• Nitric oxide (NO) has been reported to be involved in breaking seed dormancy but its mechanism of action is unclear.
• Here, we report that a rapid accumulation of NO induced an equally rapid decrease of abscisic acid (ABA) that is required for this action in Arabidopsis.
• Results of quantitative real-time polymerase chain reaction (QRT-PCR) and Western blotting indicate that the NO-induced ABA decrease correlates with the regulation of CYP707A2 transcription and (+)-abscisic acid 8′-hydroxylase (encoded by CYP707A2) protein expression. By analysing cyp707a1, cyp707a2 and cyp707a3 mutants, we found that CYP707A2 plays a major role in ABA catabolism during the first stage of imbibition.
• Fluorescent images demonstrate that NO is released rapidly in the early hours at the endosperm layer during imbibition. Evidently, such response precedes the enhancement of ABA catabolism which is required for subsequent seed germination.
Some environmental factors such as light intensity and low temperatures are known as regulators of seed dormancy (Holdsworth et al., 2008). Roles of abscisic acid (ABA) and gibberellic acid (GA) in dormancy and germination have been reported by many investigators (Bewley, 1997; Nakajima et al., 2006; Carrera et al., 2008; Holdsworth et al., 2008). Several signaling molecules, such as nitric oxide (NO) and some reactive oxygen species (ROS), have also been reported to be involved (Batak et al., 2002; Bethke et al., 2004, 2006a; Sarath et al., 2007). Components in the signaling pathway have been identified. For example, the dog1 mutant is completely nondormant and does not show obvious pleiotropic phenotypes, indicating that DOG1 plays a crucial role in dormancy (Bentsink et al., 2006). The absence of histone H2B monoubiquitination in the Arabidopsis hub1 (rdo4) mutant also reveals a role for chromatin remodeling in seed dormancy (Liu et al., 2007). However, the detailed mechanisms of dormancy holding and breaking remain unclear but interesting research topics.
The relationship between NO signaling and ABA response has been demonstrated by some investigators (Bright et al., 2006; Sarath et al., 2007; Zhang et al., 2007; Neill et al., 2008). For example, ABA-induced guard cell closure needs participation of NO (Durner & Klessig, 1999; Foissner et al., 2000; Desikan et al., 2002; Neill et al., 2002, 2008; Bright et al., 2006). However, the relationship between NO and ABA in seed germination has yet to be established. In this study, we demonstrate that rapidly accumulated NO induces an equally rapid ABA decrease that is required for the breaking of seed dormancy and germination in Arabidopsis. Such action correlates to the regulation of ABA catabolism since mutant cyp707a2 did not show the response when compared with wild-type seeds. Results of QRT-PCR and west-blotting indicate that the NO-induced ABA decrease correlates with the regulation of CYP707A2 transcription and the ABA 8′-hydroxylase protein expression.
Materials and Methods
The Arabidopsis (Arabidopsis thaliana) (Col0 and some T-DNA inserted mutants from Col0) plants were grown in a growth chamber with a 16-h photoperiod at a photoflux density of c. 200 µmol m−2 s−1 at a day temperature of 23°C and a night temperature of 20°C. In order to minimize the effect of seed maturation and storage conditions, plants of each genotype tested were grown in different sections of the same pot and seeds were harvested at the same time. Seeds were harvested in bulk 30 d after the petals appeared on the first flowers. These seeds maintained stronger dormancy. Only freshly harvested seeds were used to do the examination. The rest of seeds were stored at −80°C. Dormancy can be maintained for > 1 yr at −80°C (Millar et al., 2006; Fujii et al., 2007).
T-DNA insertion line
The seeds of Arabidopsis thaliana cyp707a1, cyp707a2, cyp707a3 (SALK_069127, SALK_083966, SALK_026545) generated by Salk Institute Genomic Analysis Laboratory (http://signal.salk.edu/) were obtained from the Arabidopsis Biological Resource Center (ABRC) (Ohio State University, Columbus, OH, USA). The seeds were planted on agar plates containing kanamycin and the kanamycin-resistant plants were transferred to soil. Seeds were harvested separately from individual plants. Subsequently, to confirm the mutant line as homozygous, PCR was performed with the genomic DNA of cyp707a1, cyp707a2, cyp707a3 mutants using gene-specific oligonucleotides: cyp707a1 (LP TAGCCATCAGGACTTTGAAGC; RP CCAAAACCCAATACGTTCATG); cyp707a2 (LP AATCCCAAATATGCCTTAGGC; RP: TATGTGGGGA CTTTGATGGAC); cyp707a3 (LP CGTAAGAATCAAACTGTTACATCAG; RP CAGGTTGGTACACCTTCAAAATG); and LB primer (GCGTGGACCGCTTGCTGCAACT).
Sodium nitroprusside (SNP) and SNAP (Sigma-Aldrich) were used as NO donor to release NO steadily and 2-(4-carboxyphenyl)-4,4,5,5-tetramentylimidazoline-1-oxyl-3-oxide (c-PTIO) (Sigma) was used as NO scavenger (Kopyra & Kopyra, 2003; Bright et al., 2006). Nordihydroguaiaretic acid (NDGA) and diniconazole (Sigma) were used as the inhibitors of 9-cis epoxcartenoid dioxygenase and ABA 8′-hydroxylase, respectively (Han et al., 2004; Kitahata et al., 2005).
Fifty seeds were placed in 55-mm diameter Petri dishes with three Whatman No. 1 filter papers and 2.2 ml of sterile double-distilled water or treatment solutions. Plates were then placed in a 21°C growth chamber under continuous light at 100 µm m−2 s−1 for 7 d. The seeds were regarded as germination when radicle emerged. Experiments were performed in quadruple for each treatment.
Determination of NO
Nitric oxide was detected by the Nitric Oxide (total) Detection Kit (Assay Designs, Ann Arbor, MI, USA). The mechanism of this kit is to transform NO to nitrite, which is then measured (Rockel et al., 2002). About 0.2 g seeds were put into 1.5 ml tubes and then with 200 µl reaction buffer and 100 µl diluted NADH solution added (100 µl water was added into a parallel tube as control). Then, 100 µl nitrate reductase (NR) was added to samples and 100 µl reaction buffer added to the control tubes. Another 400 µl reaction buffer without seeds, NADPH and NR acted as blank. The blank, control and sample tubes were mixed well and incubated at 37°C for 30 min. After incubation they were centrifuged at 3000 g for 1 min and 300 µl supernatant was transformed to a new tube. The 100 µl Griess Reagent I was then added to control, sample and blank tubes and, after being well mixed, 100 µl Griess Reagent II was added. Mixing was achieved by shaking followed by incubation at 25°C for 10 min. The optical density (OD) of samples and controls were measured at 540 nm. Each sample OD was marked as ODs and each control OD marked as ODc. The average net OD was then calculated and marked as ODn. Each average Odn = average Ods – average ODc. Each ODn could be calculated from standard curve; 0–100 µm sodium nitrate was used as standard for the standard curve. The NO content released equaled the nitrate content.
Laser scanning confocal microscopy
Laser scanning confocal microscopy (LSCM; Zeiss 510) was used to produce a NO image. Measurement of NO was performed with the specific NO dye DAF-FM (4-amino-5-methylamino-2′,7′-difluorofluorescein), using the method as described by Corpas et al. (2004) with slight modifications. Seeds were incubated in loading buffer (0.1 mm CaCl2, 10 mm KCl, 10 mm 2-(N-morpholino) ethanesulfonic acid (MES)–Tris, pH 5.6, and DAF-FM at a final concentration of 10 µm) for 1 h in the dark at 25°C, and followed by washing with loading buffer for 1 h. Before detection the testa of seed was removed and the seed was cut along the axis. All images were visualized using LSCM (excitation at 488 nm and emission at 535 nm). To enable the comparison of changes in signal intensity, confocal images were taken under identical exposure conditions for all the samples. The experiment was repeated three times for further verification.
Extraction and determination of ABA
The method of ABA extraction and determination was modified according to Cheng et al. (2002). Fresh samples c. 0.5 g were homogenized and extracted in 5 ml 80% acetone containing 2,6-di-tert-butyl-4-methylphenol (200 mg l−1) with grinding and incubated at 4°C for 16 h. One milliliter of 1 m phosphate buffer at pH 8.0 was added to the extract, 1 ng of (±)-[1,2-13C2] abscisic acid (from Hangzhou Qirui Biomedical Technology Co. Ltd (Hangzhou, China) according to Asami et al., 1999) was added to each sample as internal standard to percentage recovery. After distilling acetone on a rotary evaporator, lipids were removed by partitioning the aqueous concentrate twice with 5 ml hexanes. The pH of the aqueous phase was adjusted to 2.5 with 6 n HCl and extracted three times with 5 ml ethyl acetate. The acidic fraction was dried and dissolved in 1ml methanol, the solution subjected to HPLC on a µBondapak C18 (30 × 0.78 cm column; Waters, Milford, MA, USA). The elution Buffer was 45% methanol containing 1% acetic acid. The ABA was collected from 10.0 to 12.0 min. The fractions containing ABA were dried and methylated with diazomethane. The methylated ABA was used for GC-MS analysis.
The GC-MS analysis was performed with a mass spectrometer (HP5973; Hewlet-Packard) coupled to a gas chromatograph (HP6980; Hewlet-Packard) with a DB-1 capillary column. The injection temperature was set at 250°C, and the inlet pressure was 70 kPa. After injection, the oven temperature was maintained at 80°C for 1 min, increased to 290°C at a rate of 20°C min−1, and then kept at 290°C for 5 min. Mass spectra were obtained under the following conditions: electron energy 1.5 kV; ion source temperature 250°C; capillary interface temperature 250°C; ion range 50–450 (mass-to-charge ratio). Abscisic acid was analysed by GC-selected ion monitoring MS by monitoring m/z at 192 ((±)-[1,2-13C2] abscisic acid) and 190 (endogenous ABA).
Sodium dodecyl sulfate–polyacrylamide gel electrophoresis (SDS-PAGE) was performed as described by Laemmli (1970). Forty micrograms proteins was solubilized and separated on a 10% (w : v) acrylamide gel. Total protein (50 µg) was separated by 12% SDS-PAGE then transferred to a nitrocellulose membrane, the membrane was blocked for 60 min with 5% (w : v) nonfat milk in 0.05% (w : v) Tween 20, 10 mm Tris (pH 8.0) and 150 mm NaCl. The antibody (produced in rabbit; AbMART Inc. (Shanghai, China) against ABA 8′-hydroxylase encoded by CYP707A2 was added and incubated at 4°C overnight. After washing, the alkaline horseradish peroxidase (HRP)-coupled secondary antibody was added and incubated at room temperature for 1.5 h. The color was developed with a solution containing H2O2. The antigen obtained by chemosynthesis, the protein sequence of chemosynthesis is C-SNFFSSLYADEPALIT-NH2. This antigen was then used to immunize the rabbit and to obtained specific antibody from it.
Total RNA was isolated from seeds or leaves by RNeasy kit (Invitrogen). DNA impurities in the isolated RNA were digested before synthesizing the cDNA by adding DNase (Invitrogen) and incubating for 30 min at 37°C. DNase was then inactivated by incubating for 10 min at 65°C and removed with the digested DNA after incubation. Two micrograms of RNA was reversed to cDNA with SuperScriptIII RTS First-Strand cDNA Synthesis Kit (Invitrogen). The cDNA was then diluted for 10 times and 4 µl cDNA was used to perform the QRT-PCR. IQ SYBR Green Supermix (Bio-Rad) was used to perform the QRT-PCR. Actin2 acted as the internal standard. The QRT-PCR was executed with iCycle (Bio-Rad). The primers that were used in QRT-PCR were: CYP707A1 (forward TTGGAAAGAGGAGACTAGAG; reverse GTGAACCACAAAAGAGGAAC); CYP707A2 (forward AAATGGAGTGCACTCATGTC; reverse CCTTCTTCATCTCCAATCAC); CYP707A3 (forward ATTCTTGTCCAGGCAATGAG; reverse ATAGGCAATCCATTCTGAGG); CYP707A4 (forward GAAAGGAATACAGTACAGTC; reverse GGATTAGATTTGGCTAACTAC), NCED6 (forward TGAGAGACGAAGAGAAAGAC; reverse GTTCCTTCAACTGATTCTCG); AAO3 (forward GAAGGTCTTGGAAACACGAAGAA; reverse GAAATACACATCCCTGGTGT ACAAAAC); Actin2 (forward GTGAAGGCTGGATTTGCAGGA; reverse AACCTCCGATCC AGACACTGT).
Generation of CPY707A2 overexpressing and complementary plants
Full-length Arabidopsis CPY707A2 cDNA was obtained by using reverse-transcriptase PCR and cloned into pENTR-TOPO cloning vector (Invitrogen) and sequenced. After LR reaction, CPY707A2 cDNA was inserted into pGWB5 vector (a gift from Prof. Liang, Yangzhou University, China) which had 35S promoter, we named this vector pGWB5-CPY707A2. Transgenic Arabidopsis using cauliflower mosaic virus (CaMV) 35S promoter was generated using the floral dipping method (Clough & Bent, 1998) to transfer into Col-0 wild-type plants or CPY707A2 null plants. Transformed plants were selected by growth on Hygromycin-containing media. Plants of the second generation after transformation were used for the experiments. Vacant pGWB5 vectors, acting as control, were also transferred into Col-0 wild-type plants or CPY707A2 null plants.
Sequence data from the article can be found in the GenBank data libraries or TIGR database (Arabidopsis thaliana Genome Project) under the following accession numbers: CYP707A1, AT4G19230; CYP707A2, AT2G29090; CYP707A3, AT5G45340; CYP707A4, AT3G19270; NCED6, AT3G24220; and AAO3, AT2G27150.
Effect of NO on seed dormancy
Manipulation of NO levels in imbibed seeds regulated seed dormancy (Fig. 1).Freshly harvested Arabidopsis seeds (Col0) have strong dormancy as germination rates only reached 37% after 7 d of imbibition. When NO concentrations were increased by treating seeds with SNP, which is an NO donor, at concentrations greater than 25 µm, seed dormancy was reduced significantly. Furthermore, SNP at 200 µm broke the dormancy entirely, with germination levels reaching over 90% after 7 d of imbibition (Fig. 1a), whereas the effect of SNP could be overridden completely by treatment with 200 µm NO scavenger c-PTIO. In contrast to SNP, c-PTIO enhanced seed dormancy and decreased the germination rate of freshly harvested seeds to < 10%, though c-PTIO did not inhibit the germination of nondormancy seeds (Fig. 1b). As an additional test a second NO donor, SNAP at 100 µm was used in the experiment on seeds germination; we found that the effect of SNAP was similar as SNP and its function also could be reversed by c-PTIO (Fig. 1c).
As shown in Fig. 2a, the release of NO in the imbibed seeds was rapid and significant in the initial hours and reached a peak at 3 h before decreasing to low levels after 6 h. With the NO-specific dye, the fluorescent image intensity was enhanced for the first 3 h of imbibition before decreasing after 6 h in Arabidopsis seeds, while c-PTIO treatments reduced fluorescent intensity (Fig. 2b).
Effect of ABA biosynthesis and ABA catabolism on seed dormancy
Seed dormancy could be regulated by the endogenous ABA level (Fig. 3). The ABA contents decreased rapidly in the first 12 h of imbibition and stayed at a low level throughout the process (Fig. 3a). The decrease could be attributed to the ABA 8′-hydroxylation pathway as in other tissues (Cutler, 1999). In Arabidopsis ABA 8′-hydroxylase is encoded by CYP707A gene family (CYP707A1 to CYP707A4). Our results show that their transcription increased during the first 6 h of imbibition and decreased to a lower level after 12 h. Among them, CYP707A2 transcription was much higher than the others and CYP707A4 transcription was the lowest (Fig. 3b).
Transcripts of two ABA biosynthesis genes, AAO3 and NCED6, decreased during the first 6 h before increasing slightly thereafter (Fig. 3c). When ABA biosynthesis inhibitors NDGA (inhibiting NCED) and ABA catabolism inhibitor diniconazole (inhibiting ABA 8′-hydroxylase) (Creelman et al., 1992; Han et al., 2004; Kitahata et al., 2005) were used, NDGA did not have any effect on the germination of freshly harvested seeds whereas diniconazole enhanced seed dormancy significantly (Fig. 3d). Germination of freshly harvested seeds were shown to be inhibited by diniconazole to < 10%, indicating that ABA catabolism was the primary factor that regulates seed dormancy.
Dormancy of knockout, complementary and overexpression lines of the CYP707A family
Freshly harvested seeds of wild type, cyp707a1, cyp707a2 and cyp707a3 mutants were used to test their germination. Insertion lines of these mutants (Fig. 4a) are from ABRC, (Alonso et al., 2003) and homozygous lines were selected and analysed after their third generation. As shown in Fig. 4b cyp707a1, cyp707a2 and cyp707a3 were all homozygous.
The cyp707a1 and cyp707a3 mutants showed similar phenotype to wild type in freshly harvested or after-ripen seeds (Fig. 4c) with the seeds of cyp707a3 mutant having a slightly higher germination rate than the wild type. However, cyp707a2 mutant showed a strong dormancy in freshly harvested seeds. Seeds of wild type, cyp707a1 and cyp707a3 mutants had their dormancy entirely broken 18 d after ripening, but cyp707a2 mutant still maintained strong dormancy until 30 d (Fig. 4d).
The ABA contents of freshly harvested seeds were also measured during imbibition. As shown in Fig. 5a,b,d, ABA concentrations decreased rapidly during the first 12 h in wild type, cyp707a1 and cyp707a3 mutants, before staying at a low level after that. In comparison, ABA concentrations in cyp707a2 mutant decreased at a much slower rate and maintained a relatively high concentration of ABA after 48 h imbibition (Fig. 5c). This indicates that it is CYP707A2 that played a crucial role in the rapid decrease of ABA during the early stage of imbibition. We also produced a complementary line of cyp707a2 mutant and an overexpression line of CYP707A2 (Fig. 6). Results indicated that the complementary line showed similar phenotype to wild type but the overexpression line showed lowered dormancy compared with wild type (Fig. 6). In addition, the overexpression line also showed a stronger tolerance to ABA during germination (data not shown).
The CYP707A2 gene was involved in NO-induced dormancy break
How the ABA catabolism process was associated with the NO-induced dormancy break is shown in Fig. 7. We used the ABA catabolism mutants, cyp707a1, cyp707a2 and cyp707a3 and checked the effect of NO on their seed dormancy break. Nitric oxide effectively broke the dormancy of freshly harvested wild type, cyp707a1 and cyp707a3 seeds. With 200 µm SNP treatment, > 93% germination rate was achieved after 7 d of imbibition, and c-PTIO could override the effect of SNP completely (Fig. 7a,c). The dormancy of wild type, cyp707a1 and cyp707a3 seeds was also enhanced by c-PTIO, reducing their germination rates to < 10%. In comparison, SNP treatment did not increase the germination rate of the cyp707a2 mutant substantially, with an increase from only 8% to 16% (Fig. 7a). These results show that the function of NO in breaking seed dormancy and in germination relied heavily on the presence and function of CYP707A2.
In order to find out how the gene products of CYP707A2 are related to the effect of NO, we first used QRT-PCR to determine the changes of CYP707A2 transcript during seed germination. Freshly harvested seeds of the wild type were imbibed by water, SNP and c-PTIO solutions (Fig. 7b). For the control treated with water, the transcript of CYP707A2 increased after 1 h, reached a peak after 6 h, before decreasing after 12 h, and then showed sign of slight increase up to 48 h. If imbibed with SNP, the amount of CYP707A2 transcript increased much more substantially and stayed high during the entire process. However, CYP707A2 transcription did not increase with c-PTIO treatment or SNP added with c-PTIO. As an additional test, a second NO donor, SNAP at 100 µm was used in the experiment on CYP707A2 transcription. We found that SNAP also enhanced CYP707A2 transcription and its function could also be reversed by c-PTIO (data not shown).
The CYP707A2 transcript levels in cyp707a1 and cyp707a3 mutants exhibited a similar change to that of the wild type when treated with SNP or c-PTIO during imbibition (Fig. 7d).
The effect of NO on seed dormancy break also needed the participation of ABA 8′-hydroxylase. Western-blot analysis of ABA 8′-hydroxylase, encoded by CYP707A2, was performed during imbibition (Fig. 8). Results indicate that the concentrations of the ABA 8′-hydroxylase protein increased after the first 6 h before decreasing after 12 h (Fig. 8a). The SNP treatments enhanced 8′-hydroxylase protein significantly, while c-PTIO decreased the concentrations of the 8′-hydroxylase protein (Fig. 8b,c), confirming the results for the transcription analysis of CYP707A2 shown in the earlier figure.
NO enhanced the tolerance to ABA in wild type, cyp707a1 and cyp707a3 mutants but not in cyp707a2 mutant during germination
Similar to results reported in earlier studies (Steber et al., 2001), ABA started to inhibit seed germination of wild type from 0.5 µm with the inhibition becoming almost complete by 2 µm (Fig. 9a). Nitric oxide effectively enhanced tolerance to ABA. With 200 µm SNP treatment, the germination rate of wild-type seeds was maintained at 76% under 1 µm ABA and at 28% under 2 µm ABA. By contrast, c-PTIO increased the sensitivity of wild-type seeds to ABA treatment. When treated with 200 µm c-PTIO, only 10% of the wild-type seeds germinated under 1 µM ABA, and no seeds germinated under 2 µm ABA (Fig. 9a). Mutants cyp707a1 and cyp707a3 showed similar results (Fig. 9b,c), with the cyp707a2 mutant being more sensitive to ABA treatment than the wild type. Under 0.5 µm ABA treatment, the germination rate of cyp707a2 decreased to 38% and under 1 µm ABA < 10% germinated. The SNP treatment did not enhance the tolerance of ABA with cyp707a2 seeds significantly during germination (Fig. 9d).
Rapidly accumulated NO at the early stage of imbibition was required for seed dormancy break
The above results showed that NO increased rapidly at first 3 h of imbibition and decreased to a lower level after 6 h, Correspondingly, transcription of CYP707A2 increased at the first 6 h and rapidly decreased to a lower level after 12 h (Figs 2a and 7b). So I want to know if rapidly accumulate NO is essential for the rapidly decreasing ABA and breaking seed dormancy. As shown in Fig. 10A, when imbibed in SNP for more than 3 h, the dormancy of seeds was broken entirely. We also found when imbibed in c-PTIO for more than 3 h seeds dormancy were enhanced. Then we used c-PTIO to scavenge NO for each 3 h in the first 24 h imbibition, results indicated that scavenged NO for 3 h at first 6 h imbibition enhanced the seed dormancy significantly, while scavenged NO for 3 h after 18 h only affected seed dormancy slightly. The results of ABA contents also indicated that scavenged NO for 3 h at first 9 h decreased ABA catabolism significantly but only slightly after 18h (Fig. 10b and c).
A crucial role of NO in dormancy break or germination has been demonstrated in some plants such as Arabidopsis (Batak et al., 2002; Bethke et al., 2004, 2006a,b), barley (Bethke et al., 2004) and lettuce (Beligni & Lamattina, 2000). Earlier results show that SNP, the NO donor, breaks seed dormancy, while c-PTIO, the NO scavenger, enhances it. Our results confirm this observation and show, in addition, that NO takes part in mediating ABA catabolism, which is crucial in decreasing ABA levels and breaking dormancy. We have also demonstrated that ABA 8′-hydroxylase encoded by CYP707A2 is primarily involved in NO-mediated ABA catabolism and breaking seed dormancy in Arabidopsis.
It is well known that the CYP707A family encoding ABA 8′-hydroxylase regulates ABA catabolism (Kushiro et al., 2004; Nambara & Marion-Poll, 2005; Okamoto et al., 2006). We observed that CYP707A2 transcription increased rapidly at the first stage of imbibition and led to a correspondingly rapid ABA decrease in the seeds of Arabidopsis (Fig. 3). When three mutants of the family were used, the cyp707a2 mutant showed stronger dormancy in freshly harvest seeds and needed longer after-ripen time to break dormancy than the cyp707a1 and cyp707a3 mutants that showed a similar phenotype to wild type (Fig. 4). When the ABA changes during imbibition were followed, the ABA levels in the cyp707a2 mutant decreased at a slower rate during the early stage and stayed at higher levels after 48 h imbibition than those in wild type, cyp707a1 and cyp707a3 mutants (Fig. 5). These results indicate that CYP707A2 might play a crucial role in the rapid decrease of ABA during the early stage of imbibition. Results also showed that the CYP707A2 gene and the 8′-hydroxylase protein encoded by CYP707A2 rapidly increased in the first 6 h of imbibition and decreased after 12 h (Figs 3b, 7b, 8a).
The mechanism of NO affecting seed dormancy is not clear so far. In this study we found that NO was involved in seed dormancy and germination by regulating the ABA catabolism. Our evidence came from NO measurements as well as the use of chemicals that can manipulate the NO concentrations during imbibition. Mutants of ABA catabolism, as well as mutants of NO production, all proved the same conclusion. Results indicated that NO rapidly accumulated in the first 3 h and rapid release of NO preceded the transcription of CYP707A2 and ABA 8′-hydroxylase expression encoded by CYP707A2 during imbibition (Figs 2, 7, 8). As a consequence, ABA concentrations decreased and led to the breaking of dormancy (Figs 5, 7). Exogenous NO increased germination of wild type, cyp707a1 and cyp707a3 mutants significantly but not the cyp707a2 mutant (Figs 1, 7). As shown in Fig. 5, 200 mm SNP broke dormancy of freshly harvested seeds of wild type, cyp707a1 and cyp707a3, although the cyp707a1 and cyp707a3 mutants had similar higher ABA level when compared with the cyp707a2 mutant (Figs 5, 7). However, 200 mm SNP did not affect the germination of cyp707a2 mutant with freshly harvested seeds (Fig. 7a). We also found that cyp707a1 and cyp707a3 mutants had higher transcription of CYP707A2 when compared with the wild type (Fig. 7). The speed of ABA catabolism was also enhanced by SNP and inhibited by c-PTIO substantially in WT, cyp707a1 and cyp707a3 mutants, while SNP did not enhance the speed of ABA catabolism in cyp707a2 mutant (Fig. 5). Sodium nitroprusside enhanced tolerance of wild-type, cyp707a1 and cyp707a3 mutant seeds to ABA but not cyp707a2 (Fig. 9).
The role of CYP707A2 gene in ABA catabolism and seed dormancy was also confirmed by the complementary line of cyp707a2 mutants which showed similar phenotype to wild type. The seeds of the overexpression line of CYP707A2 showed lower dormancy when compared with wild type (Fig. 6). Overexpression of CYP707A2 also showed stronger tolerance to ABA during germination (data not shown). These results indicated that the CYP707A2 gene plays a major role in ABA catabolism during imbibition and the role of NO in breaking seed dormancy needs expression of this gene and the action of ABA 8′-hydroxylase, encoded by CYP707A2.
Nitric oxide concentration rapidly increased at the first 3 h of imbibition, reached its maximum at c. 3 h, and then decreased. These results could also be used to explain how scavenging NO at the first stage of imbibition decreased ABA catabolism and after 12 h imbibition scavenging NO affected seed dormancy and ABA catabolism slightly (Figs 2, 10). The peak of NO released appeared at the first stage of imbibition, and if the peak was scavenged by c-PTIO, the rapid increase of transcription of CYP707A2 and the 8′-hydroxylase protein encoded by CYP707A2 disappeared (Figs 7, 8). We found that transcription of ABA catabolism genes increased rapidly in the first 6 h of imbibition and decreased after 6 h, while transcription of biosynthesis genes decreased in the first 6 h and increased after 6 h (Fig. 3). There was no obvious change in ABA concentration after 12 h imbibition, although the concentration was still higher in some treatments or mutants (Figs 2, 5). These results showed that ABA catabolism and biosynthesis may have reached a balance after 12 h of imbibition. If ABA could not be rapidly decreased at the first stage of imbibition, it would stay at a higher concentration, which would maintain dormancy. Therefore the NO-induced rapid decrease in ABA is required for breaking seed dormancy.
It should be noted that although our data for NO treatments suggest that NO is required for the subsequent hormonal regulation and breaking of dormancy, the endogenous NO should be carefully measured in both dormancy and nondormancy seeds before a convincing relationship between NO production and breaking of dormancy can be established. Since NO scavenger PTIO did not affect the germination of nondormancy seeds but strongly inhibited the germination of dormancy seeds (Fig. 1), it is not sure whether NO release and its control mechanism are very different between the dormancy and nondormancy seeds. Further study with accurate NO detection is needed to clarify this.
The site of NO release during imbibition was reported by Sarath et al. (2007) as being largely in the aleurone layer in warm-season C4 grasses. Our fluorescent images suggested that NO is concentrated in the endosperm layer (or what is sometimes referred to as the aleurone layer) that surrounds the embryo and cotyledons (Fig. 2). Aleurone layer has been suggested to be the primary determinant of seed dormancy (Bethke et al., 2007). In some dicotyledonous plants it responds to NO, GA and ABA, and is sufficient to control the embryo seed dormancy if the seed coat-controlled dormancy can be considered separately (Bethke et al., 2007).
In summary, our results demonstrated that CYP707A2 and ABA 8′-hydroxylase encoded by CYP707A2 plays a central role in ABA catabolism during imbibition in Arabidopsis. Nitric oxide acts in regulating the transcription of CYP707A2. Our results also demonstrated that rapidly accumulated NO at the first stage of imbibition is required for rapid ABA catabolism and breaking of seed dormancy. Although we have demonstrated that NO is involved in seed dormancy and germination control, how NO rapidly accumulates at this stage is still unclear. Further study is required to elucidate this process.
This work was supported by Hong Kong Research Grants Council (HKBU262307) and University Grants Committee of Hong Kong (AoE/B-07/99).