BMC Plant Biology

Background: Two complete genome sequences are available for Vitis vinifera Pinot noir. Based on the sequence and gene predictions produced by the IASMA, we performed an in silico detection of putative microRNA genes and of their targets, and collected the most reliable microRNA predictions in a web database. The application is available at http://www.itb.cnr.it/ptp/grapemirna/. Description: The program FindMiRNA was used to detect putative microRNA genes in the grape genome. A very high number of predictions was retrieved, calling for validation. Nine parameters were calculated and,


Background
Chlorophyll molecules are major components of photosynthesis and play essential roles in harvesting light energy and charge separation. Chlorophylls are actively synthesized during greening and are assembled with various apoproteins to form chlorophyll-protein complexes. Apoproteins of core antenna complexes solely bind chlorophyll a, while apoproteins of peripheral antenna com-plexes bind both chlorophyll a and chlorophyll b [1]. Since these chlorophyll-protein complexes are stoichiometrically assembled to form functional photosystems, chlorophyll metabolism must be tightly regulated to coordinate with the formation of these complexes. Previous reports have suggested that the regulation of chlorophyll metabolisms is important from the viewpoint of photodamage. Most of the chlorophyll intermediates are photoreactive compounds [2]. If the regulation of 5aminolevulinic acid (ALA), a precursor of tetrapyrrole synthesis, is defective as the result of mutation of a regulation factor, protochlorophyllide is accumulated in excessive amounts under dark conditions [3]. Consequently, this overabundance of protochlorophyllide results in the generation of reactive oxygen species (ROS) and finally causes growth retardation or cell death during greening [3].
At the last step of chlorophyll biosynthesis, chlorophyll b is synthesized from chlorophyll a by chlorophyllide a oxygenase (CAO) [4]. Since an increase in CAO mRNA levels stimulates chlorophyll b synthesis, CAO activity is considered to be partly regulated by CAO gene expression [5][6][7]. In addition to the regulation on the transcriptional level, chlorophyll b synthesis is also regulated by protein stability. CAO consists of three domains, which are named sequentially from the N terminus as the A, B and C domains [8]. The C domain contains binding motifs for a Rieske center and non-heme iron and catalyzes the conversion of chlorophyll a to chlorophyll b [4,8,9]. Although the precise function of the B domain is not known, it is considered to possibly function as a linker between the A and C domains [10]. Although the A domain is not involved in catalytic function, it regulates CAO protein stability [8,11]. When the A domain was removed from CAO, and only BC domains were introduced and overexpressed in Arabidopsis (Arabidopsis thaliana), the protein levels of CAO were drastically increased and chlorophyll b was excessively accumulated [11]. We recently determined that the chloroplast Clp protease plays an important role in CAO protein stability [12]. However, excess chlorophyll b accumulation does not result in a distinct alteration of photosynthetic activity and does not induce photodamage in green leaves [13]. Therefore, at the present time it is still not known why the protein level of CAO is strictly regulated by the A domain.
In order to answer this question, we introduced genes corresponding to either the complete CAO sequence or BC domains into Arabidopsis. We subsequently investigated the greening processes of these transgenic plants under various light conditions. Interestingly, transgenic plants expressing only the BC domains died when the etiolated seedlings were exposed to continuous light. In addition, chlorophyll synthesis was disturbed in these plants during etiolation and greening of seedlings. We discuss the role of the A domain in the regulation of chlorophyll synthesis and chloroplast development.

CAO without the A domain caused photodamage during greening
In order to elucidate the physiological role of the A domain, we examined the light sensitivity of two types of transgenic plants that we named tGBCch and tGABCch in the ch1-1 mutant background. The ch1-1 mutant contains a deletion in the CAO gene and does not contain chlorophyll b [14]. The tGBCch lines overexpress a chimeric transgene containing the GFP gene and the CAO gene lacking the A-domain coding sequence [11]. In contrast, the tGABCch lines overexpress another chimeric transgene encoding GFP and the full-length CAO protein [11]. Both types of transgenic plants were grown under different light conditions and their phenotypes were compared to those of wild type and the ch1-1 mutant. When the seeds were germinated and grown under low light conditions, all four strains grew well, although the ch1-1 exhibited a pale green phenotype (Fig. 1A). When etiolated seedlings were transferred to low light conditions, most of the tGBCch plants were bleached and the growth of the tGBCch seedlings was retarded (Fig. 1C). When exposed to high light conditions, no etiolated and green seedlings of tGBCch survived ( Fig. 1B and 1D). On the contrary, seedlings from the other three strains developed normally and grew under low and high light conditions (Fig. 1B, C and 1D). These results confirmed that the overexpression of the CAO gene lacking the A domain resulted in photodamage to Arabidopsis seedlings.
GFP-CAO fusion transgenes were highly overexpressed in the etiolated and greening seedlings of all transgenic tGA-BCch and tGBCch plant lines ( Fig. 2A). The expression levels of the transgenes, which correspond to GFP-ABC or GFP-BC, were higher than that of full-length CAO in wild type seedlings ( Fig. 2A). The tGBCch transgenic lines accumulated a large amount of GFP-BC protein (a fusion of GFP and B-C domains of CAO) in both etiolated and green seedlings ( Fig. 2B and 2C). The GFP-BC proteins localized to etioplasts and chloroplasts (Fig. 3). In contrast, the full-length CAO (A-B-C domains) and the GFPfull-length CAO (a fusion of GFP and A-B-C domains of CAO) proteins were under a detectable level by immunoblotting in both etiolated and greening seedlings (Fig.  2B). It was evident that the full-length CAO and the GFPfull-length CAO proteins accumulated to certain levels in the green leaves because chlorophyll b was actively synthesized in the transgenic plants [11]. However, the protein levels of CAO in the seedlings were too low for detection by immunoblot analysis. These data suggest that the low accumulation level of CAO is sufficient to Light sensitivities of the Arabidopsis seedlings Figure 1 Light sensitivities of the Arabidopsis seedlings. A and B: Arabidopsis seedlings were germinated and grown continuously under low (70 μE m -2 s -1 ) or high light (300 μE m -2 s -1 ) conditions for 7 days. C and D: Etiolated seedlings were grown for 4 days and transferred to low or high light conditions for an additional 3 days. Numbers after transgenic plants (for example, tGBCch1, 2, 3) indicate independent transgenic lines for each type of transgenic plants.
>Expression levels of CAO in the etiolated and greening seedlings Figure 2 Expression levels of CAO in the etiolated and greening seedlings. A: The expression levels of internal CAO gene and GFP-CAO transgenes in 4-day-old etiolated seedlings and greening seedlings were determined by semi-quantitative PCR. Expression levels from the tubulin (TUB2) gene were monitored as a quantitative control. The internal CAO expression level increased in wild type and ch1-1 seedlings during greening. The CAO transcript of ch1-1 has a 31 bp deletion and it was not detected in the tGABCch and tGBCch plants. B: The accumulation levels of the CAO proteins and GFP-CAO fusion proteins in 4-day-old etiolated seedlings and in greening seedlings were determined. The estimated molecular size of the full-length CAO, GFP-ABC and GFP-BC proteins without transit peptide were 56, 83 and 70 kDa, C: The accumulation levels of the GFP-BC proteins in 4day-old etiolated tGBCch seedlings during a 24 h greening period were determined. The full-length CAO or GFP-CAO fusion proteins were not detected in wild type, ch1-1 and tGABCch seedlings (data not shown).
synthesize chlorophyll b. We should note that the excess accumulation of the fusion protein in the GFP-BC overexpressing plants was not due to the transcriptional or translational activation in these plants [11]. It was most likely that the lack of the A domain impaired the degradation of CAO by Clp protease [12].
As shown in Figs. 1 and 2, our transgenic lines did not show any obvious phenotypical differences between the lines. In the subsequent experiments described within this study, we also examined all transgenic lines shown in Figs. 1 and 2. All of the lines produced results that were nearly identical.
Next, we examined the accumulation of chlorophyll when etiolated seedlings were transferred to continuous light conditions. In comparison to wild type, chlorophyll levels were low in tGBCch and ch1-1 but slightly higher in tGA-BCch (Fig. 4). In good accordance with chlorophyll accumulation, Lhcb protein levels were higher in tGABCch and lower in tGBCch compared to wild type (Fig. 5). The chlorophyll a/b ratio was very high soon after the onset of illumination and decreased gradually during the greening in wild type (Fig. 6). In contrast, the chlorophyll a/b ratio was already low after illumination and gradually increased during greening in tGABCch and tGBCch (Fig.  6). In tGBCch, the chlorophyll a/b ratio was extremely low. Fig. 7 shows the HPLC profiles of chlorophyll that accumulated after 5 min of illumination in wild type and two types of transgenic plants. Under this condition, most of the chlorophyll molecules were derived from protochlorophyllide a that pre-accumulated during the dark condition. The wild type seedlings accumulated chlorophyll a but not chlorophyll b (Fig. 7). It should be noted that the small peak at 25 min in wild type was not chlorophyll b because the spectrum showed different peaks (data not shown). In contrast, the tGBCch only accumulated chlorophyll b, suggesting that chlorophyll a (or chlorophyllide a) in tGBCch was immediately converted to chlorophyll b (or chlorophyllide b) after the photoconversion of protochlorophyllide a into chlorophyllide a (Fig. 7). Both chlorophyll a and chlorophyll b accumulated in tGA-BCch and the level of chlorophyll a was nearly equivalent to the levels of wild type (Fig. 7).
Interestingly, we found that the fully-greened leaves of tGBCch accumulated a significant amount of 7hydroxymethyl chlorophyll, which is an intermediate molecule in the conversion between chlorophyll b and chlorophyll a (Fig. 8). The unusually high level of 7-Accumulation of chlorophyll and the GFP-BC protein in the tGBCch seedlings during greening Figure 3 Accumulation of chlorophyll and the GFP-BC protein in the tGBCch seedlings during greening. The fluorescence images of the Arabidopsis cotyledons during greening for 24 h were observed with confocal microscopy. The GFP-CAO protein and the plastids during greening are indicated by green and red fluorescence, respectively. The green fluorescence of GFP-BC protein was detected only within plastids during greening.
hydroxymethyl chlorophyll suggests that the excess GFP-BC proteins or the excess chlorophyll b might activate the conversion between chlorophyll a and b [15].

Loss of the A domain results in a perturbation of chlorophyll metabolism
The tGBCch seedlings accumulated a low level of chlorophyll during the greening process (Fig. 4). These data are consistent with the observation that the amount of protochlorophyllide, which is an end product of chlorophyll synthesis in etiolated seedlings, decreased in tGBCch transgenic plants (Table 1). When plants were fed with ALA, similar trends were observed between wild type and the transgenic plants (Table 1). These results might suggest that certain enzymatic steps between ALA and protochlorophyllide formation were disturbed and that this disturbance limited the chlorophyll synthesis in the tGB-Cch seedlings during greening.
In our next line of investigation, we examined the re-accumulation of protochlorophyllide in the dark after a brief illumination as a means to analyze the kinetics of protochlorophyllide synthesis in the seedlings. Under these conditions, protochlorophyllide levels were similar in wild type, tGBCch and tGABCch during the first day of darkness (Table 2). However, the levels decreased in the following 2-day period of darkness only in tGBCch ( Table  2). Although the etiolated tGBCch seedlings seemed to synthesize protochlorophyllide normally, the pigment appeared to be broken down during dark condition by unknown mechanisms. Alternatively, It is also possible that divinyl-protochlorophyllide a may inhibit an earlier step of chlorophyll biosynthesis in a feedback manner.
In order to investigate the chlorophyll b synthetic pathway in tGBCch, we compared the compositions of the chlorophyll intermediates in wild type, tGBCch and tGABCch. In spite of the accumulation of CAO in tGBCch, protochlorophyllide b was not detected in etiolated seedlings (Fig.  9). These data indicate that CAO is unable to catalyze the conversion of protochlorophyllide a to protochlorophyllide b in Arabidopsis. Interestingly, a significant difference was observed in the accumulation ratios of divinyl-protochlorophyllide a to total protochlorophyllide a among wild type, tGBCch and tGABCch ( Fig. 9 and Table 1). The etiolated seedlings of wild type and tGABCch predominately accumulated monovinyl-protochlorophyllide a (Table 1). Conversely, monovinyl-protochlorophyllide a was accumulated to very low levels in tGBCch (Table 1). This observation indicates that the loss of the A domain from CAO resulted in an inactivation of 3, 8-divinyl protochlorophyllide a 8-vinyl reductase [16].

The A domain protects cells against photodamage during greening
We previously reported on the light sensitivity of transgenic Arabidopsis plants overexpressing prokaryotic CAO Lhcb protein levels in the greening seedlings Figure 5 Lhcb protein levels in the greening seedlings. Etiolated seedlings were first grown for 4 days under dark conditions and were subsequently exposed to low light conditions for 24 h. The accumulation levels of the Lhcb proteins in the greening seedling were determined by immunoblotting using anti-Lhcb antibody. Lhcb protein levels in the ch1-1 seedlings after 24 h of greening were nearly equivalent to those from tGBCch (data not shown). Chlorophyll levels in the seedlings during greening Figure 4 Chlorophyll levels in the seedlings during greening. Etiolated seedlings were grown for 4 days under dark conditions. The etiolated seedlings were then exposed to low light conditions and were harvested during greening at the indicated times. The total amount of chlorophyll per seedling was determined with HPLC analysis. The error bars represent standard deviations (n = 4).
which does not have the A domain. These transgenic plants accumulated much more chlorophyll b than tGB-Cch plants [13]. In this previous report, we concluded that the excess accumulation of chlorophyll b did not cause photodamage in the fully greened leaves of the transgenic plants [13]. However, in this study, we found that most of the etiolated tGBCch seedlings died during greening not only under high light conditions, but also under low light conditions. Without the presence of chlorophyll a, chlorophyll b would not be incorporated into apoproteins. As a consequence, the free chlorophyll b might generate ROS [2]. In contrast to tGBCch, transgenic plants overexpressing the full-length CAO accumulated both chlorophyll a and chlorophyll b immediately after the onset of illumina-tion. Unlike tGBCch plants, the full-length-CAO overexpression lines did not exhibit photodamage. It is possible that chlorophyll b is assembled with apoproteins in the presence of chlorophyll a. As a result, the generation of ROS might have been suppressed. Collectively, these results indicate that the A domain is essential in preventing photodamages that could be caused by excess chlorophyll b accumulation during greening. The etiolated seedlings of the transgenic Arabidopsis overexpressing prokaryotic CAO might cause obvious photodamage.
An unusually high level of 7-hydroxymethyl chlorophyll a, an intermediate molecule in the conversion between chlorophyll b to chlorophyll a, accumulated in the green leaves of tGBCch (Fig. 8). This observation might indicate that excess chlorophyll b is reconverted to chlorophyll a by the chlorophyll cycle [15]. It is not likely that 7hydroxymethyl chlorophyll was formed in the course of the chlorophyll a to chlorophyll b conversion. Since CAO catalyzes the two-step reaction from chlorophyll a to chlorophyll b by itself, the intermediate molecule, 7hydroxymethyl chlorophyll, is supposed to be associated with the CAO enzyme and is not released during the reaction [14]. In our study, 7-hydroxymethyl chlorophyll was not found just after the start of the illumination of etiolated seedlings, although chlorophyll b was actively synthesized (Fig. 7). Therefore, it would be reasonable to assume that 7-hydroxymethyl chlorophyll was predominantly formed in the course of the chlorophyll b to chlorophyll a conversion.

Accumulation of GFP-BC protein interferes with the regulation of chlorophyll biosynthesis
In comparison to wild type plants, chlorophyll accumulation during greening was suppressed in tGBCch seedlings, even under low light conditions. It is possible that this suppression might be due to the oxidative stress generated by free chlorophyll b. The overabundance of protochlorophyllide generates ROS, which causes growth retardation or cell death during greening in the flu mutant [3]. Recently, we found that the chlorophyll biosynthesis pathway is a target of oxidative stress [17]. However, all of the phenotypes of tGBCch could not be explained by oxi- The seedlings were grown for 6 days under dark conditions with or without ALA (0.5 mg ml -1 ) on wet filter paper. Seedlings were subsequently harvested under dark conditions and their extracts were analyzed for their protochlorophyllide composition by HPLC analysis (n = 5).
Molar ratio of the chlorophyll a and b contents in the seed-lings during greening Figure 6 Molar ratio of the chlorophyll a and b contents in the seedlings during greening. The etiolated seedlings were first grown for 4 days under dark conditions and were subsequently exposed to low light. The compositions of chlorophyll a and b were calculated from chromatographic peak profiles at 650 nm with HPLC analysis. The error bars represent standard deviations (n = 4).
dative stress alone. For example, the level of protochlorophyllide a in etiolated tissues of tGBCch was lower than that of wild type. Since these plants were not exposed to light, the involvement of ROS in the inhibition of protochlorophyllide accumulation can be excluded. We also observed an increase in the ratio of divinyl-protochlorophyllide to the total protochlorophyllide in tGBCch transgenic plants. These data indicate that the accumulation of the BC domains of CAO inhibits the activity of 3, 8-divinyl protochlorophyllide a 8-vinyl reductase. On the contrary, overexpression of the GFP-ABC protein enhanced chlorophyll accumulation. These phenomena are very interesting in terms of biological regulation since both CAO and truncated CAO affected not only chlorophyll b synthesis but also other chlorophyll metabolic steps.
These results are reasonable because some enzymes, which are related to chlorophyll biosynthesis, are reported to form complexes or interact with other enzymes. For example, glutamyl-tRNA reductase and glutamate-1-semialdehyde aminotransferase form a complex which enables an efficient substrate trafficking between two enzymes [18]. In addition, BchH, one of the three subunits of Mg chelatase of Rhodobacter capsulatus, accelerated the Mg-protoporphyrin IX monomethyl transferase activity [19]. If CAO does indeed interact with other enzymes, the accumulation of truncated CAO could potentially interfere with the activity of the target enzymes.
In addition, the loss of the A domain in CAO impaired the conversion of divinyl-protochlorophyllide a to monovinyl protochlorophyllide a, and obviously reduced total protochlorophyllide a level under dark conditions. A possible scenario for the reduction of protochlorophyllide accumulation in the tGBCch seedlings is that the excessively-accumulated CAO in these seedlings interfered with the enzyme that catalyzes the formation of monovinylprotochlorophyllide a from divinyl-protochlorophyllide a, the latter of which may inhibit an earlier step of chlorophyll biosynthesis in a feedback manner. Further studies are necessary and warranted to clarify this question.

CAO does not catalyze the conversion of protochlorophyllide a to protochlorophyllide b
When etiolated seedlings of wild type are exposed to continuous light, chlorophyll a immediately accumulates after the onset of illumination. Subsequent to illumination, chlorophyll a and b levels gradually increase after several hours of a lag phase. Conversely, in tGBCch plants, chlorophyll b accumulated but chlorophyll a was not detected after 5 min of illumination (Fig. 7). Considering that protochlorophyllide b was not found in etiolated seedlings, these data indicate that chlorophyll a was immediately converted to chlorophyll b in tGBCch seedlings during the early phase of greening (Fig. 9).
Several research groups report that protochlorophyllide b does not exist in etioplasts [20,21]. Contrary to these reports, another group claimed that the protochlorophyllide b could be detected in etiolated seedlings when they use diethylpyrrocarbonate-containing acetone for the extraction of pigments [22]. It was also proposed that protochlorophyllide b is synthesized via a pathway that is independent from the CAO protein [23]. In addition, a cyanobacteria transformant with the CAO and Lhcb transgenes, accumulated protochlorophyllide b [24]. This observation indicates that CAO is capable of converting protochlorophyllide a to protochlorophyllide b. In this study, we used 100% acetone as an extraction solvent for chlorophyll intermediates in order to avoid the artificial Chlorophyll accumulation in etiolated Arabidopsis seedlings after light treatments Figure 7 Chlorophyll accumulation in etiolated Arabidopsis seedlings after light treatments. Chlorophyll accumulation after exposure to low light for 5 min in 6-day-old etiolated seedlings. The chlorophyll a and b extracted from the seedlings were separated by HPLC analysis. Chl a, chlorophyll a; Chl b, chlorophyll b.
Chlorophyll accumulation in greened seedlings Figure 8 Chlorophyll accumulation in greened seedlings. Chlorophyll accumulation in fully greened seedlings that were grown for 8 days under low light conditions. The chlorophyll a, b and 7-hydroxymethyl chlorophyll extracted from the seedlings were separated by HPLC analysis. The spectrum of a peak that was eluted at 24.6 min in the extract from the tGBCch leaves is shown. Chl a, chlorophyll a; Chl b, chlorophyll b; 7-HM-Chl, 7-hydroxymethyl chlorophyll.
formation of chlorophyll derivatives and to maximize our extraction of the chlorophyll intermediates from seedlings [21]. Consequently, we did not detect any protochloro-phyllide b in etiolated tGBCch seedlings, even though a large amount of the CAO protein accumulated. These results indicate that protochlorophyllide b is not a major intermediate product for chlorophyll b synthesis in planta.

Conclusion
Taken together, our results demonstrated that excessive accumulation of chlorophyll b is harmful to plants. This deleterious effect likely disturbs the proper synthesis of chlorophyll intermediates and the accumulation of photosynthetic proteins. Therefore, we conclude that the regulation of the CAO protein by the A domain is essential for the survival of etiolated seedlings under light conditions. Since the A domain functions to maintain proper chlorophyll b levels throughout the development of plants [11], it would be reasonable to assume that the regulation of the CAO protein level is essential in various aspects of plant development and acclimation to light conditions. Further studies on the regulatory mechanism Six-day-old etiolated seedlings were exposed to low light for 5 min and were returned to dark conditions. The seedlings after a 1 or 3 day additional dark incubation period were subsequently harvested under dark conditions. Their extracts were analyzed for their protochlorophyllide composition by HPLC analysis (n = 3). The n.d. abbreviation indicates "not detected".
Chromatographic peaks and spectra of divinyl-and monovinyl-protochlorophyllide a in etiolated seedlings Figure 9 Chromatographic peaks and spectra of divinyl-and monovinyl-protochlorophyllide a in etiolated seedlings. Seedlings were grown for 6 days under dark conditions on wet filter paper without ALA. The seedlings were harvested under dark conditions and their extracts were analyzed for their protochlorophyllide composition by HPLC. DV, divinyl-protochlorophyllide a; MV, monovinyl-protochlorophyllide a.
of CAO may elucidate the physiological roles of this mechanism throughout the life cycle of plants.

Plant materials and growth conditions
The Arabidopsis CAO protein consists of three domains, which are respectively named: A (V37-L170), B (P171-G200) and C (A201-G537) domains [8]. In this study, we used Arabidopsis thaliana Columbia wild type and the ch1-1 mutant which contains a deletion in the CAO gene [14].
In addition, we also used two types of transgenic plants overexpressing chimeric fusions of the CAO protein and the GFP. The two Arabidopsis transformants, tGBCch and tGABCch, expressed fusion transgenes corresponding to transit peptide-GFP-B-C domains and transit peptide-GFP-A-B-C domains in the ch1-1 background, respectively [11]. Three homozygous lines for each transgenic plant tGBCch and tGABCch, were used in this study. All of the transgenic plants were isolated independently. Arabidopsis seedlings were grown at 22°C on agar plates (0.7% [w/ v]) containing 1/2 diluted Murashige-Skoog medium. Plants were grown under low light (70 μE m -2 s -1 from fluorescent bulbs) or high light (300 μE m -2 s -1 from xenon light bulbs [3.6 kW] equipped with ND filters, TGE-2S3, Tabai, Osaka, Japan) conditions.

Reverse transcriptase PCR
Total RNA was extracted using an RNeasy Plant Mini Kit (Qiagen, Hilden, Germany) from 200 etiolated seedlings which were grown under dark conditions for 4 days. RNA was extracted from etiolated seedlings after a period of greening for 24 hours. One μg of isolated RNA was reverse-transcribed into cDNA using a SuperScript III Kit (Invitrogen, Carlsbad, CA). PCR was performed to determine gene expression level using the synthesized firststranded cDNA sample in a 25 μl reaction. PCR conditions were as follows: 96°C (2 min (25 μl) were subsequently subjected to SDS-PAGE. The resolved proteins were blotted onto a hybond-P membrane (GE Healthcare, Buckinghamshire, UK). A 1/5000-diluted anti-GFP (Invitrogen) antibody, a 1/2000-diluted anti-CAO antibody [11] and a 1/5000diluted anti-Lhcb [6] rabbit antibody were used to detect specific proteins as previously described [11]. Cross-reactive protein bands were developed using anti-rabbit IgG that was linked to horseradish peroxidase (GE Healthcare). An ECL plus western blotting analysis kit was use for the chemiluminescent detection of antigens (GE Healthcare).