OsWRKY5 Promotes Rice Leaf Senescence via Senescence-Associated NAC and Abscisic Acid Biosynthesis Pathway

The onset of leaf senescence is triggered by external cues and internal factors such as phytohormones and signaling pathways involving transcription factors (TFs). Abscisic acid (ABA) strongly induces senescence and endogenous ABA levels are finely tuned by many senescence-associated TFs. Here, we report on the regulatory function of the senescence-induced TF OsWRKY5 TF in rice (Oryza sativa). OsWRKY5 expression was rapidly upregulated in senescing leaves, especially in yellowing sectors initiated by aging or dark treatment. A T-DNA insertion activation-tagged OsWRKY5-overexpressing mutant (termed oswrky5-D) promoted leaf senescence under natural and dark-induced senescence (DIS) conditions. By contrast, a T-DNA insertion oswrky5-knockdown mutant (termed oswrky5) retained leaf greenness during DIS. Reverse-transcription quantitative PCR (RT-qPCR) showed that OsWRKY5 upregulates the expression of genes controlling chlorophyll degradation and leaf senescence. Furthermore, RT-qPCR and yeast one-hybrid analysis demonstrated that OsWRKY5 indirectly upregulates the expression of senescence-associated NAM/ATAF1/2/CUC2 (NAC) genes including OsNAP and OsNAC2. Precocious leaf yellowing in the oswrky5-D mutant might be caused by elevated endogenous ABA concentrations resulting from upregulated expression of ABA biosynthesis genes OsNCED3, OsNCED4, and OsNCED5, indicating that OsWRKY is a positive regulator of ABA biosynthesis during leaf senescence. Furthermore, OsWRKY5 expression was suppressed by ABA treatment. Taken together, OsWRKY5 is a positive regulator of leaf senescence that upregulates senescence-induced NAC, ABA biosynthesis, and chlorophyll degradation genes.


Introduction
Leaf senescence is the final stage of plant development and involves diverse molecular and cellular processes such as degradation of chlorophylls and macromolecules, and remobilization of nutrients into newly developing or storage organs through expression of senescence-associated genes (SAGs). The onset of leaf senescence begins with chlorophyll degradation and proceeds to hydrolysis of macromolecules (proteins, lipids, and nucleic acids); this is followed by cell death [1][2][3].
Genetic studies have revealed the contribution of chlorophyll catabolic enzymes to sequential reactions of chlorophyll degradation. The STAY-GREEN (SGR) protein is a magnesium (Mg)-dechelatase, which produces pheophytin a by removing Mg from chlorophyll a [4,5]. Thus, functional deficiency of SGR orthologs leads to a strong stay-green phenotype in diverse plant species including Arabidopsis thaliana [6], rice (Oryza sativa) [4], pea (Pisum sativum) [7], tomato (Solanum lycopersicum), bell pepper (Capsicum annuum) [8], and soybean (Glycine max) [9]. Failure to convert pheophytin a to pheophorbide a due to mutation of NON-YELLOW COLORING3 (NYC3), encoding an α/β hydrolase-fold family protein, delays leaf senescence during dark-induced senescence (DIS) [10]. Knockdown of rice pheophorbide a oxygenase (OsPAO) leads to accumulation of pheide a and prolongs leaf greenness during dark incubation [11]. SAGs identified during leaf senescence in rice encode putative proteins involved in metabolic programing [12]; Osh36 and Osl85 encode an aminotransferase and isocitrate lyase, which participate in amino acid and fatty acid metabolism, respectively.
The WRKY TFs participate in various biological processes such as biotic and abiotic stress, seed development, seed dormancy, and germination [26]. Genome-wide analyses have revealed that many WRKY genes are strongly induced by leaf senescence [27,28], suggesting that WRKY TFs are involved in regulating leaf senescence. Following identification of AtWRKY6 as a regulator of leaf senescence [29], other WRKY TFs regulating leaf senescence have been functionally characterized. For example, mutation of Arabidopsis AtWRKY53 confers a delayed leaf senescence phenotype by specifically altering regulation of its target genes [30]. Overexpression of AtWRKY22, a target gene of AtWRKY53, accelerates leaf senescence [31]. AtWRKY54 and AtWRKY70 act as negative regulators of leaf senescence by interacting independently with WRKY30 [32]. AtWRKY45 mediates gibberellic acid (GA)-induced leaf senescence by interacting with a DELLA protein, RGL1 [33]. AtWRKY75 increases salicylic acid (SA) and H 2 O 2 levels by activating SID2 and repressing CAT2, respectively, resulting in early leaf senescence [34]. Heterologous expression of rice OsWRKY23 promotes leaf senescence in Arabidopsis [35]. Rice OsWRKY42 induces reactive oxygen species (ROS) by directly downregulating the expression of OsMT1d encoding metallothionein protein and thereby promoting leaf senescence [36]. Unlike Arabidopsis WRKY TFs involved in the regulation of leaf senescence; however, few OsWRKY TFs have been identified as functioning in the execution of leaf senescence.
In this study, we found that OsWRKY5 expression is upregulated at the onset of leaf senescence. The OsWRKY5-overexpressing oswrky5-D mutation promoted leaf yellowing under aging and dark treatment, while an oswrky5-knockdown mutant exhibited a delayed senescence phenotype during DIS. Reverse-transcription quantitative PCR (RT-qPCR) analysis suggested that CDGs and SAGs were upregulated by senescence-induced OsWRKY5. Furthermore, OsWRKY5 seemed to indirectly regulate the expression of senescence-associated NAC (senNAC) genes such as OsNAP and OsNAC2, which are upstream regulators of CDGs and SAGs. OsWRKY5 elevated endogenous ABA levels by upregulating the expression of ABA biosynthetic genes. Our results thus provide evidence that OsWRKY5 acts as a positive regulator of leaf senescence in rice.

Characterization of OsWRKY5
OsWRKY5 (Os05g04640), a member of rice WRKY TF family, consists of six exons with 1509 bp of open reading frame in 5379 bp of genomic DNA. OsWRKY5 is predicted to encode a 502 amino acid protein with a molecular mass of 52.3 kDa (http://rice.plantbiology.msu.edu/index.shtml). The WRKY domain of OsWRKY5 has a single consensus motif (WRKYGQK) and a zinc-finger C 2 H 2 motif (Cx 5 Cx 23 HxH), indicating that OsWRKY5 belongs to the group II WRKY TF family [37]. From sequence alignment of WRKY domains between OsWRKY5 and group II Arabidopsis thaliana WRKY (AtWRKY) proteins, we found that the domain sequences of OsWRKY5 are quite similar to those of AtWRKY6 and AtWRKY47, members of the subgroup IIb AtWRKY TF family ( Figure S1). To examine the subcellular localization of OsWRKY5, we transiently expressed the 35S::YFP-OsWRKY5 construct in onion epidermal cells. The fluorescent signal of YFP-OsWRKY5 fusions was observed exclusively in nuclei ( Figure S2), indicating that OsWRKY5 is a nuclear-localized protein.

OsWRKY5 Is Upregulated during Leaf Senescence
To examine the spatial expression of OsWRKY5, we investigated transcript levels of OsWRKY5 in rice organs including root, culm, leaf blade, leaf sheath, and panicle at the reproductive stage ( Figure 1a). OsWRKY5 was preferentially expressed in the leaf blade and leaf sheath. Previous transcriptome analysis [28] showed upregulation of 47 rice WRKY TFs including OsWRKY5 in flag leaves during natural senescence (NS). Therefore, we determined age-dependent changes in OsWRKY5 expression in flag leaves of wild-type (WT; japonica cultivar 'Dongjin') rice grown in a paddy field under natural long-day (NLD) conditions (>14 h light/day). While OsWRKY5 was constitutively expressed in developing leaves at the vegetative stage, OsWRKY5 expression was dramatically upregulated in senescing leaves at the reproductive stage ( Figure 1b). In addition, OsWRKY5 expression gradually increased in detached leaves of four-week-old WT leaves during DIS (Figure 1c). We further found that OsWRKY5 transcripts accumulated in the yellowing sector (region c) more than in the green sector (region a) of senescing flag leaves ( Figure 1d). These results suggest that OsWRKY5 is involved in the onset and progression of leaf senescence in rice.

OsWRKY5 Positively Regulates the Progression of Leaf Senescence
To examine the function of OsWRKY5 in leaf senescence, we identified activation-tagged and loss-of-function mutants. To this end, we obtained two independent T-DNA insertion lines (PFG_3A-15928 and PFG_3A-06060) from the RiceGE database (http://signal.salk.edu/cgi-bin/RiceGE) in which each T-DNA fragment with an activation tag (4 × 35S promoter) was integrated into the promoter region of OsWRKY5 (Figure 2a). Nucleotide sequences used for determining the locations of T-DNA insertions have been posted on the RiceGE database. Based on the comparison with OsWRKY5 promoter sequence, we predicted that T-DNA fragments of oswrky5-D and oswrky5 are inserted in the 1602-bp and 1637-bp upstream of the start codon of OsWRKY5 gene, respectively. They were confirmed by our re-sequencing the regions of T-DNA insertion regions of two mutant lines. To verify the expression levels of OsWRKY5 in these mutant lines, we measured OsWRKY5 expression levels in detached leaves of four-week-old mutant plants during DIS. RT-qPCR showed that OsWRKY5 transcripts accumulated to high levels in PFG_3A-15928 compared to the WT due to the activation-tagged T-DNA insertion; by contrast, in PFG_3A-06060, the T-DNA insertion in the promoter region of OsWRKY5 reduced expression of OsWRKY5 (Figure 2b,c). These results indicate that PFG_3A-15928 and PFG_3A-06060 are dominant activation and recessive knockdown mutants, respectively (hereafter termed oswrky5-D and oswrky5, respectively). To confirm this, we further investigated the OsWRKY5 expression in leaf blade, leaf sheath, and root separated from WT and mutant lines grown in paddy soil for three weeks. Similar to expression patterns of OsWRKY5 shown in detached leaves during DIS (Figure 2b,c), OsWRKY5 transcripts highly accumulated in all tissues of oswrky5-D compared with the WT, while they were significantly lower in oswrky5 than in the WT (Figure 2d,e).
To determine the phenotypic difference between WT and mutant lines during DIS, we next incubated detached leaves of three-week-old WT, oswrky5-D, and oswrky5 plants in 3 mM MES buffer (pH 5.8) at 28 • C under complete darkness. While oswrky5-D showed accelerated leaf yellowing compared with the WT, the oswrky5 leaves retained their green color longer than the WT leaves ( Figure 3a,b). Consistent with the leaf color, the total chlorophyll content of oswrky5-D was less than that of the WT after DIS, whereas oswrky5 maintained higher total chlorophyll levels during DIS compared with the WT (Figure 3c,d).
In senescing leaves, chlorophylls are sequentially degraded by chlorophyll-degrading enzymes associated with upregulation of CDGs, including SGR [4], NYC3 [10], and OsPAO [11]. Many other SAGs are also upregulated during DIS in rice, with products identified as seed imbibition protein (Osh69), glyoxylate aminotransferase (Osh36), and isocitrate lyase (Osl85) [12]. We therefore measured transcript levels of CDGs and SAGs in detached leaves of three-week-old WT, oswrky5-D, and oswrky5 plants under DIS conditions as shown in Figure 2. RT-qPCR analysis revealed that expression of CDGs and SAGs was upregulated in oswrky5-D after 4 days of dark incubation (DDI) (Figure 4a-f) but downregulated in oswrky5 after 5 DDI when compared with the WT (Figure 4g-l). These results demonstrate that OsWRKY5 promotes the onset and progression of leaf senescence by upregulating expression of CDGs and SAGs.
To further examine how OsWRKY5 overexpression affects leaf senescence during vegetative and reproductive stages, we monitored age-dependent leaf yellowing in WT and oswrky5-D plants grown under NLD conditions (>14 h daylight) in the field (37 • N latitude, Suwon, South Korea). While there was no significant difference in leaf color between the WT and oswrky5-D until heading (Figure 5a), the leaves of oswrky5-D showed a precocious leaf senescence phenotype at 40 days after heading (DAH) (Figure 5b,c). The SPAD value, a parameter for leaf greenness, indicated lower levels of green pigments in flag leaves of oswrky5-D compared with the WT at 24 DAH (Figure 5d). Leaf greenness is closely linked to photosynthetic capacity [38,39]. Thus, reduced SPAD value led to a relatively lower Fv/Fm ratio (efficiency of photosystem II) in oswrky5-D than in WT at 24 DAH (Figure 5e). Similar to expression patterns of CDGs and SAGs during DIS, CDG and SAG transcripts were more abundant in the senescing flag leaves of oswrky5-D than in those of WT at 40 DAH (Figure 5f). These results indicate that OsWRKY5 acts as a positive regulator of leaf senescence during both NS and DIS. OsWRKY5 mRNA levels were measured in rice tissues separated from three-week-old WT and mutant lines. Transcript levels of OsWRKY5 in oswrky5-D (b,d) and oswrky5 (c,e) were determined by RT-qPCR and normalized to the transcript levels of OsUBQ5. Mean and SD values were obtained from more than three biological replicates. Asterisks indicate a statistically significant difference from WT, as determined by Student's t-test (* p < 0.05, ** p < 0.01, *** p < 0.001). DDI, day(s) of dark incubation.   Figure 3 (3a,b). Expression of CDGs and SAGs in oswrky5-D (a-f) or oswrky5 (g-l) was compared with that in the WT after 4 or 5 DDI, respectively. Transcript levels of CDGs (a-c and g-i) and SAGs (d-f and j-l) were determined by RT-qPCR analysis and normalized to that of OsUBQ5. Mean and SD values were obtained from more than three biological replicates. Asterisks indicate a statistically significant difference from WT, as determined by Student's t-test (* p < 0.05, ** p < 0.01, *** p < 0.001). Experiments were repeated twice with similar results. CDGs, Chl degradation genes; DDI, day(s) of dark incubation; SAGs, senescence-associated genes. Transcript levels were determined by RT-qPCR analysis and normalized to that of OsUBQ5. Mean and SD values were obtained from more than three biological replicates. Asterisks indicate a statistically significant difference from WT, as determined by Student's t-test (* p < 0.05, ** p < 0.01, *** p < 0.001).

OsWRKY5 Upregulates SenNAC Genes
Previous studies have shown that senNACs including OsNAP and OsNAC2 promote leaf senescence by upregulating expression of CDGs and SAGs [25,40]. To determine whether OsWRKY5 participates in NAC TF-mediated senescence pathways, we examined the expression levels of OsNAP and OsNAC2 in detached leaves of WT, oswrky5-D, and oswrky5 under DIS conditions. RT-qPCR showed that compared with WT, the expression levels of OsNAP and OsNAC2 were higher in oswrky5-D during dark incubation (Figure 6a,b), while they were downregulated at 0 and 5 DDI in oswrky5 compared with WT (Figure 6c,d). These results suggest that OsWRKY5 promote leaf senescence by upregulating the expression of OsNAP and OsNAC2. WRKY TFs regulate the transcription of their target genes by recognizing a consensus cis-element, the so-called W-box [26]. The W-box has been generally defined as 5 -TTGAC(C/T)-3 with an invariant TGAC core sequence essential for WRKY binding [37,41]. Since repetitive TGAC sequences enhance WRKY binding efficiency, we searched for the TGAC core sequence within 2 kb upstream of the transcriptional initiation sites of OsNAP and OsNAC2, and found two regions (-1001~-765 and -657-575) in the promoter of OsNAP and five regions (-1760~-1574, -1411~-1309, -1135~-1021, -660-480, and -352~-98) in the promoter of OsNAC2 (Figure 6e). To investigate whether the OsWRKY5 TF directly binds to the promoters of OsNAP and OsNAC2, we performed yeast one-hybrid assays. However, we could not find any difference between GAL4AD and GAL4AD-OsWRKY5 by measuring β-galactosidase activity of lacZ reporter genes, indicating that OsWRKY5 seems not to interact with the promoter of OsNAP or OsNAC2 (Figure 6f).
Previous studies have reported that the microRNA osa-miR164b is closely associated with the post-transcriptional regulation of OsNAC2, resulting in reduction of OsNAC2 mRNA levels [42,43]. To examine whether OsWRKY5 regulates endogenous levels of osa-miR164b during DIS, we determined the expression of osa-miR164b in detached leaves of three-week-old WT, oswrky5-D, and oswrky5 plants using stem-loop RT-PCR analysis. It revealed no difference in the levels of osa-miR164b among genotypes ( Figure S3), indicating that OsWRKY5 seems not to participate in osa-miR164b-mediated senescence pathway.

OsWRKY5 Is Involved in ABA Biosynthesis Pathway
Among phytohormones affecting the onset and progression of leaf senescence [2], ABA activates senescence-associated regulatory pathways, leading to acceleration of leaf senescence [44]. Genetic studies have revealed that the endogenous ABA concentration is delicately controlled by senNACs such as OsNAP and OsNAC2 [25,40]. Considering that OsWRKY5 upregulated the expression of OsNAP and OsNAC2, we speculated that OsWRKY5 is mainly involved in regulating ABA biosynthesis. Indeed, the endogenous ABA concentration was significantly higher in leaves of three-week-old oswrky5-D plants than in those of the WT (Figure 7a). RT-qPCR analysis showed that ABA biosynthesis genes including OsNCED3, OsNCED4, and OsNCED5 were significantly upregulated in oswrky5-D leaves compared with the WT (Figure 7b). This strongly suggests that the early leaf senescence of oswrky5-D is mainly due to an increase in ABA biosynthesis after heading.  Figure 6A. Transcript levels of ABA biosynthetic genes including OsNCED3, OsNCED4, and OsNCED5 were determined by RT-qPCR analysis and normalized to transcript levels of OsUBQ5. Mean and SD values were obtained from more than three biological replicates. Asterisks indicate a statistically significant difference from WT, as determined by Student's t-test (* p < 0.05, ** p < 0.01). (c) Ten-day-old WT seedlings grown on 0.5X MS phytoagar medium at 28 • C under continuous light conditions were transferred to 0.5X MS liquid medium only (control) or 0.5X MS liquid medium containing 50 µM epibrassinolide (BR), 50 µM gibberellic acid (GA), 50 µM 3-indoleacetic acid (IAA), 50 µM 6-benzylaminopurine (6-BA), 100 µM salicylic acid (SA), 50 µM methyl jasmonic acid (MeJA), 50 µM abscisic acid (ABA), or 50 µM 1-aminocyclopropane-1-carboxylic acid (ACC). Total RNA was isolated from leaves after 4 h of treatment. OsWRKY5 mRNA levels were determined by RT-qPCR analysis and normalized to transcript levels of OsUBQ5. Mean and SD values were obtained from more than three biological replicates. Asterisks on orange bars indicate a statistically significant difference from the control, as determined by Student's t-test (*p < 0.05, ** p < 0.01). Experiments were repeated twice with similar results.

OsWRKY5 Promotes Leaf Yellowing during NS and DIS
We found that OsWRKY5 participates in the ABA-mediated regulatory pathways of leaf senescence. OsWRKY5 was expressed in leaves and its transcription was activated by aging and dark treatment (Figure 1b,c). In addition, we found four W-boxes and a single G-box in the 1500-bp upstream of transcription initiation site of OsWRKY5 ( Figure S4). These cis-elements are recognized by WRKY, bZIP, bHLH, and NAC transcription factors which are involved in the regulation of leaf senescence [28], implying that expression of OsWRKY5 could be regulated by senescence-induced transcription factors. The WRKY domain of OsWRKY5 has the highest amino acid similarity to that of AtWRKY6 ( Figure S1). Similar to the early leaf senescence phenotype of AtWRKY6-OX in Arabidopsis [29], the progression of leaf senescence was much faster in the oswrky5-D mutant than in WT plants under NS and DIS conditions. (Figures 3 and 5), and the oswrky5 knockdown mutant showed markedly delayed leaf senescence (Figure 3).
Many senescence-induced TFs directly or indirectly regulate expression of their target genes, including CDGs, SAGs, and other senescence-associated TFs. OsNAP directly binds to the promoters of SGR, NYC1, NYC3, RCCR1, and Osl57 (encoding a putative 3-ketoacyl-CoA thiolase). OsNAP also indirectly regulates the expression of Osh36 and Osh69, whose amino acid sequences are quite similar to those of Arabidopsis thaliana aminotransferase and Brassica oleracea seed imbibition protein, respectively [25]. OsNAC2 enhances chlorophyll degradation by directly interacting with the promoters of SGR and NYC3 [40]. We therefore speculate that OsWRKY5 upregulates the expression of CDGs and SAGs by regulating senNACs. Upregulation of OsNAP and OsNAC2 was observed in oswrky5-D, resulting in early leaf yellowing (Figures 4 and 5). However, OsWRKY5 does not bind to the promoter regions of OsNAP and OsNAC2 despite the presence of repetitive TGAC core sequences (Figure 6e,f), suggesting that it indirectly regulates expression of these genes.
WRKY TFs can physically interact with other TFs involved in leaf senescence. For example, Besseau et al. (2012) showed that expression of Arabidopsis WRKY30, WRKY53, WRKY54, and WRKY70 is induced during leaf senescence and WRKY53, WRKY54, and WRKY70 interact independently with WRKY30 in yeast two-hybrid assays [32]. In Arabidopsis, WRKY45 functions in GA-mediated leaf senescence by interacting with the DELLA protein RGA-LIKE1 (RGL1) characterized as a repressor of GA signaling [33]. Recently, TT2, a MYB family member, was identified as an interacting partner of WRKY27 in upland cotton (Gossypium hirsutum L.) [45]. Therefore, exploring the possible interaction networks of the OsWRKY5 TF in the regulation of senNACs should provide more insight into the mechanism of leaf senescence.
Interestingly, despite of constitutive expression of OsWRKY5 in oswrky5-D, there were no significant difference in senescence phenotype and expression of CDGs and SAGs between WT and oswrky5-D before leaf senescence (Figures 3a and 5a). Leaf senescence is a complex process involving numerous regulators [27,46]. These regulators are mostly induced by the onset of senescence. OsWRKY5 may require other cofactors to upregulate the expression levels of CDGs and SAGs during leaf senescence. Thus, even though OsWRKY5 transcripts are highly accumulated in oswrky5-D during vegetative growth, it seems that OsWRKY5 overexpression is not much as effective as a senescence promoter, possibly due to a lack of senescence-induced cofactors.

OsWRKY5 Mediates ABA-Induced Leaf Senescence
ABA promotes the onset and progression of leaf senescence [47]. Thus, endogenous ABA levels are elevated by upregulation of ABA biosynthesis genes during leaf senescence, promoting further ABA-induced leaf senescence [44,48]. Arabidopsis 9-CIS-EPOXYCAROTENOID DIOXYGENASE (NCED) genes, encoding a rate-limiting enzyme in ABA biosynthesis, are upregulated during NS [19,49]. Dark incubation induces expression of OsNCED3 in rice leaves [50], and overexpression of OsNCED3 accelerates leaf yellowing in rice during dark incubation. NAP increases ABA biosynthesis by directly upregulating transcription of ABSCISIC ALDEHYDE OXIDASE3 (AAO3), leading to chlorophyll degradation during dark incubation in Arabidopsis [20]. In rice, although transcription of ABA biosynthesis genes, such as OsNCED1, OsNCED3, OsNCED4, and OsZEP, is inhibited by OsNAP, the functional ortholog of Arabidopsis NAP, overexpression of OsNAP leads to precocious leaf senescence by directly regulating CDGs and SAGs [25]. OsNAC2 elevates endogenous ABA content by directly binding to the promoters of OsNCED3 and OsZEP, thereby promoting leaf senescence [40]. Transgenic Arabidopsis plants heterologously expressing foxtail millet (Setaria italica) NAC1 (SiNAC1) show enhanced transcription of ABA biosynthesis genes, NCED2 and NCED3, resulting in early leaf senescence [51]. Although ABA signaling pathways mediated by WRKY TFs are involved in multiple aspects of plant development including leaf senescence [52], molecular evidence for WRKY TF involvement in ABA biosynthesis is limited. In this study, we found that OsWRKY5 upregulates transcription of ABA biosynthesis genes, OsNCED3, OsNCED4, and OsNCED5 (Figure 7b), suggesting that OsWRKY5 functions in the promotion of leaf senescence by increasing ABA biosynthesis (Figure 7a). Furthermore, based on the involvement of OsWRKY5 in OsNAC2 expression (Figure 6b,d), OsWRKY5 probably activates an OsNAC2-mediated ABA biosynthetic pathway (Figure 8). Plants have developed several regulatory mechanisms to restore ABA homeostasis during leaf senescence. For example, in tomato, NAP2 directly regulates expression of genes regulating ABA biosynthesis (NCED1) and ABA degradation (CYP707A2) to establish ABA homeostasis during leaf senescence [53]. ABA-induced OsNAP represses the accumulation of endogenous ABA in rice, indicating that OsNAP participates in a negative feedback mechanism on ABA biosynthesis [25]. Expression of OsNAC2 is differentially regulated by ABA concentration; OsNAC2 expression is upregulated by 20 µM ABA, but inhibited by ABA concentrations over 40 µM [40]. Because the expression of OsWRKY5 is reduced by excessive ABA treatment (Figure 7c), it is highly possible that OsWRKY5 transcription is repressed by excessive concentrations of ABA via a negative feedback regulatory mechanism.

Plant Materials, Growth Conditions, and Experimental Treatments
The Oryza sativa japonica cultivar 'Dongjin' (parental line), and the oswrky5-D and oswrky5 mutants were grown in a growth chamber under LD conditions (14 h light at 28 • C/10 h dark at 25 • C) and in a rice paddy field under NLD conditions (≥14 h sunlight/day, 37 • N latitude) in Seoul, Republic of Korea.
The T-DNA insertion activation-tagged oswrky5-D and knockdown oswrky5 mutants were obtained from the Crop Biotech Institute at Kyung Hee University, Republic of Korea [54,55].

Determination of Photosynthetic Activity, Total Chlorophyll, and SPAD Value
To evaluate photosynthetic activity, the middle section of the flag leaf of plants grown in a paddy field under NLD conditions was adapted in the dark for 10 min. The Fv/Fm ratio was then measured using an OS-30p+ instrument (Opti-Sciences, Hudson, NH, USA). Total chlorophyll content was measured in rice leaves grown in the growth chamber for 4 weeks. Pigment was extracted from detached leaves incubated in complete darkness using 80% ice-cold acetone. After centrifugation at 10,000× g for 15 min at 10 • C, the absorbance of supernatants was measured at 647 and 663 nm using a UV/VIS spectrophotometer (BioTek Instruments, Winooski, VT, USA). The concentration of chlorophyll was calculated as previously described [56]. The change of SPAD value was determined in the flag leaf of plants grown in a paddy field under NLD conditions using a SPAD-502 instrument (KONICA MINOLTA, Tokyo, Japan).

RT-qPCR and Stem-Loop RT-qPCR Analysis
Total RNA was extracted from rice tissues using an RNA Extraction kit (MG Med, Seoul, Korea), according to the manufacturer's instructions. For synthesis of first-strand cDNA, 2 µg of total RNA was used for reverse transcription (RT) in 20 µL volume with oligo(dT) 15 primer and M-MLV reverse transcriptase (Promega, Madison, WI, USA). For quantification of miR164b, stem-loop pulsed RT was conducted from 2 µg of total RNA in 20 µL volume using a miR164b-specific stem-loop primer and M-MLV reverse transcriptase (Promega) with the following conditions: 16 • C for 30 min followed by pulsed RT of 40 cycles at 16 • C for 2 min, 42 • C for 1 min, and 50 • C for 1 s, and then inactivation of reverse transcription at 70 • C for 5 min [57]. All RT products were diluted with 80 µL distilled water.

Determination of ABA Content
Four-week-old leaves of WT and oswrky5-D were pulverized in liquid nitrogen and then homogenized in 80% methanol containing 1 mM butylated hydroxytoluene as an antioxidant. Extracts incubated for 12 h at 4 • C were centrifuged at 4000× g for 20 min. The supernatant was passed through a Sep-Pak C18 cartridge (Waters, Milford, MA, USA) as described previously [59], and the eluate was subjected to an enzyme-linked immunosorbent assay (ELISA) using an ABA ELISA kit (MyBioSource, San Diego, CA, USA) according to the manufacturer's instructions.

Conclusions
We investigated a WRKY TF family member in rice, OsWRKY5, which acts as a positive regulator of leaf senescence. OsWRKY5 upregulates both ABA biosynthesis and transcription of CDGs and SAGs during leaf senescence, thereby promoting leaf yellowing. OsWRKY5 transcription contributes to regulating leaf senescence in rice.

Conflicts of Interest:
The authors declare no conflict of interest.