologs of these A. thaliana genes were identified using Blast analysis. Based on this soybean gene list, DEGs putatively involved in the defense response to UV-B stress were identified (Figure S2).

To better understand the biological functions of the DEGs in response to UV-B, enrichment of GO categories among the DEGs was assessed using BINGO software (http://www.psb.ugent.be/cbd/papers/BiNGO/Home.html) (Maere et al., 2005). Significantly over-represented GO categories were visualized in Cytoscape (http://www.cytoscape.org). The biochemical pathways involving the DEGs were predicted using Kyoto Encyclopedia of Genes and Genomes (KEGG) (http://www.genome.jp/kegg/).

Gene-specific primers were designed based on the nucleotide sequences of the chosen DEGs for qRT-PCR analysis using Primer3 software (http://primer3plus.com/web_3.0.0/primer3web_input.htm) (Supplementary Table 3). Total RNA from six samples (Cheongja 3 and Buseok; control, 0.5, and 6 h UV-B treatments) was used to synthesize cDNA using a Bio-Rad iScript™ cDNA Synthesis Kit (Cat. 170-8891, Hercules, CA, USA). The synthesized cDNAs were used for qRT-PCR with a Bio-Rad iQ™ SYBR Green Supermix Kit (Cat. 170-8882) using a LightCycler® 480 (Roche Diagnostics, Laval, QC, Canada). The qRT-PCR reaction mixtures (total volume of 20 μl) contained 100 ng of cDNA, each primer at 300 μM, 8 μl of sterile water, and 10 μl of Bio-Rad iQ™ SYBR Green Supermix. The amplification conditions were as follow: 5 min denaturation at 95°C followed by 40 cycles of 95°C for 10 s and 60°C for 1 min. The samples were analyzed in triplicate to ensure statistical significance, and the tubulin gene was used as a reference gene for normalization of target gene expression in soybean. Data were analyzed based on the stable expression level of the reference gene according to the method of Livak and Schmittgen (2001). To assess treatment effects on soybean genotypes, significances were analyzed with Fisher's least significant difference tests (P < 0.05) using the Statistical Analysis System (SAS 9.4 for window, SAS Institute Inc., Cary, NC, USA).

Seeds of knock-out mutants for the target genes were obtained from the Arabidopsis Biological Resource Center (ABRC, http://abrc.osu.edu/) at Ohio State University, USA. Seven mutant lines were utilized, including mutants for five members of TIR-NBS-LRR family (AT1G64070.1; SALK_042846C, AT5G17680.1; SALK_004241C, AT5G36930.1; SALK_124056C, AT5G41540.1; SALK_034471C, and AT4G36150.1; SALK_084909C), one DGK mutant (AT5G07920.1; SALK_033664C), and one PIP5K mutant (AT1G34260.1; SALK_047604C). To identify homozygous mutants for the target genes, seeds incubated at 4°C were sown in commercial peat soil and grown under 16/8 h light conditions in a growth chamber for 14 days. Homozygous mutant plants were confirmed by RT-PCR using primers designed based on the inserted T-DNA and target gene DNA sequences, resulting in the identification of only four knock-out mutants (Figure S3; Supplementary Table 4). Both knock-out mutant and Columbia ecotype (Col-0) seeds were sterilized and incubated at 4°C in the dark for 4 days to synchronize germination and to ensure uniform growth, followed by cultivation under 16 h white fluorescent light for 7 days in a growth chamber. As a pilot UV-B irradiation test to determine suitable irradiation times for A. thaliana, wild-type (Col-0) plants were subjected to different UV-B treatments (1, 2, 4, and 8 h); 4 h UV-B treatment was determined to be suitable (Figure S4). Seven-days-old knock-out mutants were treated with 4 h UV-B irradiation at 22°C. For qRT-PCR analysis, rosette leaves were collected from UV-B-treated and non-treated mutant plants, frozen immediately in liquid nitrogen, and stored at −80°C until use. The specific primers for qRT-PCR were designed using primer 3 (Supplementary Table 5).

To compare transcriptomic variation in the soybean lines in response to UV-B stress, Cheongja3 and Buseok leaves were collected after continuous UV-B exposure at a high fluence rate for 0 (control), 0.5, and 6 h, with a total of six samples. Using Illumina HiSeq 2000 system, 0.3 billion 100 bp paired-end reads were produced, ranging from 47 to 62 million reads per sample (Supplementary Table 6), which were mapped against the soybean reference sequence. The RNA-seq analysis workflow depicted in Figure S2 was implemented to analyze the sequencing data. Of the total reads, 78% were properly mapped to the reference sequence, resulting in approximately 25-fold average coverage (Supplementary Table 6). About 20% unmapped reads appears to be attributable to either misassembled or absent sequences in the reference assembly.

To identify DEGs related to UV-B stress, we compared transcript abundances among the six samples from two soybean genotypes differing in UV-B tolerance based on FPKM values. Two-way comparison data analysis was used to investigate transcriptomic variations, representing the comparison (i) between Cheongja 3 and Buseok by UV-B treatment and (ii) between control and UV-B treatments by genotype (Figure S2). To further identify DEGs in response to supplementary UV-B irradiation, we then utilized a set of 4666 A. thaliana genes predicted to be implicated in UV-B stress defense, either directly or indirectly, via Aranet (Figure S2; Supplementary Tables 1, 2). Homology comparative analysis revealed that 1875 DEGs detected in the Cheongja 3 vs. Buseok comparison were orthologs of UV-B-related Arabidopsis genes (Figure S2). In the control vs. treatment comparison, 228 DEGs in Buseok and 129 in Cheongja 3 were found to be homologous to the UV-B related Arabidopsis genes (Figure S2; Supplementary Table 7–9).

Using the final sets of DEGs, we generated Venn diagrams to identify treatment- or genotype-specific DEGs and common DEGs (Figure 1A). A larger number of DEGs were specific to each genotype and each UV-B treatment time. In the Cheongja 3 vs. Buseok comparison, 499 DEGs specific to the controls were considered to be due to differences in the genetic backgrounds between the genotypes and were thus excluded from further functional classification. Of the 1376 DEGs in Buseok relative to Cheongja 3 (Supplementary Table 9), 511 were upregulated and 345 were downregulated under 0.5 h UV-B treatment (Figure 1B). Under 6 h UV-B treatment, the number of up- and downregulated DEGs was similar. The control vs. treatment comparisons by genotype showed that Buseok had more DEGs than Cheongja 3 both for 0 vs. 5 h and 0 vs. 6 h (Figure 1B). In Cheongja 3, the number of DEG is higher in upregulation than downregulation for 0 vs. 6 h. There were 149 and 50 Buseok- and Cheongja 3-specific DEGs, respectively (Figure 1A). Most DEGs in both genotypes were upregulated by 0.5 h UV-B treatment but more strongly downregulated by 6 h UV-B exposure (Figure 1B).

We investigated the biological significance of the changes in transcript abundance provoked by UV-B irradiation stress in soybean. To address the functional distribution of the identified DEGs, we performed GO term enrichment using Cytoscape plug-in BINGO. GO assignment of 690 DEGs among the 1376 DEGs identified in the Cheongja 3 vs. Buseok comparison by UV-B treatment resulted in significant overrepresentation of 31 GO terms, including 17 in the biological process category and 14 in the metabolic process category (Figure 2; Supplementary Table 10). In the biological process category, over-represented GO terms were mainly classified into four clusters, as shown in Figure 2, including cell morphogenesis, immune response, signaling, and apoptosis. A large portion of the DEGs involved in the GO clusters immune response, signaling, and apoptosis were overlapping, i.e., toll/interleukin-1 nucleotide-binding-site leucine-rich repeat (TIR-NBS-LRR) genes. The GO terms over-represented in metabolic process were primarily divided into five clusters including light harvesting system, fatty acid biosynthetic process, small molecule biosynthetic process, ROS metabolic process, and oxidation reduction. From the control vs. treatment comparisons in UV-B-resistant Buseok, 176 DEGs induced by 0.5 h UV-B treatment were grouped into six GO clusters, including biological regulation, immune response, signaling, apoptosis, ion transport, and polysaccharide metabolic process (Figure S5; Supplementary Table 11). However, no GO term was over-represented by the 124 DEGs in Buseok in response to 6 h UV-B treatment. In the control vs. treatment comparisons in UV-B-sensitive Cheongja 3, 9 DEGs under 0.5 h UV-B treatment were grouped into the GO term metal ion transport (GO:0030001) and 26 DEGs under 6 h UV-B treatment were grouped into oxidation reduction (GO:0055114) (Figure S6; Supplementary Table 12). According to BINGO analysis, the GO terms involved in immune response, cell death, and signaling were the most dominant.

KEGG pathway mapping of the 1376 DEGs identified in the Cheongja 3 vs. Buseok comparison under UV-B treatments revealed the involvement of 101 biological pathways (Supplementary Table 13). The 228 and 129 DEGs in the control vs. treatment comparisons in Buseok and Cheongja 3 were assigned to 28 and 52 KEGG pathways, respectively (Supplementary Tables 14, 15). Interestingly, phosphatidylinositol signaling system (gmx04070) and mTOR signaling pathway (map04150), which are involved in environmental information processing and signal transduction, were also identified.

Based on the results of BINGO and KEGG analysis, we specifically focused on DEGs in over-represented GO and KEGG terms such as cell death and immune system, stress defense signaling, and ROS metabolism, as described below. We also investigated the expression patterns of genes in the photosystem and circadian rhythm categories.

Supplemental UV-B radiation downregulates genes encoding several key photosynthetic proteins, including small subunits of Rubisco, subunits of ATP synthase, and chlorophyll a/b binding protein of the light harvesting antenna complex of photosystem II (Jenkins, 2009). In the present study, there are no photosynthetic DEGs by UV-B treatments in both Buseok and Cheongja 3 compared to the control (Supplementary Tables 7–9, 16). However, differences between the two soybean genotypes were observed in the transcript abundance of chlorophyll a/b binding protein and photosystem light harvesting complex genes under both control and elevated UV-B conditions (Supplementary Table 16). Inherent variations in the expression of two genes encoding light harvesting complex subunits were detected between Buseok and Cheongja 3 under ambient light (control). In addition, the expression patterns of six genes associated with photosystem during UV-B irradiation significantly differed between genotypes, which included two homologs (Glyma09g0826 and Glyma15g19810) for photosystem I light harvesting complex gene 6, one (Glyma04g33360) for light-harvesting chlorophyll-protein complex I subunit A4, one (Glyma16g27995) for light-harvesting chlorophyll-protein complex II subunit B1, and one (Glyma11g35130) for light harvesting complex photosystem II (Supplementary Table 16). These results indicate that the transcriptional regulation of photosynthetic genes in response to supplementary UV-B irradiation is different between soybean genotypes.

Photomorphogenic UV-B responses are specifically mediated by (but are not independent of) other aspects of light signaling, including photoreceptors and circadian rhythms. We observed significant differences in the transcript abundance of genes involved in UV-B specific photomorphogenic pathways between UV-B-resistant Buseok and UV-B-sensitive Cheongja 3. Specifically, HY5 (Glyma08g41450) and COP1 (Glyma14g05430) were upregulated by 6 h UV-B treatment (Supplementary Table 9). Most DEGs encoding CHS proteins, which were expressed at lower levels in Buseok compared to Cheongja 3 under control conditions, were upregulated by UV-B irradiation (Supplementary Table 16). Additional key genes involved in the circadian rhythm pathway, such as PHYTOCHROME A, GIGANTEA PHYTOCHROME INTERACTING FACTOR 3, and PSEUDO-RESPONSE REGULATOR, were also differentially expressed in response to supplementary UV-B exposure (Supplementary Table 16). In the control vs. treatment comparisons by genotype, UVR8, HY5, and COP1 were not detected among DEGs (data not shown).

To investigate which components in defense signaling pathways are activated in response to UV-B stress, we investigated significant differences in the expression of genes participating in programmed cell death and the immune system. The comparison of Cheongja 3 vs. Buseok under UV-B treatments revealed differential expression of 20 TIR-NBS-LRR genes, 11 and 9 of which were upregulated in Buseok and Cheongja 3, respectively (Table 1; Figure 3). Seven additional TIR-NBS-LRR and two coiled-coil-nucleotide-binding-site leucine-rich repeat (CC-NBS-LRR) genes were induced by 0.5 h UV-B irradiation only in Buseok (Table 1), while none of these genes were induced by 6 h treatment.

Nine members of the heat shock 70 (HSP70) protein family were differentially expressed in response to UV-B irradiation in Buseok relative to Cheongja 3, six of which were down-regulated (Table 1; Figure 3). However, there was no significant difference in the expression patterns of HSP70 genes between control and 0.5 or 6 h UV-B treatment in Buseok or Cheongja 3.

Genome-wide transcript analysis in response to supplemental UV-B irradiation revealed DEGs involved in the phosphatidylinositol signaling pathway (Table 2). The phosphatidylinositol signaling pathway generates phosphatidic acid (PA) from structural phospholipids as a secondary messenger. Under UV-B stress, transcripts for phosphatidylinositol-4-phosphate 5-kinase (PIP5K) family protein were more abundant in Buseok than in Cheongja 3 (Table 2); this protein generates phosphatidylinositol 4,5-bisphosphate [PtdIns(4,5)P₂] as a substrate of phospholipase C (PLC) from phosphatidylinositol 4-phosphate (Ptdins4P) in the plasma membrane (Mueller-Roeber and Pical, 2002). Also, PtdIns(4,5)P₂ is produced from the phosphorylation of phosphatidylinositol-3,4,5-trisphosphate [PtdIns(3,4,5)P₃] by phosphatidylinositol-3,4,5-trisphosphate 3-phosphatase and dual-specificity protein phosphatase (PTEN 2), for which one gene (Glyma20g26280) was upregulated in Buseok under UV-B irradiation (Table 2). PtdIns(4,5)P₂ is hydrolyzed into inositol 1,4,5-trisphosphate (InsP₃) and diacylglycerol (DAG) by phosphatidylinositol-specific PLC (PI-PLC) proteins. One DEG (Glyma18g03090) encoding PI-PLC was upregulated in Cheongja 3 under 6 h UV-B treatment relative to Buseok (Table 2). In UV-B-resistant Buseok, PI-PLC genes were expressed at low levels and independent of UV-B treatment. Meanwhile, two DEGs (Glyma17g08510 and Glyma06g39760) for diacylglycerol kinase (DGK) family proteins were upregulated in Buseok by UV-B stress (Table 2); these proteins convert DAG produced by PI-PLC to PA (Laxalt and Munnik, 2002). The coupling of DGK to the activation of PI-PLC may be induced in response to UV-B stress in Buseok, leading to increased levels of PA in the cells, thereby activating several downstream defense responses.

InsP3 generated via the hydrolysis of PI-PLC diffuses into the cytosol and is involved in the release of Ca²⁺ from intracellular stores (Ruelland et al., 2015). Owing to a lack of InsP3 receptors in plants, InsP3 is converted into the more phosphorylated forms of inositol, i.e., tetra, penta, and hexaphosphates (InsP4, InsP5, and InsP6), through further phosphorylation steps involving at least two types of inositol polyphosphate 2-kinase (IPK1) and inositol polyphosphate kinase 2 (IPK2, synonym for inositol 1,3,4-trisphosphate 5/6-kinase) (Munnik and Vermeer, 2010; Zhou et al., 2012; Sparvoli and Cominelli, 2015). Among these enzymes, Buseok exhibited enhanced expression of two IPK2 family genes under 0.5 and 6 h UV-B irradiation compared to Cheongja 3 (Table 2). Therefore, in UV-B-resistant Buseok, the genes for four enzymes involved in the PA-dependent signaling pathway were upregulated in response to UV-B stress (Figure 3).

We also identified some DEGs encoding putative target proteins that interact with PA, including two DEGs (Glyma08g20830, Glyma19g24375) encoding protein phosphatase 2 family proteins, which were upregulated in Buseok under 0.5 h UV-B treatment (Figure 3; Supplementary Table 16); these proteins carry a PA binding motif. One example of a functionally characterized protein phosphatase targeted by PA is the protein phosphatase 2C ABI1 (ABA insensitive 1), which is bound by PA to negatively regulate ABA signaling (Zhang et al., 2004). We detected differential transcript accumulation of three genes encoding MAPKs in response to UV-B stress between Cheongja 3 and Buseok (Figure 3; Supplementary Table 16); MAPK 19 (Glyma15g38490) and MAPK 20 (Glyma14g03190) were upregulated in Buseok under 6 h UV-B treatment and MAPK 3 (Glyma11g15700) was upregulated in Cheongja 3 under 0.5 h UV-B treatment (Figure 3; Supplementary Table 16). In both Buseok and Cheongja 3, however, no significant difference in MAPK expression was detected in the control vs. UV-B treatment comparison.

Abiotic stresses including UV-B stress induce ROS production and scavenging. Four genes encoding copper/zinc superoxide dismutase (Cu/Zn SOD) family proteins, which catalyze the dismutation of superoxide anion (O2•-) to oxygen (O₂) and hydrogen peroxide (H₂O₂), were upregulated in Buseok in response to UV-B stress, while in Cheongja 3, the iron (Fe)-SOD gene was upregulated in response to 6 h UV-B treatment. Among the genes assigned to the GO term “oxidation reduction,” 21 cytochrome P450 (Cyt P₄₅₀) family genes were identified, 12 and 9 of which were upregulated in Buseok and Cheongja 3, respectively (Figure 3; Supplementary Table 16). Six of eight Cyt P₄₅₀ DEGs were downregulated in Cheongja 3 under 6 h UV-B treatment compared to the control (Supplementary Table 17). NAD(P)H-dependent electron transport involving cytochrome P450 produces O2•- in the endoplasmic reticulum (Sharma et al., 2012). Genes of the Cyt P₄₅₀ family 76 subfamily C2 and the Cyt P₄₅₀ family 706 subfamily A are induced by UV-C irradiation (Narusaka et al., 2004). To avoid injury from ROS overproduction, ROS scavenging or detoxification is performed by antioxidative systems consisting of both nonenzymatic and enzymatic antioxidants (Sharma et al., 2012). In Buseok, we identified upregulated DEGs involved in flavonoid biosynthesis to produce phenolic compounds with antioxidant properties, such as anthocyanidin and tannin, including genes encoding flavanone-3-hydroxylase (F3H) and cinnamate-4-hydroxylase (C4H) (Supplementary Table 18). By contrast, a gene encoding the enzymatic oxidant ascorbate peroxidase was upregulated in Cheongja 3. Under 6 h UV-B treatment, genes encoding glutathione peroxidase were also upregulated only in Cheongja 3 compared to the control (Supplementary Table 17).

Using qRT-PCR, we confirmed that five genes (Glyma03g07121, Glyma06g41880, Glyma12g16450, Glyma16g34086, and Glyma19g07650) in the TIR-NBS-LRR family were upregulated in Buseok in response to 0.5 h UV-B irradiation (Figures 4A–E). The TIR-NBS-LRRs' expression patterns obtained by qRT-PCR also showed significantly increased transcript abundance in Buseok by UV-B treatment. Interestingly, transcript accumulation of Glyma03g07121 (TIR-NBS-LRR) was not observed in Cheongja 3 by qRT-PCR or RNA-seq. In addition, the expression profiles of individual DEGs encoding IPK2 (Glyma06g09430), PIP5K (Glyma10g36251), DGK 1 (Glyma17g08510) and DGK 5 (Glyma06g39760) in the PA signaling pathway were obtained by qRT-PCR (Figures 4F–I). Significant upregulation of IPK2 (Glyma06g09430), PIP5K (Glyma10g36251) and DGK 5 (Glyma06g39760) was observed in Buseok compared to Cheongja 3 under UV-B stress. DGK 1 (Glyma17g08510) was significantly activated only in Buseok by 6 h UV-B irradiation relative to control. Finally, we investigated the expression patterns of three DEGs encoding Cyt P₄₅₀ (Glyma06g21920), Cu/Zn SOD (Glyma11g19840), and C4H (Glyma14g38580), which are involved in ROS production and scavenging, using qRT-PCR (Figures 4J–L). Two genes Cu/Zn SOD (Glyma11g19840) and C4H (Glyma14g38580), except Cyt P₄₅₀ (Glyma06g21920) showed significant higher expression in Buseok by both RNA-seq and qRT-PCR analyses, There was no significantly differential expression of Cyt P450 (Glyma06g21920) between Cheongja 3 and Buseok but 6 h UV-B treatment induced upregulation of Cyt P450 in both genotypes. All but one of the gene expression patterns measured by qRT-PCR agreed with ones analyzed by RNA-seq.

We identified four A. thaliana knock-out mutant lines, including two for TIR-NBS-LRR (AT5G36930:SALK_124056C and AT4G36150: SALK_084909C), one for DGK (AT5G07920:SALK_033664C), and one for PIP5K (AT1G34260:SALK_047604C) (Figure 5). The knock-out mutant lines did not show any noticeable morphological changes compared to wild-type (Col-0) under normal growing condition, suggesting that these genes do not play vital roles in basic growth and development. However, 2 days after 4 h UV-B stress, three of the mutant lines (AT5G36930:SALK_124056C, AT5G07920:SALK_033664C, and AT1G34260:SALK_047604C) showed severe chlorosis and stagnant growth, indicating increased sensitivity to UV-B stress (Figure 5). By contrast, the wild type and the remaining mutant lines (affected in the TIR-NBS-LRR gene) exhibited continued growth despite the presence of leaf curling and slight wilting. We measured the expression patterns of the four target genes in the wild type and mutant lines in response to 4 h UV-B irradiation by qRT-PCR. In the wild type, all of the genes except AT5G36930 (TIR-NBS-LRR) were induced by 4 h UV-B irradiation. As expected, the expression levels of all target genes were highly reduced in the corresponding mutant lines under both control and UV-B treatment. Functional validation using the A. thaliana knock-out mutants revealed that some genes implicated in the phosphatidic acid signaling pathway and immune response play key roles in UV-B stress defense.

The authors declare that the research was conducted in the absence of any commercial or financial relationships that could be construed as a potential conflict of interest.