Nitric Oxide Regulates Dark-Induced Leaf Senescence Through EIN2 in Arabidopsis

Authors

  • Yun-Han Niu,

    1. The National Key Laboratory of Plant Molecular Genetics and National Center for Plant Gene Research (Shanghai), Institute of Plant Physiology & Ecology, Shanghai Institutes for Biological Sciences, the Chinese Academy of Sciences, Shanghai 200032, China
    Search for more papers by this author
  • Fang-Qing Guo

    Corresponding author
    1. The National Key Laboratory of Plant Molecular Genetics and National Center for Plant Gene Research (Shanghai), Institute of Plant Physiology & Ecology, Shanghai Institutes for Biological Sciences, the Chinese Academy of Sciences, Shanghai 200032, China
    Search for more papers by this author

Tel: +86 21 5492 4098; Fax: +86 21 5492 4015; E-mail: fqguo@sibs.ac.cn

Abstract

The nitric oxide (NO)-deficient mutant nos1/noa1 exhibited an early leaf senescence phenotype. ETHYLENE INSENSITIVE 2 (EIN2) was previously reported to function as a positive regulator of ethylene-induced senescence. The aim of this study was to address the question of how NO interacts with ethylene to regulate leaf senescence by characterizing the double mutant ein2-1 nos1/noa1 (Arabidopsis thaliana). Double mutant analysis revealed that the nos1/noa1-mediated, dark-induced early senescence phenotype was suppressed by mutations in EIN2, suggesting that EIN2 is involved in nitric oxide signaling in the regulation of leaf senescence. The results showed that chlorophyll degradation in the double mutant leaves was significantly delayed. In addition, nos1/noa1-mediated impairment in photochemical efficiency and integrity of thylakoid membranes was reverted by EIN2 mutations. The rapid upregulation of the known senescence marker genes in the nos1/noa1 mutant was severely inhibited in the double mutant during leaf senescence. Interestingly, the response of dark-grown nos1/noa1 mutant seedlings to ethylene was similar to that of wild type seedlings. Taken together, our findings suggest that EIN2 is involved in the regulation of early leaf senescence caused by NO deficiency, but NO deficiency caused by NOS1/NOA1 mutations does not affect ethylene signaling.

Introduction

As an essential programmed event rather than a passive process at the ultimate stage of leaf development, leaf senescence is an evolutionarily-selected developmental process that enables nutrient transfer from aging leaves to younger or reproductive structures of the plant (Lim et al. 2007). During leaf senescence, the degeneration of chloroplasts and the disruption in photosynthesis occur first, which are then followed by deterioration and leakiness of the membrane, which finally results in degradation of macromolecules and disintegration of mitochondria and nuclei (Jing et al. 2003; Lim et al. 2007). Interestingly, approximately 10% of Arabidopsis genes are differentially expressed during leaf senescence (Guo et al. 2004). Those genes involved in photosynthesis are downregulated, while the senescence-associated genes (SAGs) and related transcription factors are upregulated. On the other hand, the process of leaf senescence appears to be affected if the transcription and translation processes are disrupted (Buchanan-Wollaston et al. 2003; Lim et al. 2003; Lin and Wu 2004; Buchanan-Wollaston et al. 2005; Lim et al. 2007).

It is known that leaf senescence can be induced by a variety of environmental factors such as extended darkness, drought, salt, heat, cold, and pathogen attacks (Lim et al. 2007). Dark-induced senescence partially mimics natural senescence (Weaver et al. 1998; Guo and Gan 2005). In comparison with natural senescence, dark-induced leaf senescence has the advantage that the initiation of senescence is synchronized among individuals and genotypes. Therefore, dark-induced senescence is a frequently used model system for studies on leaf senescence. Leaf senescence is also regulated by various hormones: ethylene, cytokinin, abscisic acid, salicylic acid (SA) and jasmonic acid (JA) (Lim et al. 2007). Known for its role in fruit ripening, ethylene has also been found to cause premature senescence (Grbic and Bleecker 1995). The ethylene biosynthetic genes, such as ACC synthase (ACS) and ACC oxidase (ACO), were shown to be upregulated in senescing leaves (van der Graaff et al. 2006) with the increased ethylene levels during leaf senescence. Furthermore, leaf senescence was delayed in ein2 (ethylene-insensitive 2), which is deficient in ethylene signal transduction (Grbic and Bleecker 1995). In addition, Arabidopsis mutant lines involved in the ethylene signaling pathway also showed a delayed-senescence phenotype (Oh et al. 1997). Recently, it was reported that ORE1, a NAC transcription factor that positively regulates aging-induced cell death in Arabidopsis and its expression is upregulated concurrently with leaf aging by EIN2 but is negatively regulated by miR164 during leaf senescence (Kim et al. 2009).

Nitric oxide (NO) has long been characterized as an anti-senescence agent. It has been shown that the NO content of peroxisomes isolated from senescent pea leaves is less than that of young leaves (Del Rio et al. 2004). NO emission is negatively correlated with senescence processes (Leshem et al. 1998). Moreover, the NO-deficient mutant nos1/noa1 exhibits an early-senescence phenotype, and dark-induced senescence and chlorophyll loss can be rescued in nos1/noa1 leaves by treatments with NO donors (Guo and Crawford 2005). Interestingly, expressing a NO-degrading enzyme in Arabidopsis led to early leaf senescence phenotypes with a massive upregulation of SAGs (Mishina et al. 2007).

Nitric oxide and ethylene may act antagonistically during the process of maturation and senescence in higher plants. However, the mechanism through which NO modulates leaf senescence processes by interacting with ethylene signaling pathways remains elusive. Given that the NO-deficient mutant nos1/noa1 exhibited an early-senescence phenotype (Guo and Crawford 2005) whereas ein2-1 displayed a delayed-senescence phenotype (Oh et al. 1997; Alonso et al. 1999), we isolated the double mutant ein2-1 nos1/noa1 in order to further investigate the relationships between NO and ethylene pathways during leaf senescence. Our data suggest that EIN2 is involved in NO signaling in the regulation of leaf senescence, but NO deficiency caused by NOS1/NOA1mutations does not affect ethylene signaling.

Results

The double mutant ein2-1 nos1/noa1 shows a delayed leaf senescence phenotype

To analyze the possible genetic linkage between NOS1/NOA1 and EIN2 in the regulation of leaf senescence, we obtained the double mutant ein2-1 nos1/noa1 by crossing the ein2-1 mutant and the nos1/noa1 mutant. Despite the controversy about the role of NOS1/NOA1 in NO biosynthesis (Guo et al. 2003; Crawford et al. 2006; Guo 2006; Zemojtel et al. 2006; Moreau et al. 2008), it has been demonstrated that the mutation of NOS1/NOA1 causes a significant reduction in NO production in Arabidopsis (Guo et al. 2003; He et al. 2004; Zeidler et al. 2004; Bright et al. 2006; Zhao et al. 2007; Zottini et al. 2007) and Nicotiana benthamiana (Asai et al. 2008). Since the nos1/noa1 mutant seedlings had a remarkable yellowish-leaf phenotype (Guo et al. 2003) (Figure 1A), this was used as a phenotypic marker for isolating the ein2-1 nos1/noa1 double mutant. The putative double mutants were confirmed by performing polymerase chain reaction (PCR) and sequence analyses. As shown in Figure 1A, we found the double mutant ein2-1 nos1/noa1 plants exhibited a yellowish-leaf phenotype like the nos1/noa1 mutant, but the leaf size of the double mutant plants was larger with the leaf shape resembling the ein2-1 mutant.

Figure 1.

Dark-induced leaf senescence phenotypes. 

(A) Phenotypes of wild type, nos1/noa1, ein2-1 and the double mutant ein2-1 nos1/noa1 at days 7 and 21 after germination. Bars in upper panels = 1 mm. Bars in lower panels = 1 cm. 
(B) Dark-induced leaf senescence phenotypes of the fully-expanded leaves detached from plants of wild type, nos1/noa1, ein2-1, the double mutant ein2-1 nos1/noa1.
WT, wild type.

To further analyze the genetic relationship between NOS1/NOA1 and EIN2 involved in the regulation of leaf senescence, we characterized the dark-induced senescence phenotype of the double mutant ein2-1 nos1/noa1. A dark-induced leaf senescence assay was performed using fourth rosette leaves detached from 3-week-old plants grown on long-day conditions. As shown in Figure 1B, the visible yellowing phenotypes were more pronounced in the detached leaves of the nos1/noa1 mutant at day 4 in darkness, whereas this nos1/noa1-mediated effect was greatly reduced in the double mutant ein2-1 nos1/noa1. In contrast, dark-induced senescence occurred slowly in the ein2-1 mutant leaves, which stayed green even at day 6 after dark treatment. Interestingly, the dark-induced senescence patterns of the double mutant ein2-1 nos1/noa1 were similar to those of ein2-1, since chlorophyll was almost completely eliminated in both the wild type and the nos1/noa1 mutant by day 6 of darkness.

Leaf yellowing as a consequence of chlorophyll degradation is a convenient visible indicator of leaf senescence. To monitor the leaf senescence processes, the total chlorophyll content was quantified by spectrophotometer. As shown in Figure 2A, the chlorophyll level dropped dramatically to less than 20% of its starting level at day zero, whereas the double mutant ein2-1 nos1/noa1 still maintained a 60% chlorophyll level at days 4 and 5 after dark treatment. In general, the rates of chlorophyll degradation in the ein2-1 and double mutant ein2-1 nos1/noa1 were obviously lower during 6-d-dark treatment. Next, the photosynthetic chlorophyll fluorescence parameter Fv/Fm was measured to determine the maximal quantum yield of photosystem II (PSII). Interestingly, the optimal value for Fv/Fm (0.8) was detected in both wild-type and ein2-1 leaves at the beginning of dark treatment, but a lower Fv/Fm value (0.7) was measured in the nos1/noa1 mutant and in the double mutant ein2-1 nos1/noa1 leaves, indicating that the photochemical efficiency was constitutively impaired in the nos1/noa1 mutant, which cannot be reverted by EIN2 mutations. During leaf senescence at days 4 and 5, the Fv/Fm ratio of the nos1/noa1 mutant leaves was severely reduced to less than 0.1 while the double mutant ein2-1 nos1/noa1 retained a much higher Fv/Fm ratio around 0.7, which was similar to the ratio of ein2-1 (Figure 2B). These results suggest that EIN2 acts downstream of NOS1/NOA1 in regulating dark-induced leaf senescence processes.

Figure 2.

Analysis of chlorophyll degradation and photochemical efficiency of photosystem II (PSII) (Fv/Fm) of detached leaves in darkness.  

(A) Chlorophyll measurements in the fully-expanded leaves detached from the indicated genotypes during dark-induced leaf senescence (n= 5). 
(B) Analysis of photochemical efficiency of PSII (Fv/Fm) in the detached leaves of the indicated genotypes during dark-induced leaf senescence (n= 3). 
Error bars indicate standard deviation. WT, wild type.

EIN2 mutations rescue the susceptible integrity of thylakoid membranes in the nos1/noa1 mutant under dark treatment

It is well known that the degeneration of chloroplasts and the disruption of photosynthesis occurs first during leaf senescence with deterioration and leakiness of the membrane (Jing et al. 2003; Guo and Gan 2005; Lim et al. 2007). Given that the double mutant ein2-1 nos1/noa1 shows a delayed leaf senescence phenotype, we attempted to investigate the effect of EIN2 mutations on the integrity of thylakoid membranes in chloroplasts of the nos1/noa1 mutant leaves during dark-induced leaf senescence. In the representative leaves detached from the mature plants, detailed examinations by transmission electron microscopy (TEM) revealed that chloroplast structures were severely disrupted and the configurations of thylakoid systems nearly disappeared in the nos1/noa1 mutant under dark treatment for 4 d (Figure 3B, F). In contrast, the granum-stroma thylakoid membranes were still visible with broken stromal membranes and thick granal stacks in the double mutant ein2-1 nos1/noa1 (Figure 3D, H). Interestingly, large and thick granal stacks were observed in the chloroplasts of darkened ein2-1 mutant leaves (Figure 3C, G), while there were no large and thick granal stacks in the wild type chloroplasts (Figure 3A, E). It should be noted that a lot of electron-dense deposits appeared in the chloroplasts of the nos1/noa1 mutant and the double mutant ein2-1 nos1/noa1, respectively, and that these deposits are likely to be osmiophilic granules that increased and then converged into bigger granules along with the decomposition of the thylakoid membranes during the process of dark-induced leaf senescence. According to the TEM images shown in Figure 3, most of the starch granules were decomposed in the chloroplasts of the nos1/noa1, ein2-1 and ein2-1 nos1/noa1 double mutants, whereas a few starch granules were still observed in wild-type chloroplasts under dark treatment for 4 d. These results further support the conclusion that NO plays an important role in maintaining the stability and integrity of thylakoid membranes through EIN2 during dark-induced leaf senescence.

Figure 3.

Chloroplast structures examined by transmission electron microscopy (TEM). 

The chloroplast structures in the fully-expanded leaves detached from the plants of wild type, nos1/noa1, ein2-1 and the double mutant ein2-1 nos1/noa1 under control conditions (A–D) or in darkness for 4 d (E–H), respectively. Bars = 0.5 μm. GL, granal lamellar thylakoids; OG, osmiophilic granules; SG, starch granules; SL, stroma lamellar thylakoids.

EIN2 mutations inhibit the early activation of some known senescence-associated genes in the nos1/noa1 mutant under dark treatment

To further investigate the role of EIN2 in NOS1/NOA1-mediated leaf senescence, we analyzed the relative expression of known senescence-associated genes in the wild type, nos1/noa1, ein2-1 and ein2-1 nos1/noa1 under dark treatment conditions by quantitative reverse transcription (qRT)-PCR. We first found that the relative expression levels of both ACD1 (Pruzinska et al. 2003) and PPH (Schelbert et al. 2009), which are well-known as the key components involved in the chlorophyll breakdown pathway, were rapidly upregulated by dark treatment in the nos1/noa1 mutant compared with that in the wild type leaves, whereas the expression levels of these two genes were severely inhibited in the darkened leaves of both ein2-1 and ein2-1 nos1/noa1 plants (Figure 4). Next, we examined the expression levels of three senescence-associated transcription factors NAP, NAC2 and WRKY6. The results showed that the expression levels of three transcription factors in the nos1/noa1 mutant leaves were significantly upregulated compared with that in the wild type leaves at days 3 after dark treatment. However, such an upregulated expression pattern was not detected in the ein2-1 and double mutant ein2-1 nos1/noa1 during 5-d-dark treatment (Figure 4). Furthermore, the expression patterns of another two senescence-associated genes SEN1 and APG7 were similar to the patterns detected in ACD1 and PPH in the leaves of indicated genotypes (Figure 4). Taken together, the expression patterns of these senescence-associated genes mentioned above are in agreement with the delayed-senescence phenotypes as exhibited in the double mutant ein2-1 nos1/noa1 under dark treatment.

Figure 4.

Expression analysis of the known senescence-associated genes during dark-induced leaf senescence. 

mRNA levels in the fully-expanded leaves detached from the plants of the indicated genotypes were examined by quantitative reverse transcription polymerase chain reaction (qRT-PCR). For qRT-PCR analysis, Actin2 was used as the internal standard. Error bars indicate standard deviations of three technical replicates, and the results were consistent in three biological replicates. ACD1, ACCELERATED CELL DEATH 1; APG7, AUTOPHAGE 7; NAC2, NAC DOMAIN CONTAINING PROTEIN 2; NAP, NAC DOMAIN CONTAINING PROTEIN; PPHPHEOPHYTINASE; SEN1, SENESCENCE 1; WRKY6, WRKY DNA-BINDING PROTEIN 6; WT, wild type.

NO deficiency caused by NOS1/NOA1 mutations does not affect ethylene responses in the nos1/noa1 mutant

Considering that NOS1/NOA1 regulates dark-induced leaf senescence through EIN2, we tested whether NOS1/NOA1 mutations also disrupt the ethylene signal transduction pathway. We grew etiolated seedlings of wild type, nos1/noa1, ein2-1 and ein2-1 nos1/noa1 on medium with different concentrations of ACC (0 to 100 μM) in order to test the ethylene triple responses. As shown in Figure 5A and B, the triple responses in the nos1/noa1 mutant seedlings were almost the same as the responses observed in the wild type seedlings when treated with different concentrations of ACC, indicating that NO deficiency caused by NOS1/NOA1 mutations has little effect on the triple responses to ethylene. As expected, the triple responses of ein2-1 nos1/noa1 were similar to the response observed in the ein2-1 mutant, exhibiting an ethylene-insensitive phenotype (Figure 5A, B). Thus, we proposed that NOS1/NOA1 regulates dark-induced leaf senescence through EIN2, but NOS1/NOA1 is not involved in the regulation of ethylene signal transduction.

Figure 5.

Sensitivity analysis in response to ethylene treatments. 

(A) The triple response phenotypes of the indicated genotypes to different concentrations of ACC. Bars = 1 mm. 
(B) Measurements of hypocotyl length of the indicated genotype seedlings in response to different concentrations of ACC. Error bars indicate standard deviation (n= 30). 
ACC, 1-Aminocyclopropane-1-Carboxylate; WT, wild type.

Discussion

Senescence is a highly-regulated process terminating with programmed cell death (PCD). Leaf senescence is regulated by different endogenous factors resulting from developmental and environmental signals. With regard to regulatory networks, a feature of leaf senescence is hormonal crosstalk to regulate the expression of senescence-associated genes (SAGs) (Jing et al. 2003; Guo and Gan 2005; Lim et al. 2007). NO has long been noted as an anti-senescence signal in that the exogenous application of NO extends the postharvest life of fruits and vegetables (Leshem et al. 1998) and counteracts rice leaf senescence induced by the treatment of abscisic acid (Hung and Kao 2003). Senescence-dependent chlorophyll degradation can be delayed by NO treatments in soybean (Glycine max) cotyledons (Jasid et al. 2009). NO-deficient mutants showed an early leaf senescence phenotype compared with wild type plants (Guo and Crawford 2005; Mishina et al. 2007). By contrast, it is well known that ethylene plays an important role in promoting leaf senescence (Grbic and Bleecker 1995). Mutations in EIN2, encoding a master positive regulator in the ethylene signaling pathway, were shown to delay leaf senescence (Grbic and Bleecker 1995; Oh et al. 1997; Jing et al. 2002). Although accumulated evidence suggests that EIN2 plays an important role in leaf senescence, how EIN2 functions in the regulation of senescence processes is not fully understood. Recently, it was reported that EIN2 together with a NAC transcription factor ORE1 and the miRNA164 regulates age-dependent cell death and leaf senescence in Arabidopsis (Kim et al. 2009). In recent years, there have been major advances in our understanding of the molecular mechanisms regulating leaf senescence processes with respect to NO and ethylene signaling pathways, respectively. However, the genetic evidence for the linkage between the two signaling pathways remains unexplored.

In this study, we demonstrate that EIN2 is involved in NO signaling in the regulation of dark-induced leaf senescence in Arabidopsis. The results described in this paper provide several lines of evidence indicating that EIN2 acts downstream of NO signaling as a regulatory cross point between NO and ethylene signaling pathways. First, the nos1/noa1-mediated, dark-induced early senescence phenotype was partially suppressed by mutations in EIN2, suggesting that EIN2 functions in nitric oxide signaling to regulate leaf senescence. Second, detailed examinations by transmission electron microscopy (TEM) revealed that chloroplast structure in the nos1/noa1 mutant was severely degraded after 4 d of darkness, whereas this nos1/noa1-mediated degradation was partially reverted in the ein2-1 nos1/noa1 double mutant. These results further support the conclusion that EIN2 functions downstream of NO signaling and plays a critical role in maintaining the integrity of thylakoid membranes during dark-induced leaf senescence. Third, in agreement with changes in leaf senescence phenotypes, the rapid upregulation of the known senescence marker genes in the nos1/noa1 mutant was significantly delayed in the ein2-1 nos1/noa1 double mutant. These results confirm that NO acts as a negative regulator in activating the expression of some senescence-marker genes through EIN2.

It has long been known that NO levels are negatively correlated with ethylene levels in natural senescence (Leshem et al. 1998). By detecting emissions of NO and ethylene in postharvest strawberries and avocados, a clear-cut stoichiometric relationship was found between the two gases: unripe fruit exhibits high NO levels and low ethylene levels, with the reverse situation in ripe fruit (Leshem and Pinchasov 2000). Treating tomatoes with NO delayed the burst of ethylene production and color development in both mature green (MG) and breaker (BR) stage fruits (Eum et al. 2009). Accumulated evidence suggests a possible linkage between NO and ethylene pathways in determining the onset of senescence. These findings suggest that endogenous NO may regulate ethylene-promoted fruit senescence. According to our results, EIN2 appears to be one of the key components involved in NO signaling to modulate leaf senescence.

It has been reported that attached and detached nos1/noa1 mutants showed strongly accelerated dark-induced senescence compared with leaves of wild type plants (Guo and Crawford 2005). Leaf senescence is characterized by an increase in expression of a multitude of genes that are known as SAGs (BuchananWollaston 1997; Gepstein et al. 2003). In accordance with NO acting as a factor delaying leaf senescence, the expression of SAGs in plants made to be NO-deficient through the expression of an NO-degrading enzyme occurred earlier than that in wild type and in empty vector control plants (Mishina et al. 2007). In this study, we have also found that the expression of several known SAGs was rapidly upregulated in nos1/noa1 mutants compared with wild type plants during dark-induced leaf senescence. Importantly, we demonstrate that the rapid upregulation of SAGs is dramatically inhibited, comparable to the delayed expression patterns observed in ein2-1 mutant leaves, indicating that EIN2 acts as a determinant factor in activating the expression of SAGs caused by endogenous NO-deficiency during leaf senescence. Interestingly, we also demonstrate that NOS1/NOA1-dependent NO generation is not required for ethylene responses because no difference was observed between the nos1/noa1 mutant and wild type. The ein2-1 nos1/noa1 double mutant showed a similar insensitive phenotype as exhibited in the ein2-1 mutant. This suggests that during leaf senescence, NO signals are mediated by EIN2 in regulating the expression of SAGs, whereas NO is not involved in ethylene responses mediated by EIN2. Considering that NOS1/NOA1 and NR-mediated biosynthesis of NO is responsible for most of the NO production in Arabidopsis (Lozano-Juste and Leon 2010), we can not exclude that NO generated by other sources in the nos1/noa1 mutant might be sufficient to support ethylene responses.

Our analysis of the ein2-1 nos1/noa1 double mutant has revealed the role of EIN2 in NO signaling in the regulation of dark-induced leaf senescence. In this study, we demonstrate that EIN2 is required for the NO-deficiency induced early senescence phenotypes in Arabidopsis. Given that NO signaling and responses are part of an intricate cross-talk network with other hormones, future work is needed to determine the NO signaling components that participate in hormonal regulation of leaf senescence mediated by EIN2.

Materials and Methods

Plant materials and growth conditions

Arabidopsis thaliana Columbia ecotype was used as WT. The nos1/noa1 and ein2-1 mutants were kindly provided by Dr Crawford (University of California at San Diego, USA). Seeds were germinated and grown under a long-day condition, 16 h of white light (80 μmol/m2 per s) and 8 h in dark, with 60% relative air humidity at 21 °C. For dark-induced leaf senescence, fully extended rosette leaves were detached from 3-week-old plants, and then incubated on Petri dishes containing three layers of filter papers soaked in 15 mL distilled water. Petri dishes were wrapped with double-layer aluminum foil, and kept in the dark at 22 °C.

ACC treatments

Seeds were surface-sterilized with 70% ethanol and 30% bleach for 5 min, then washed with 75% ethanol three times and dried in the air. The sterilized seeds were plated on 9 cm diameter Petri dishes with half-strength Murashige and Skoog (MS) medium (1% sucrose), with and without different concentrations of ACC. Petri dishes were sealed with Parafilm tape and kept in the dark at 4 °C for 4 d. After that, Petri dishes were transferred to 22 °C and then germinated in the dark for 80 h. The hypocotyl length of the ACC-treated seedlings was individually measured.

Chlorophyll measurements

The chlorophyll content was determined as described previously (Arnon 1949). The chlorophyll in leaf samples was extracted with 80% (v/v) acetone. Chlorophyll content was determined spectrophotometrically.

Chlorophyll fluorescence measurements

Chlorophyll fluorescence emissions were detected with an LI-6400XT Portable Photosynthesis System (LI-COR Biosciences, Lincoln, NE, USA). The maximum photochemical efficiency of PSII was determined from the ratio of variable (Fv) to maximum (Fm) fluorescence (Fv/Fm).

Transmission electron microscopy

The fully expanded leaves were cut into several sections and fixed with 2.5% (v/v) glutaraldehyde and 2% (v/v) paraformaldehyde (pH 7.4) at 4 °C overnight. Thin sections were examined by a transmission electron microscope (H7650, Hitachi, Tokyo, Japan) using a voltage of 120 kV.

RNA Isolation and Analysis

Total RNA was extracted using the TRIzol reagent according to the manufacturer's recommendations (Takara, Dalian, China). DNA contaminated in total RNA samples was digested with RNase-free DNase (Takara). Complementary DNA was produced using total RNA and an oligo (dT) 18 primer. Quantitative real-time PCR was performed with SYBR Premix Ex TaqII (Takara) using a MyiQ5 single color Real-Time PCR Detection System (Bio-Rad, Hercules, CA, USA). The comparative threshold cycle (Ct) method was used for determining relative transcript levels (iQ5 admin, Bio-Rad) using ACTIN2 as an internal control. The primers used for the real-time PCR analysis are as follows: ACD1(AT3G44880), 5′-ACGGCATGGTAAGAGTCAGC-3′ (forward) and 5′-AAACCAGCAAGAACCAGTCG-3′ (reverse); PPH (AT5G13800), 5′-CAATCATGCTTGCTCCTGGTG-3′ (forward) and 5′-CTACCAATCCTGGACTCCTCC-3′ (reverse); NAP (AT1G69490), 5′-GCCATTCACAGCGGTTCAAG-3′ (forward) and 5′-CAACAAATGAGCCAGCGAAC-3′ (reverse); NAC2 (AT5G39610), 5′-TCGTCTCACGTGACCTGCTT-3′ (forward) and 5′-GGACTCGTGGACAAGTCTTTTGT-3′ (reverse); WRKY6 (AT1G62300), 5′-TAGTCACGACGGGATGATGA-3′ (forward) and 5′-ATTAGGAGGCGGAGGTGAGT-3′ (reverse); SEN1 (AT4G35770), 5′-AACATGTGGATCTTTCAAGTGCC-3′ (forward) and 5′-GTCGTTGCTTTCCTCCATCG-3′ (reverse); APG7 (AT5G45900), 5′-CCACCTCTCTGTCTGATGATATGC-3′ (forward) and 5′-GACCAAGCAGCGTGATCTGTGAG-3′ (reverse); and ACTIN2 (AT3G18780), 5′-GCCATCCAAGCTGTTCTCTC-3′ (forward) and 5′-GCTCGTAGTCAACAGCAACAA-3′ (reverse).

(Co-Editor: Hai-Chun Jing)

Acknowledgements

We thank X.Y. Gao (Shanghai Institutes for Biological Sciences, the Chinese Academy of Sciences) for assistance with electron microscopy. This research was supported by the National Natural Science Foundation of China (31170244 and 30770198).

Ancillary