The rice faded green leaf locus encodes protochlorophyllide oxidoreductase B and is essential for chlorophyll synthesis under high light conditions

Authors

  • Yasuhito Sakuraba,

    1. Department of Plant Science, Plant Genomics and Breeding Institute, Research Institute for Agriculture and Life Sciences, Seoul National University, Seoul, Korea
    Search for more papers by this author
    • These authors contributed equally to the work.
  • Md Lutfor Rahman,

    1. Department of Plant Science, Plant Genomics and Breeding Institute, Research Institute for Agriculture and Life Sciences, Seoul National University, Seoul, Korea
    Search for more papers by this author
    • These authors contributed equally to the work.
  • Sung-Hwan Cho,

    1. Department of Plant Science, Plant Genomics and Breeding Institute, Research Institute for Agriculture and Life Sciences, Seoul National University, Seoul, Korea
    Search for more papers by this author
  • Ye-Sol Kim,

    1. Department of Plant Science, Plant Genomics and Breeding Institute, Research Institute for Agriculture and Life Sciences, Seoul National University, Seoul, Korea
    Search for more papers by this author
  • Hee-Jong Koh,

    1. Department of Plant Science, Plant Genomics and Breeding Institute, Research Institute for Agriculture and Life Sciences, Seoul National University, Seoul, Korea
    Search for more papers by this author
  • Soo-Cheul Yoo,

    Corresponding author
    • Department of Plant Science, Plant Genomics and Breeding Institute, Research Institute for Agriculture and Life Sciences, Seoul National University, Seoul, Korea
    Search for more papers by this author
  • Nam-Chon Paek

    Corresponding author
    • Department of Plant Science, Plant Genomics and Breeding Institute, Research Institute for Agriculture and Life Sciences, Seoul National University, Seoul, Korea
    Search for more papers by this author

For correspondence (e-mail ncpaek@snu.ac.kr or suchyoo3@snu.ac.kr).

Summary

NADPH:protochlorophyllide oxidoreductase (POR) catalyzes photoreduction of protochlorophyllide (Pchlide) to chlorophyllide in chlorophyll (Chl) synthesis, and is required for prolamellar body (PLB) formation in etioplasts. Rice faded green leaf (fgl) mutants develop yellow/white leaf variegation and necrotic lesions during leaf elongation in field-grown plants. Map-based cloning revealed that FGL encodes OsPORB, one of two rice POR isoforms. In fgl, etiolated seedlings contained smaller PLBs in etioplasts, and lower levels of total and photoactive Pchlide. Under constant or high light (HL) conditions, newly emerging green leaves rapidly turned yellow and formed lesions. Increased levels of non-photoactive Pchlide, which acts as a photosensitizer, may cause reactive oxygen accumulation and lesion formation. OsPORA expression is repressed by light and OsPORB expression is regulated in a circadian rhythm in short-day conditions. OsPORA was expressed at high levels in developing leaves and decreased dramatically in fully mature leaves, whereas OsPORB expression was relatively constant throughout leaf development, similar to expression patterns of AtPORA and AtPORB in Arabidopsis. However, OsPORB expression is rapidly upregulated by HL treatment, similar to the fluence rate-dependent regulation of AtPORC. This suggests that OsPORB function is equivalent to both AtPORB and AtPORC functions. Our results demonstrate that OsPORB is essential for maintaining light-dependent Chl synthesis throughout leaf development, especially under HL conditions, whereas OsPORA mainly functions in the early stages of leaf development. Developmentally and physiologically distinct roles of monocot OsPORs are discussed by comparing with those of dicot AtPORs.

Introduction

In photosynthetic organisms, chlorophyll (Chl) harvests light energy and transfers the excitation energy to other components of the photosynthetic electron transport chain (Grossman et al., 1995; Barber et al., 2000). In contrast to its essential, energy-harvesting role, Chl and its intermediate derivatives can also interact with oxygen molecules, giving rise to toxic singlet oxygen radicals (op den Camp et al., 2003). Therefore, chloroplast development and maintenance requires the finely tuned control of Chl synthesis. One of the most important steps in Chl synthesis is the reduction of protochlorophyllide a (Pchlide a) to chlorophyllide a (Chlide a), which is the sole light-requiring reaction in the Chl synthetic pathway (Griffiths, 1978). This photoreduction is catalyzed by NADPH:protochlorophyllide oxidoreductases (PORs), which are highly conserved in all the oxygenic photosynthetic organisms (Masuda and Takamiya, 2004). The catalytic capacity of PORs is regulated in a light-dependent manner for the strict control of Chl synthesis (Apel et al., 1980). Whereas PORs reduce Pchlide a to Chlide a upon illumination, the disintegration of prolamellar bodies (PLBs) occurs as etioplasts begin to transform into chloroplasts (Virgin et al., 1963; Henningsen, 1970.

POR genes or PORs have been identified in monocot plants, including Hordeum vulgare (barley; Holtorf et al., 1995; Schulz et al., 1989), Avena sativa (oat; Darrah et al., 1990), Triticum aestivum (wheat; Teakle and Griffiths, 1993) and Zea mays (maize; Hopkins, 1982; Hopkins and Elfman, 1984; Millerd and McWilliam, 1968), and in dicot plants including Arabidopsis thaliana (Arabidopsis; Armstrong et al., 1995; Benli et al., 1991; Oosawa et al., 2000), Pisum sativum (pea; Spano et al., 1992), Nicotiana tabacum (tobacco; Masuda et al., 2002; Zavaleta-Mancera et al., 1999), Cucumis sativus (cucumber; Yoshida et al., 1995) and Amaranthus tricolor (Iwamoto et al., 2001). Among them, biochemical and physiological functions of PORs are well studied in barley and Arabidopsis. In barley, two POR isoforms have been identified: HvPORA and HvPORB (Holtorf et al., 1995). Although their enzymatic activities are quite similar in vitro (light- and NADPH-dependent), their gene expression patterns are distinct: HvPORA transcripts are abundant in etiolated seedlings, but decrease rapidly after illumination (Holtorf et al., 1995). However, HvPORB is expressed constitutively throughout leaf development, not only in dark-grown seedlings but also in light-adapted green tissues (Holtorf et al., 1995). In Arabidopsis, AtPORA and AtPORB were initially identified, and their catalytic functions and expression patterns are considerably similar to those of barley (Armstrong et al., 1995). Thereafter, another isoform, AtPORC, was identified (Oosawa et al., 2000). In contrast with AtPORA, AtPORC expression is fluence rate-dependent, increasing under light conditions and decreasing under dark conditions (Oosawa et al., 2000; Su et al., 2001). Noticeably, AtPORC was drastically upregulated by high light (HL) treatment, whereas both AtPORA and AtPORB were downregulated (Masuda et al., 2003). In the Arabidopsis porC (atporC) mutant, Chl levels decreased drastically under HL treatment (Masuda et al., 2003), and overexpression of AtPORC enhanced tolerance to photo-oxidative damage (Pattanayak and Tripathy, 2011). These results indicate that AtPORC activity is required for light-dependent Chl synthesis and/or maintenance of a threshold Chl level, especially under HL conditions during leaf development (Masuda et al., 2003).

The number of POR isoforms differs among angiosperms: for example, Arabidopsis contains three PORs, barley, rice, and tobacco have two PORs and cucumber and pea harbor only one POR (Sundqvist and Dahlin, 1997; Fusada et al., 2000). In addition, the distinct roles of PORs in each plant species seem quite different, even in the species with the same number of PORs. For example, both barley (monocot) and tobacco (dicot) have two PORs, but their expression patterns in etiolated seedlings are considerably different from each other (Holtorf et al., 1995). Determination of the developmentally and physiologically divergent roles of POR isoforms in detail requires isolation and characterization of por mutants in each plant species, as shown in the studies of the three por mutants in Arabidopsis (Frick et al., 2003; Masuda et al., 2003); however, studies in other plants have been considerably limited by a lack of por mutants.

In this study, we carried out map-based cloning of the faded green leaf (fgl) locus in Oryza sativa (Os, rice), and revealed that fgl harbors a 1–bp deletion in the coding region of OsPORB, resulting in a frameshift mutation and premature translational termination. Based on our expression analysis of two OsPOR genes and phenotypic characterization of the fgl mutant, we proposed that OsPORB plays important roles in maintaining a threshold Chl level throughout leaf development, especially under HL field conditions, whereas OsPORA mainly functions during early leaf elongation in the field. The detailed developmental and physiological roles of two OsPORs in Chl synthesis, depending on light conditions and leaf developmental stages, are discussed.

Results

Phenotypic characteristics of the rice fgl mutant

In the paddy field, fgl plants were easily distinguished by their production of light-green leaf blades during early vegetative stages and then severe degreening (yellow/white variegation), commencing with the leaf-tip area and proceeding throughout development (Figure 1a). To characterize the fgl phenotype in more detail, we examined five leaf blades from top (youngest, I) to bottom (oldest, V) in the main stem at 80 days after seeding (DAS) (Figure 1a,b). Leaf variegation was always initiated from older leaves; this degreening proceeded from the distal region of leaf blades and spread downwards. Afterwards, several necrotic lesions formed in the yellow sectors. Next, we examined photosynthetic pigments and proteins in each leaf tissue. As the fgl leaves became older, the concentrations of total Chl gradually decreased (Figure 1c), and the Chl a/b ratio increased (Figure 1d). The levels of photosynthetic proteins (Lhcb1, PsbC, Lhca1, PsaF and RbcL) also decreased, especially Lhcb1 and Lhca1 (Figure 1e). Lower levels of photosystem proteins cause a decrease in the concentrations of photosystem-binding pigments (Tzvetkova-Chevolleau et al., 2007). Similarly, total carotenoid levels were significantly diminished in fgl, possibly because of decreasing levels of photosystem proteins in older leaves (Figure 1f). This suggests that the Chl synthetic rate and/or photosynthetic apparatus in the chloroplasts of fgl are not properly maintained during leaf maturation, leading to yellow/white leaf variegation. For this reason, the fgl mutation negatively affects the agronomic traits of rice (Figure S1), such as the number of tillers/plant, the number of panicles/plant and 1000–grain weight. The fgl phenotypic characterization suggests that during leaf development, FGL is essential for Chl synthesis and/or stability of photosynthetic proteins in mature chloroplasts.

Figure 1.

Phenotypic characterization of the fgl mutant. (a) Phenotypes of wild-type (WT) and fgl plants at 80 days after seeding (DAS) in the paddy field. (b) Phenotypes of leaf blades in the main culm of the WT and fgl plants in (a). (b–f) I, first (youngest, upper position); II, second; III, third; IV, fourth; and V, fifth (oldest, lower position) leaf blades. (c, d) Total chlorophyll (Chl) contents (c) and Chl a/b ratio (d) according to leaf age (see Experimental procedures). Mean and SD values were obtained with four biological replicates. (e) Immunoblot analysis of photosystem proteins (see Experimental procedures). Total protein extracted from 1.0 mg (fresh weight) leaf tissues was loaded in each lane. α–Lhcb1, anti-Lhcb1antibody used for detection. Rubisco large subunit (RbcL) protein was visualized by Coomassie Brilliant Blue staining. Similar results were obtained in three independent experiments. (f) Total carotenoid (Car) contents according to leaf age. Mean and SD values were obtained from four biological replicates.

Map-based cloning of the fgl locus

The single recessive fgl locus was previously mapped to the long arm of chromosome 10 (Iwata and Omura, 1975). In the current study, genetic mapping was initially performed using 243 fgl plants out of 1032 plants in an F2 mapping population derived from the cross of fgl (japonica) and Dasanbyeo (indica). Using four SSR and eight STS markers, fgl was initially mapped to an 8.5–cM interval (Figure 2a). For physical mapping, 789 additional wild-type F2 plants whose genotypes were determined by F3 test were used, and new STS markers were developed by aligning genomic DNA sequences of japonica and indica types (Figure 2b). Finally, fgl was narrowed down to 47 kb between two STS markers, S210 and fg31, in the BAC clones, AC092489 and AC068923 (Figure 2b). Ten candidate genes (Figure 2c) were found in the 47–kb genomic region from the Rice Genome Research Program (http://rgp.dan.affrc.go.jp), and cloned from fgl by RT-PCR or genomic PCR. We found a one base pair (1–bp) deletion in the second exon of OsPORB (Os10g35370) in fgl; this deletion produces a frameshift mutation and leads to premature translational termination, as the wild-type OsPORB protein has 402 amino acids (aa) but the predicted mutant protein has only 51 aa (Figure 2d and Figure S2). The 1–bp deletion of osporB in fgl was further confirmed by dCAPS analysis (Figure 2e).

Figure 2.

Map-based cloning of the fgl locus. (a) Genetic mapping of the fgl locus. The fgl locus was initially mapped to within 8.5 cM between two STS markers, S10053 and S10061.5, on chromosome 10. The marker information is listed in Table S1. (b) Physical mapping of the fgl locus with STS markers. Numbers below the thick lines represent the number of recombinants between two flanking markers. (c) Candidate genes (black boxes) in the 47–kb genomic DNA. (d) A frameshift mutation of OsPORB in the fgl mutant. Four exons and three introns are indicated by rectangles and thick lines, respectively. The position of the 1–bp deletion (G109) is marked in the second exon. (e) dCAPS analysis was performed to confirm the 1–bp deletion in OsPORB. AatII digests the genomic PCR products from the fgl mutant (lane 1) but not from the wild type ‘Kinmaze’ (lane 2). Lane 3, a mapping parent ‘Dasanbyeo’; lane 4, F1 hybrid (Dasanbyeo/fgl).

To confirm that the mutation in OsPORB caused the fgl phenotype, we used OsPORB to complement the fgl mutation. For the complementation test, the 35S:OsPORB cDNA construct in a pMDC32 binary plasmid (see Experimental procedures) was introduced into calli generated from mature embryos of fgl seeds by Agrobacterium-mediated transformation (Jeon et al., 2000). Eight independent transgenic lines (35S:OsPORB/fgl) grew normally, and their leaf color was indistinguishable from that of wild-type plants (Figure S3), confirming that the loss-of-function mutation in OsPORB is responsible for the leaf variegation of fgl. The rice genome has two POR isoforms: OsPORA and OsPORB. OsPORA (Os04g58200) encodes a 388–aa protein, the sequence of which is considerably similar to OsPORB (Figure S4). Also, the POR proteins are very conserved among higher plants (Figure S4).

PLB formation and photoactive Pchlide levels in the etiolated seedling of fgl

It has been shown that both POR and Pchlide levels are closely associated with the development of prolamellar bodies (PLBs; a uniformly curved lattice of tubular membranes) in etioplasts (Ryberg and Sundqvist, 1991; Engdahl et al., 2001), because the size of PLBs is significantly reduced in the etioplasts of both atporA and atporB mutants in Arabidopsis (Masuda et al., 2003; Paddock et al., 2012). To confirm whether the size reduction of PLBs occurs in the fgl/osporB mutant, we examined the ultrastructure of plastids in 10–day-old seedlings grown in darkness or in the paddy field. In the leaf tissues of etiolated seedlings, PLBs were smaller in the etioplasts of fgl than in the wild type (Figure 3a,b). In field-grown seedlings, the chloroplasts in the light-green sectors of fgl leaves exhibited decreased thylakoid stacking with a few plastglobules (Figure 3d). Furthermore, in the chloroplasts of yellow sectors, thylakoid stacking was hardly detected and plastoglobule number increased significantly (Figure 3e).

Figure 3.

Transmission electron microscopy (TEM) analysis of plastids in fgl leaves. (a, b) Etioplasts in the mesophyll cells of wild type WT (a) and fgl (b). Leaf samples were obtained from the etiolated seedlings grown for 10 days in darkness. (c–e) Chloroplasts in the green leaves of WT (c), and in the light-green (d) and yellow (e) sectors of fgl leaves. Leaf samples were obtained from 1–month-old plants grown in the paddy field. G, grana thylakoid; PG, plastoglobule; PLB, prolamellar body. Scale bars: 0.5 μm.

It was previously reported that the size reduction of PLBs leads to a decrease of total Pchlide levels (Franck et al., 2000). We also found that in etiolated seedlings, the total Pchlide level in the fgl mutant was reduced to about half of the wild-type level (Figure 4), consistent with the reduced size of PLBs in the fgl etioplasts (Figure 3b). Also, the ratio of photoactive Pchlide to total Pchlide was 0.22 in fgl and 0.78 in wild type, indicating that the photoactive Pchlide level in PLBs is remarkably reduced, and that relatively higher levels of non-photoactive Pchlide accumulate in the fgl leaves (Figure 4). These indicate that in rice, OsPORB is essential for both PLB and photoactive Pchlide formation in dark conditions for light-dependent Chl synthesis.

Figure 4.

Levels of total Pchlide and photoactive Pchlide were significantly reduced in fgl. Pchlide levels were measured in (1) the 10–day-old etiolated seedlings for total Pchlide, and (2) 10–min light-treated etiolated seedlings for non-photoactive Pchlide. Photoactive Pchlide levels were obtained by subtraction: (1) – (2). Mean and SD values were obtained from six replications. This analysis was performed twice with similar results.

OsPORB is not required for light-dependent Chl synthesis during greening of etiolated plants

There are three POR isoforms in Arabidopsis (Su et al., 2001) and two POR isoforms in rice (Masuda and Takamiya, 2004). Therefore, we examined the molecular characteristics and distinct functions of the two OsPORs in Chl synthesis, depending on light or developmental stage. First, we examined the effect of light on OsPOR expression during greening. When 6–day-old etiolated wild-type seedlings were exposed to light for 24 h, OsPORA expression was rapidly downregulated after illumination, whereas OsPORB expression decreased only slightly (Figure 5a). Consistent with its mRNA levels, OsPORA protein levels were severely reduced after 6 h of light treatment, and were maintained at low levels until 12 h in both wild-type and fgl seedlings (Figure 5b). OsPORB remained relatively constant in the wild type, but OsPORB did not accumulate in fgl because of the frameshift mutation in OsPORB.

Figure 5.

Abundance of OsPOR transcripts and OsPOR proteins during the greening of etiolated seedlings. The 6–day-old etiolated seedlings of wild type (WT) and fgl were exposed to light (200 μmol m−2 s−1) for 0 h (L0) up to 24 h (L24). (a) Changes in transcript levels of OsPORA and OsPORB in WT seedlings during greening. In this study, qRT-PCR analysis was performed to measure their relative mRNA levels, which were normalized to the transcript levels of Glyceraldehyde phosphate dehydrogenase (GAPDH; GenBank accession number AK064960). (a, d) Mean and SD values were obtained with three biological replicates. (b) Immunoblot analysis of OsPORs during greening (see Experimental procedures). OsPORA and OsPORB proteins were visualized with an anti-POR antibody using an ECL detection system. Anti-RbcL antibody was used to detect the Rubisco large subunit. Total protein extracted from 5.0 mg (fresh weight) leaf tissue was loaded in each lane. Similar results were obtained in three independent experiments. (c) Comparison of greening speed between WT and fgl etiolated seedlings. (d) Chl contents in the WT and fgl seedlings during greening. This analysis was performed twice with similar results.

Next, to examine whether OsPORA contributes to light-dependent Chl synthesis during leaf greening, we compared the greening speed of wild-type and fgl seedlings. The visible speed of greening (Figure 5c) and the rates of Chl synthesis (Figure 5d) were not significantly altered in fgl, indicating that low levels of OsPORA are sufficient for leaf greening in rice, even in the absence of OsPORB activity in fgl mutants.

The rapid yellowing of newly emerging fgl leaves is caused by the rapid downregulation of OsPORA expression under continuous light

Because levels of OsPORA mRNA and OsPORA protein are drastically downregulated upon illumination (Figure 5a), it can be assumed that OsPORA activity would be nearly absent in rice plants grown under continuous light (CL) conditions. To examine this, we grew wild-type and fgl plants for 2 weeks in two photoperiodic conditions: short days (SDs; 10–h light/14–h dark) as the control and CL conditions. The leaves of fgl mutant plants showed almost normal green phenotype with no leaf variegation in SD conditions (Figure 6a), with slightly reduced Chl synthesis (Figure 6b). Under CL conditions, however, fgl mutants displayed a yellowish leaf color, with higher necrotic lesion formation, indicating severe repression of Chl synthesis (Figure 6a,b). We next examined the mRNA and protein levels of OsPORA and OsPORB in 2–week-old wild-type leaves under SD or CL conditions. Expression levels of OsPORA under CL were substantially lower than under SDs, leading to almost undetectable levels of OsPORA under CL (Figure 6c,e), whereas the mRNA and protein levels of OsPORB were slightly reduced (Figure 6d,e). This indicates that the severe leaf variegation of fgl under CL conditions is mainly attributable to the rapid reduction of OsPORA activity.

Figure 6.

fgl leaves exhibit a severe xantha phenotype under continuous light (CL) conditions. (a–e) Plants were grown for 14 days in short-day (SD; 10–h light/14–h dark) and CL conditions in a growth chamber (200 μmol m−2 s−1). Leaf phenotype (a) and total chlorophyll concentration (b) were analyzed with the second leaves of wild-type (WT) and fgl plants. (c, d) Expression levels of OsPORA and OsPORB in the WT. The second leaves for qRT-PCR analysis were sampled at ZT0 for OsPORA and at ZT3 for OsPORB. (e) Immunoblot analysis of OsPORA and OsPORB proteins. RbcL was detected as a loading control. Similar results were obtained in three independent experiments. (f, g) Daily expression patterns of OsPORA and OsPORB in WT seedlings under SD conditions. WT seeds were germinated and planted in soil, and grown in SD conditions for 7 days. At 8 days after seeding (DAS), all leaf tissues of two plants were harvested at 3–h intervals. (b–d) Student's t–test: *P < 0.05; **P < 0.01. (b–g) Mean and SD values were obtained from three biological replicates.

We further examined the diurnal expression of OsPORA and OsPORB in young wild-type seedlings under SD conditions. OsPORA expression was dark-induced, with a peak at ZT0 (at the end of night period), and rapidly decreased under the light period (Figure 6f), whereas OsPORB is expressed in a circadian rhythm with a peak at ZT3 (3 h after light exposure) (Figure 6g), similar to the expression profiles of PORA and PORB in Arabidopsis and barley (Holtorf et al., 1995; Su et al., 2001). These results suggest that in fgl mutants, OsPORA expressed during the night period contributes to light-dependent Chl synthesis, and inhibits leaf variegation during early leaf development under natural growth conditions in the paddy field (Figure 1a).

Enhanced OsPORB activity is essential for Chl synthesis under HL

In Arabidopsis, AtPORC expression increases under HL conditions, whereas both AtPORA and AtPORB are downregulated (Masuda et al., 2003), indicating that AtPORC is responsible for Chl synthesis under HL conditions. Thus, we examined whether OsPORB has a similar function to AtPORC for Chl synthesis in response to HL. The 14–day-old wild-type and fgl seedlings grown under mild light intensity (200 μmol m−2 s−1) were transferred to HL (1500 μmol m−2 s−1) for 3 days in SD conditions (Figure 7). Compared with the wild type, fgl exhibited a severely photobleached phenotype (Figure 7a,b), possibly as a result of photooxidative damage. In the wild type, OsPORA mRNA levels were lower after 3 h of HL treatment (Figure 7c), whereas OsPORB mRNA levels drastically increased (Figure 7d), indicating that the rice plant requires increased expression of OsPORB in response to HL conditions. The levels of OsPORA and OsPORB were consistent with their mRNA levels (Figure 7e), indicating that rice POR activity is regulated at the transcriptional level. These results strongly suggest that OsPORB activity is necessary for Chl synthesis under HL conditions in the paddy field, reminiscent of AtPORC function in Arabidopsis (Masuda et al., 2003).

Figure 7.

Effect of high light (HL) on Chl synthesis and OsPOR expression. Plants grown under short-day (SD) conditions (200 μmol m−2 s−1 for 10 h) for 14 days were used for HL experiment. (a, b) Phenotypes (a) and total Chl contents (b) of the leaves in 17–day-old plants treated with different daylight intensity (–HL, 200 μmol m−2 s−1; +HL, 1500 μmol m−2 s−1) for 3 days in SD conditions. (c, d) Expression levels of OsPORA (c) and OsPORB (d) in the SD-grown WT plants that were treated by HL for 3 h. Mean and SD values were obtained from three replicates. (e) Immunoblot analysis of OsPORA and OsPORB in (d). RbcL was detected as a loading control. Similar results were obtained in three independent experiments. (b–d) Student's t–test: *P < 0.05; **P < 0.01.

OsPORA functions in early leaf development and OsPORB functions throughout leaf development

Degreening and necrotic lesion formation in fgl always occurred from the top area of older leaf blades in paddy field conditions. We thus examined OsPOR expression levels in two different sections (top and middle) of the leaf blades (Figure 8a–c). OsPORA mRNA levels were relatively higher in the younger third leaf (L3) than in the older second leaf (L2), and were higher in the middle area than the top area of both leaf blades (Figure 8a,b). As the wild-type leaves matured, OsPORA transcript levels dramatically decreased in the leaf tip. OsPORB expression, however, decreased more slowly, retaining relatively higher levels in the older leaves compared with OsPORA (Figure 8a–c). We next examined their expression patterns in the second leaves of 80–DAS plants grown in the paddy field (Figures 1a,b and 8d). OsPORA mRNA levels in the top section of both wild-type and fgl leaves were significantly lower than those in the middle section (Figure 8e), whereas OsPORB mRNA levels showed no significant difference (Figure 8f). These results indicate that the reduced expression of OsPORA in older leaves can explain why leaf variegation always initiates from the top of fgl leaves in paddy field conditions.

Figure 8.

Differential expression of OsPOR genes depending on leaf age and position. (a–c) Two-week-old wild-type (WT) seedlings grown under continuous light (CL) conditions were examined. The middle (I, III) and top (II, IV) parts of the second (older) and third (younger) leaves (a) were used to determine the expression levels of OsPORA (b) and OsPORB (c), and were examined in each indicated part of the leaves by using qRT-PCR with three replications. Scale bar: 1 cm. (d–f) The fourth leaf blades of the main stem in the WT and fgl at 80 days after seeding (DAS) in the paddy field (Figure 1b; IV) were examined: T, top; M, middle region. (b, c, e, f) Student's t–test: *P < 0.05; **P < 0.01.

The significance of OsPORA in Chl synthesis during early leaf development was further examined by monitoring the appearance of leaf variegation and necrotic lesion formation of the third leaf blade emerging from the second leaf sheath in fgl seedlings grown under CL conditions (Figure S5). Interestingly, the newly emerging leaf blade exhibited normal green color at 1 day after emergence (DAE), similar to the fgl phenotype under SD conditions. As the leaf blade elongated, however, degreening occurred rapidly from the top of the leaf blade (Figure 8a). This strongly suggests that OsPORA is responsible for Chl synthesis during early leaf development, at least before the emergence of the leaf sheath.

To further understand the function of OsPORs after the seedling stage, we examined OsPOR mRNA levels in wild-type leaves at 7, 21 and 60 DAS in the field. OsPORA expression drastically declined at 21 and 60 DAS, compared with that at 7 DAS (Figure 9a), whereas OsPORB expression was relatively constant, even at 60 DAS (Figure 9b). These results were quite similar to leaf age-dependent mRNA variations of Arabidopsis and barley PORs (Armstrong et al., 1995; Schunmann and Ougham, 1996). This result demonstrates that OsPORA contributes to light-dependent Chl synthesis during early leaf development, and that OsPORB functions throughout leaf development, especially during the late stages of leaf development in rice.

Figure 9.

Effect of leaf age on OsPOR expression. Wild-type (WT) plants were grown under SD conditions (cool white light, 500 μmol m−2 s−1) in the growth chambers. The third leaves from the bottom of the main culm were marked and sampled at ZT0; young (7 days after seeding, DAS), fully elongated (21 DAS) and old (60 DAS) leaf stages, and expression levels of OsPORA (a) and OsPORB (b) were examined. Mean and SD values were obtained with three biological replicates. Similar results were obtained in two independent experiments.

ROS production and reduced expression of Chl synthesis- and photosynthesis-associated genes in fgl leaves

It has been reported that leaf variegation and/or necrotic lesions in some rice mutants can be caused by excessive accumulation of ROS (Li et al., 2010; Jiang et al., 2011). We thus examined the levels of three types of ROS (singlet oxygen, superoxide and hydrogen peroxide) in the fgl leaves grown under CL conditions. Singlet oxygen was highly accumulated in the fgl mutant leaves compared with the wild type (Figure 10a), and the accumulation levels of the other reactive oxygen species (ROS), superoxide and hydrogen peroxide, were also relatively higher in the fgl leaves than in the wild type (Figure 10b,c). To further understand the molecular basis for the fgl phenotype, we examined the expression levels of Chl synthesis-associated genes (HEMA, GSAT, CHLH, PORA, DVR and CAO) and photosynthesis-associated genes (Lhcb1 and Lhcb4) in the leaves of wild-type and fgl plants grown for 2 weeks under CL conditions. Based on the expression levels of the wild type, two sectors (light-green and yellow) of fgl leaves were compared. In the light-green middle sectors of fgl leaves, mRNA levels of HEMA, GSAT, CHLH, DVR, CAO, Lhcb1 and Lhcb4 were downregulated (Figure 10d). Interestingly, the OsPORA mRNA level was markedly increased, possibly compensating for the lack of OsPORB activity in fgl; however, all eight genes were significantly downregulated in the yellow sectors of fgl leaves. Reduced expression of the genes involved in Chl synthesis often causes a reduction of the levels of several Chl intermediates (Alawady and Grimm, 2005). Thus, we further examined the levels of three Chl biosynthetic intermediates, including 5–aminolevulinic acid (ALA), protoporphyrin IX (Proto IX) and Mg protoporphyrin IX monoester (MPE), in 2–week-old plants grown under SD and CL conditions (Figure S6). As expected, their levels were slightly lower under SDs, but decreased drastically under CL, indicating that the repressed expression of the Chl biosynthetic genes caused the subsequent reduction of Chl intermediates in fgl leaves under CL conditions. Together, these results suggest that leaf variegation is closely associated with an increase of ROS levels, which leads to the downregulation of Chl synthesis- and photosynthesis-associated genes, and finally to the formation of necrotic lesions in the yellow sectors.

Figure 10.

Reactive oxygen species (ROS) and transcriptional analyses in fgl leaves. The leaves of 2–week-old WT and fgl seedlings grown under CL conditions were examined. (a) Singlet oxygen (1O2) accumulation. 1O2 was detected by Singlet Oxygen Sensor Green® reagent (SOSG) fluorescence (see Experimental procedures). Red Chl autofluorescence (left), green SOSG fluorescence (middle) and the merged images (right) are shown in the upper (WT) and lower (fgl) panels. Scale bar: 50 μm. (b, c) Hydrogen peroxide (H2O2) and superoxide anion radicals (O2) in fgl leaves. The accumulation of H2O2 and O2 were visualized by staining with DAB (b) and NBT (c), respectively. (d) Expression levels of Chl synthesis- and photosynthesis-associated genes. Black and gray bars represent relative mRNA levels in the light-green and yellow sectors of fgl, respectively, that are normalized to their levels in the green tissues of WT. Mean and SD values were obtained in one experiment with three biological replicates. This qRT-PCR analysis was performed twice with similar results. HEMA encodes glutamyl-tRNA reductase (Os10g35840); GSAT, glutamate-1-semialdehyde aminotransferase (Os08 g41990); CHLH, Mg-chelatase H subunit (Os03g20700); DVR, 3,8–divinyl chlorophyllide a 8–vinyl reductase (Os03g22780); CAO, chlorophyllide a oxygenase (Os10g41780); Lhcb1, light-harvesting Chl a/b-binding protein of PSII (Os01g41740); Lhcb4 (Os07g37240).

Discussion

By map-based cloning of the fgl locus in rice, we found that FGL encodes OsPORB, which acts in the light-dependent reduction of Pchlide a to Chlide a in Chl synthesis. fgl is a osporB-null mutant, the first to be identified in a monocot species (Figure 2). We next investigated the distinct roles of OsPORA and OsPORB depending on light conditions and leaf developmental stages. In the fgl mutants, degreening and necrotic lesion formation always initiated in the distal area of leaves and proceeded downwards, which is attributed to the precocious downregulation of OsPORA in older leaves and in the distal region of leaf blades in the absence of OsPORB activity. Under HL conditions, the whole leaves of fgl turned yellow rapidly and showed necrotic lesions (Figure 7a). Because OsPORB is highly upregulated under HL conditions (Figure 7d), this photobleaching is closely associated with a lack of OsPORB, which functions mainly in fluence rate-dependent Chl synthesis. Taking these results together, we concluded that in field conditions, both degreening and lesion formation in fgl mutants, commencing from the distal regions of older leaves, is caused by the fluence rate- and age-dependent suppression of OsPORA, resulting in no POR activity.

The necrotic lesion formation in fgl mutant leaves is likely to result from the photosensitization caused by the increased accumulation of non-photoactive Pchlide (Figure 4), which produces singlet oxygen upon illumination (Chakraborty and Tripathy, 1992; op den Camp et al., 2003). In fgl, all three types of ROS (singlet oxygen, superoxide and hydrogen peroxide) accumulated to high levels in the mutant leaf tissues (Figure 10a,b). The excessive ROS causes photo-oxidative damage and cell death by non-enzymatic oxidation of cellular compounds, such as the direct peroxidation of lipids, or by activating signal transduction pathways triggering cellular processes such as the alteration of photosynthetic functions (Kim et al., 2008). In addition, ROS causes an alteration of gene expression in the nucleus. The excessive ROS generated in the underdeveloped fgl chloroplasts (Figure 3d,e) obviously downregulates Chl synthesis- and photosynthesis-associated genes through the retrograde signaling from the chloroplast to the nucleus (Gadjev et al. 2006). We found that the expression levels of upstream (HEMA, GSAT and CHLH) and downstream (CAO) genes of POR in the Chl synthesis pathway, as well as photosynthetic genes (Lhcb1 and Lhcb4), were downregulated in fgl (Figure 10d). Thus, we concluded that underdeveloped chloroplasts and necrotic lesions in the fgl leaves are caused by oxidative damage as well as the alteration of gene expressions, possibly because of an excessive accumulation of ROS derived from non-photoactive Pchlide in older leaf tissues and/or HL conditions in the field.

The expression patterns of the two POR genes in barley and rice are substantially similar to each other; however, the fgl phenotypes at different developmental stages and in different environmental conditions strongly suggest that the molecular functions of PORA and PORB differ in rice and barley. A light-harvesting POR:Pchlide complex (LHPP) has been reported in barley etioplasts, but not in other plants (Reinbothe et al., 1999, 2003). In barley, PORB bound to Pchlide a, which is a photoactive form of LHPP that can be converted to Chlide a, whereas PORA bound to Pchlide b remains photo-inactive (Reinbothe et al., 2003). However, fgl mutants can synthesize Chl under light conditions, at least during the greening of etiolated seedlings and in the early stages of leaf development (Figures 1 and 5 and Figure S5), indicating that in rice, Pchlide can be converted to Chlide in the absence of OsPORB activity. This indicates that OsPORA is also able to form a photoactive LHPP complex, or that complex formation is not necessary for OsPORA function. Yuan et al. (2012) recently demonstrated that LHPP assembly is required for POR function and seedling greening in barley, but not in Arabidopsis. Thus, further study will be required to elucidate the mechanism of functional OsPOR:Pchlide assembly in rice.

Most higher plants have one or two POR isoforms (Masuda et al., 2003), although the Arabidopsis genome contains three POR genes (Su et al., 2001). Therefore, the POR functions in each plant species are likely to be variable. Here, we demonstrate the developmental and physiological functions of rice PORs by comparing the rice fgl/osporB mutant with the previously reported Arabidopsis por single or double mutants. Both Arabidopsis and rice PORB genes are constitutively expressed throughout development (Figure 9b) (Armstrong et al., 1995), but their responses to light intensity were somewhat different. We found that the fgl mutant phenotype became more severe as light intensity increased, and necrotic spots increased drastically under HL conditions in the growth chamber (Figure 7) or in field conditions (Figure 1); however, it has shown that the atporB mutant displayed almost the same phenotype as the wild type under HL (1000 μmol m−2 s−1; Masuda et al., 2003). Interestingly, in response to HL treatment, leaf phenotype and Chl levels of the the fgl/osporB mutant were reminiscent of those of atporC mutants (Masuda et al., 2003). In addition, OsPORB expression was rapidly upregulated by HL treatment (Figure 5h), similar to the expression of AtPORC (Masuda et al., 2003). In spite of some discrepancies between the rice fgl and Arabidopsis atporB mutants in response to HL, their etiolated seedlings exhibited similar defective phenotypes in PLB size and Pchlide levels (Figures 3b and 4). Together, our results strongly suggest that OsPORB has overlapping functions corresponding to both AtPORB and AtPORC.

Next, we compared PORA functions between the atporB atporC double mutant and the fgl/osporB mutant, which both have only PORA activity throughout development. During plant growth, AtPORA is transiently expressed in early seedling development (Armstrong et al., 1995) and atporB atporC double mutants exhibited severe growth defects (Frick et al., 2003). By contrast, we found that fgl mutants displayed a much milder growth defect (Figure 1) because OsPORA expression expanded to developing leaves after the seedling stage (Figure 8a). The atporB atporC double mutant showed a seedling-lethal xantha (highly chlorophyll-deficient) phenotype at the cotyledon stage, with a severe defect in plant growth and development (Frick et al., 2003). This indicates that OsPORA activity is still critical for Chl synthesis after the seedling stage in rice, which corresponds to AtPORB function in Arabidopsis. Probably the difference of PORA activities in two species may be caused by the difference in temporal and spatial expression of PORA, not by functional difference. The functional redundancy among Arabidopsis PORs has been well characterized: i.e. the constitutive expression of AtPORA fully rescued the atporB atporC double mutant (Paddock et al., 2010). In this respect, it should be examined whether the overexpression of OsPORA could rescue the fgl phenotype, similar to the case in Arabidopsis. Thus, OsPORA appears to have a function equivalent to both AtPORA and AtPORB. In agreement with this hypothesis, a phylogenic tree of Arabidopsis, rice and barley PORs (Figure S7) suggests that OsPORA and HvPORA are evolutionarily closer to AtPORA and AtPORB, but that OsPORB and HvPORB are similar to AtPORC.

In summary, we propose that OsPORA has overlapping functions corresponding to AtPORA and AtPORB, but that OsPORB has functions corresponding to AtPORB and AtPORC. In higher plants, Chl synthesis must be finely controlled in various light conditions and throughout development. For the strict control of bulk Chl synthesis in each higher plant species, PORs seem to have been duplicated and subsequently subfunctionalized, resulting in the generation of paralogous copies with conserved functions, but different expression patterns, over time or in response to environmental conditions, as observed in the dicot Arabidopsis and the monocot rice. Such gene duplication and subsequent functional specialization among POR paralogs seems to facilitate adaptive evolution in plants (Flagel and Wendel, 2009).

Experimental procedures

Plant materials and growth conditions

The single recessive fgl mutant was previously generated from a japonica rice cultivar ‘Kinmaze’ by methyl nitrosourea (MNU) mutagenesis, as previously described (Iwata and Omura, 1975). For phenotypic characterization, plants were grown in the paddy field or in a growth chamber. The chamber conditions were as follows: switching between 10 h of light (200 μmol m−2 s−1) at 30°C and 14 h of dark at 20°C (SD conditions) or continuous light, with 10 h at 30°C and 14 h at 20°C (CL conditions). Samples of etiolated seedlings were collected under a dim green safe light for transmission electron microscopy (TEM), Pchlide quantification and quantitative real-time PCR (qRT-PCR) analyses.

Measurement of photosynthetic pigments

For the measurement of total Chl and carotenoid concentration, pigments were extracted from leaf tissues of wild-type (Kinmaze) and fgl plants with 80% ice-cold acetone. The concentrations of Chl and carotenoid were determined with a UV/VIS spectrophotometer, as described previously (Lichtenthaler, 1987). Etiolated seedlings were used to measure total Pchlide, and light (200 μmol m−2 s−1) treatment was used to measure non-photoactive Pchlide. Total photosynthetic pigments were extracted from 100 mg of tissue in 2 ml of 80% acetone with 0.1 N ammonium hydroxide at 4°C. Fluorescence emission of the pigment extracts were recorded using a fluorescence spectrophotometer XRF-1700 (Shimadzu, http://www.shimadzu.com). An excitation wavelength of 433 nm was used to record emissions at 634 nm. Pchlide concentrations were calculated as previously described (Sperling et al., 1998). The concentrations of ALA, Proto IX and MPE in 2–week-old WT and fgl leaves grown under SD and CL conditions were measured from fluorescence emission spectra (553, 400 and 420 nm) using XRF1700 (Shimadzu), as described previously (Hukmani and Tripathy, 1992; Sood et al., 2005).

SDS-PAGE and immunoblot analysis

The detection of photosynthesis-related and OsPOR proteins by immunoblot analysis was carried out as previously described (Han et al., 2012). Leaf tissue (1.0 mg) was homogenized with 10 μl of sample buffer (50 mM Tris, pH 6.8, 2 mM EDTA, 10% w/v glycerol, 2% SDS and 6% 2–mercaptoethanol), then denatured at 75°C for 3 min, and finally subjected to SDS-PAGE. The antibodies for detection of Lhcb1, PsbC, Lhca1, PsaF and POR were obtained from Agrisera (http://www.agrisera.com). The RbcL protein was visualized by staining the immunoblotted membranes with Coomassie Brilliant Blue reagent (Sigma-Aldrich, http://www.sigmaaldrich.com).

Map-based cloning of fgl

A mapping population of 1032 F2 individuals was generated by crossing a japonica-type fgl mutant and an indica-type cultivar, Dasanbyeo. Initial localization was determined using polymorphic SSR and STS markers between two mapping parents (Dasanbyeo and fgl). For fine mapping, STS markers were designed by comparison of genomic DNA sequences between ‘9311’ and Nipponbare using the Kropbase database (http://kropbase.snu.ac.kr). SSR marker information was obtained from the Gramene website (http://www.gramene.org/markers). For dCAPS (derived cleaved amplified polymorphic sequence) analysis, a mismatch forward primer (5′-GGCGTTCTTGGGCGTTCGTCTCGGCGACGT-3′) and the reverse primer (5′-GCACAGAGCAAGCAGAGTGCTG-3′) were designed using dcaps finder 2.0 (http://helix.wustl.edu/dcaps/dcaps.html). Each PCR product was digested with the AatII restriction enzyme, and digested PCR products were separated on agarose gels for visualization.

Rice transformation

For complementation of the fgl mutation, a full-length cDNA of OsPORB was cloned into the pMDC32 gateway binary vector containing the cauliflower mosaic virus 35S promoter. The 35S:OsPORB in pMDC32 was introduced into the calli generated from the mature seed embryos of fgl mutants through the Agrobacterium (strain LBA4404)-mediated method (Jeon et al., 2000). Transformants were confirmed by PCR amplification using the specific primers listed in Table S1.

TEM analysis

Leaf samples for TEM analysis were harvested from dark-grown etiolated or light-grown seedlings. Fixation and polymerization of leaf samples were carried out as described previously (Park et al., 2007). The sections were sliced to 60 nm with an ultramicrotome (MT–X; RMC, http://www.rmcproducts.com/) and were stained with 2% uranyl acetale for 5 min and Reynold's lead citrate for 2 min at 25°C. The processed samples were finally examined using a JEM-1010 EX electron microscope (JEOL, http://www.jeol.com).

Analysis of reactive oxygen species

For singlet oxygen staining, the second leaf blades of the first panicle were used. Two-week-old wild-type and fgl leaves were infiltrated in the dark with a solution of 100 μM Singlet Oxygen Sensor Green® reagent (SOSG; Invitrogen, http://www.invitrogen.com) in 50 mM phosphate potassium buffer (pH 7.5). After 30 min of incubation leaf discs were washed with distilled water for 5 min. Following excitation at 480 nm, the fluorescence emission at 520 nm was detected by laser scanning confocal microscopy (LSM510; Carl Zeiss, http://corporate.zeiss.com). Red Chl autofluorescence was collected at 680 nm. Detection of hydrogen peroxide (H2O2) and superoxide (O2) using DAB and NBT, respectively, was performed as previously described (Han et al., 2012).

Pchlide quantification by fluorescence spectroscopy

Etiolated seedlings were used to measure total Pchlide, and then10–min light (200 μmol m−2 s−1)-treated etiolated seedlings were used to measure non-photoactive Pchlide by fluorescence emission analyses. Total photosynthetic pigments were extracted from 100 mg tissues in 2 ml of 80% acetone with 0.1 N ammonium hydroxide at 4°C. Fluorescence emission of the pigment extracts was recorded using a fluorescence spectrophotometer XRF-1700 (Shimadzu, http://www.shimadzu.com). An excitation wavelength of 433 nm was used to record emission spectra. The resulting emission bands were measured at 634 nm. Pchlide concentrations were calculated according to the method described by Sperling et al. (1998).

qRT-PCR analysis

The expression levels of OsPORs, Chl synthesis- and photosynthesis-associated genes were determined by qRT-PCR analysis. Total RNA extraction, reverse transcription and qRT-PCR analysis were carried out as previously described (Han et al., 2012). Primers used for qRT-PCR are listed in Table S2.

Acknowledgements

We thank Do-In Kim for her excellent technical support, and the ABRC for pMDC32 plasmids. This work was supported by a grant from the Next-Generation BioGreen 21 Program (Plant Molecular Breeding Center No. PJ008128), Rural Development Administration, Republic of Korea.

Ancillary

Advertisement