Cell Growth Defect Factor1/CHAPERONE-LIKE PROTEIN OF POR1 Plays a Role in Stabilization of Light-Dependent Protochlorophyllide Oxidoreductase in Nicotiana benthamiana and Arabidopsis^,

Protochlorophyllide oxidoreductase (POR), which catalyzes the protochlorophyllide reduction step in the chlorophyll biosynthetic pathway, is activated by light, and its substrate and product are photosensitizers. This work reports that CHAPERONE-LIKE PROTEIN OF POR1 interacts with POR isoforms in chloroplast membranes and protects POR proteins from photooxidative damage with its chaperone activity.

Abstract

Angiosperms require light for chlorophyll biosynthesis because one reaction in the pathway, the reduction of protochlorophyllide (Pchlide) to chlorophyllide, is catalyzed by the light-dependent protochlorophyllide oxidoreductase (POR). Here, we report that Cell growth defect factor1 (Cdf1), renamed here as CHAPERONE-LIKE PROTEIN OF POR1 (CPP1), an essential protein for chloroplast development, plays a role in the regulation of POR stability and function. Cdf1/CPP1 contains a J-like domain and three transmembrane domains, is localized in the thylakoid and envelope membranes, and interacts with POR isoforms in chloroplasts. CPP1 can stabilize POR proteins with its holdase chaperone activity. CPP1 deficiency results in diminished POR protein accumulation and defective chlorophyll synthesis, leading to photobleaching and growth inhibition of plants under light conditions. CPP1 depletion also causes reduced POR accumulation in etioplasts of dark-grown plants and as a result impairs the formation of prolamellar bodies, which subsequently affects chloroplast biogenesis upon illumination. Furthermore, in cyanobacteria, the CPP1 homolog critically regulates POR accumulation and chlorophyll synthesis under high-light conditions, in which the dark-operative Pchlide oxidoreductase is repressed by its oxygen sensitivity. These findings and the ubiquitous presence of CPP1 in oxygenic photosynthetic organisms suggest the conserved nature of CPP1 function in the regulation of POR.

INTRODUCTION

Chaperones play essential roles in the regulation of protein activity and interactions under changing environments. In chloroplasts of oxygenic photosynthetic organisms, chaperones such as heat shock proteins, Hsp70, Hsp90, and Hsp60, and proteases, including Caseinolytic protease, Filamentous temperature sensitive H, and Degradation of periplasmic protein participate in protein folding/unfolding, translocation, assembly/disassembly of protein complexes, and degradation (Mulo et al., 2008; Nordhues et al., 2010). In addition to these conventional chaperones and proteases, many chloroplast proteins play specific roles in the assembly of protein complexes and in the maintenance of functional protein conformations within chloroplasts. For example, LOW PHOTOSYSTEM II ACCUMULATION1 (LPA1) and Albino (Alb) family proteins are involved in assembly of the photosystems in Arabidopsis thaliana and Chlamydomonas reinhardtii, respectively (Göhre et al., 2006; Peng et al., 2006). cpSRP43, a targeting factor in the chloroplast signal-recognition particle, functions as a specific molecular chaperone of the light-harvesting chlorophyll a,b– binding protein (CAB) to prevent its aggregation (Falk and Sinning, 2010). C. reinhardtii Calvin Cycle Protein12 (CP12) protects glyceraldehyde-3-phosphate dehydrogenase (GAPDH) from denaturation and aggregation as a permanent chaperone-like protein in chloroplasts (Erales et al., 2009).

In angiosperms, the chlorophyll biosynthetic pathway consists of a complex series of reactions catalyzed by multiple nuclear-encoded enzymes (Masuda and Fujita, 2008). The reduction of protochlorophyllide (Pchlide) to chlorophyllide via the trans addition of hydrogen across the C17-C18 double bond of the D-ring of Pchlide is a key regulatory step in chlorophyll biosynthesis and is catalyzed by two distinct enzymes, the light-dependent Pchlide oxidoreductase (POR) and the light-independent (dark-operative) Pchlide oxidoreductase (DPOR) (Masuda and Takamiya, 2004; Reinbothe et al., 2010). Both enzymes are widely distributed in oxygenic photosynthetic organisms, but angiosperms have evolved to have only POR. Due to the strict light dependence of POR for catalysis, chlorophyll biosynthesis is arrested at the Pchlide stage in dark-grown angiosperm seedlings; instead, the highly accumulated POR, together with Pchlide, NADPH, lipids, and carotenoids, form a paracrystalline membrane structure, known as the prolamellar body (PLB), within etioplasts (Masuda and Takamiya, 2004; Reinbothe et al., 2010). Illumination triggers rapid degradation of POR proteins and breakdown of PLBs, in parallel with chlorophyll synthesis and chloroplast biogenesis (Reinbothe et al., 1995; Masuda and Takamiya, 2004; Reinbothe et al., 2010).

The x-ray crystal structure of POR has not been determined at present. However, homology modeling suggests that the structure of POR is similar to those of members of the short-chain alcohol dehydrogenase family with characteristics of a globular and soluble protein (Heyes and Hunter, 2005; Sytina et al., 2008). Nevertheless, an important feature of POR assembly is its association with plastid membranes: In chloroplasts, POR is associated with the thylakoid and envelope membranes as a peripheral protein, whereas in etioplasts, POR forms large aggregates as the PLB (Masuda and Takamiya, 2004; Reinbothe et al., 2010). The mechanism of membrane binding of POR despite its lack of predictable transmembrane regions has not yet been revealed. Catalytic mechanisms of POR were studied using ultrafast pump-probe absorption difference spectroscopy (Heyes and Hunter, 2005; Sytina et al., 2008). The results indicate that the light-driven activation of POR is critically required before catalysis: The reaction is triggered by light absorption of the POR-Pchlide complex followed by light-independent steps to produce chlorophyllide.

POR genes are ubiquitously found in oxygenic photosynthetic organisms, such as cyanobacteria, green algae, mosses, gymnosperms, and angiosperms, and share a high degree of sequence similarity (Masuda and Takamiya, 2004). There are varying numbers of POR genes depending on the species: Among angiosperms, tobacco (Nicotiana tabacum) has two POR genes, POR1 and POR2 (Masuda et al., 2002), whereas Arabidopsis has three POR genes, PORA, PORB, and PORC, which show different expression profiles (Frick et al., 2003; Masuda et al., 2003; Masuda and Takamiya, 2004; Paddock et al., 2010). Arabidopsis PORA mRNA and protein are both detected abundantly in etiolated seedlings, but their levels dramatically decrease upon illumination. PORB mRNA and protein are present in both etiolated seedlings and light-grown plants, whereas the mRNA and protein levels of PORC are very low in etiolated seedlings but are induced by light. During the last two decades, the information known about POR has dramatically increased with regard to its in vivo function, catalytic mechanism, localization, evolution, and gene expression pattern. However, it remains unknown how POR is protected from photooxidative damage, considering that POR is a light-activated enzyme (Heyes and Hunter, 2005; Sytina et al., 2008) and that its substrate and product are photosensitizers (Reinbothe et al., 1996; Sperling et al., 1997; Buhr et al., 2008).

Cell growth defect factor1 (Cdf1; At5g23040) was identified by a yeast genetic screen for the isolation of Arabidopsis cell death–related genes (Kawai-Yamada et al., 2005). Cdf1 overexpression resulted in apoptosis-like cell death in yeast, accompanied by excessive accumulation of reactive oxygen species (ROS), although Cdf1 function in planta has not been characterized (Kawai-Yamada et al., 2005). During a virus-induced gene silencing (VIGS)–based screening of chloroplast development genes (Cho et al., 2004; Jeon et al., 2012), we found that silencing of Nicotiana benthamiana Cdf1 causes a significant reduction in POR protein accumulation. Here, we report that Cdf1, designated herein as chaperone-like protein of POR1 (CPP1), plays a crucial role in chlorophyll synthesis and chloroplast biogenesis by regulating the stability and function of POR.

RESULTS

Cdf1/CPP1 Is a Membrane Protein with a J-Like Domain

Cdf1/CPP1 from Arabidopsis (At-CPP1) and N. benthamiana (Nb-CPP1) is predicted to have a chloroplast transit peptide, a J-like (JL) domain, and three transmembrane domains (see Supplemental Figures 1A and 1C online). The J-domain is found in DnaJ/Hsp40 cochaperones that function as partners of the highly conserved Hsp70; it is required for both interaction with Hsp70 and stimulation of Hsp70 ATPase activity (Qiu et al., 2006; Rajan and D’Silva, 2009). CPP1 homologs are ubiquitously present in oxygenic photosynthetic organisms that have POR, including cyanobacteria, green algae, mosses, gymnosperms, and angiosperms (see Supplemental Figures 1B and 1C and Supplemental Data Set 1 online). The JL domains of Nb-CPP1, At-CPP1, and slr1918 (the CPP1 homolog of Synechocystis sp PCC6803) are similar in structure to the J-domain of Tid1 (a human mitochondrial homolog of bacterial DnaJ cochaperones; Lo et al., 2004) based on computational models, albeit with low sequence similarity (see Supplemental Figure 2 online). Interestingly, the JL domain of CPP1 lacks a highly conserved tripeptide of His, Pro, and Asp (the HPD motif) of the J-domain, which is required for its interaction with Hsp70 (Qiu et al., 2006; Rajan and D’Silva, 2009) (see Supplemental Figures 1C and 2 online).

VIGS of Nb-CPP1 Results in Defective Chloroplast Biogenesis and Chlorophyll Synthesis

We examined the in vivo effects of CPP1 deficiency in N. benthamiana and Arabidopsis using VIGS and a dexamethasone (DEX)–inducible RNA interference (RNAi) strategy. Our analyses suggest that a null mutation of CPP1 is embryo lethal in Arabidopsis (see Supplemental Figure 3 online). VIGS of Nb-CPP1 in N. benthamiana resulted in leaf yellowing, similar to the phenotypes from VIGS-mediated double silencing of N. benthamiana POR1 (Nb-POR1) and Nb-POR2 (Figures 1A and 1C). VIGS of either Nb-POR1 or Nb-POR2 alone caused no visible phenotype. Real-time quantitative RT-PCR showed silencing of Nb-CPP1 in tobacco rattle virus (TRV):NbCPP1 lines, but not in TRV:NbPOR1/2 lines, compared with the TRV control (Figure 1B). Semiquantitative RT-PCR showed silencing of both Nb-POR1 and Nb-POR2 in TRV:NbPOR1/2 double-silenced lines compared with the TRV control (Figure 1D). Differential interference contrast and confocal laser scanning microscopy revealed that the average number and size of chloroplasts in the yellow sector of TRV:NbCPP1 leaves were reduced to ∼31.2 and ∼66.9% of the TRV control, respectively (Figure 1E; see Supplemental Figure 4 online). Chlorophyll autofluorescence of the TRV:NbCPP1 chloroplasts diminished to ∼4.9% of the control chloroplasts (Figure 1E). Total chlorophyll content in the yellow sectors of the TRV:NbCPP1 leaves was also reduced to ∼15.4% of levels in the TRV control, suggesting defective chlorophyll synthesis (Figure 1F). Transmission electron microscopy (TEM) of the spongy mesophyll cells in the yellow sectors revealed significantly decreased chloroplast numbers and abnormal chloroplast morphology in Nb-CPP1 VIGS plants (Figure 1G). The affected chloroplasts were small and elongated and contained poorly developed thylakoids. Taken together, these results demonstrate that Nb-CPP1 deficiency impairs chlorophyll synthesis and chloroplast biogenesis.

Figure 1.

VIGS of Nb-CPP1 and Nb-POR1/2.

(A) For VIGS of Nb-CPP1, a 726-bp Nb-CPP1 cDNA fragment was cloned into the TRV-based VIGS vector pTV00. To simultaneously silence Nb-POR1 and Nb-POR2 using VIGS, a 402-bp Nb-POR1 and a 400-bp Nb-POR2 cDNA fragment were cloned together into pTV00. aa, amino acids.

(B) The amount of endogenous Nb-CPP1 mRNA in TRV:NbCPP1 and TRV:NbPOR1/2 VIGS lines was measured by real-time quantitative RT-PCR to confirm gene silencing. β-Tubulin mRNA levels were used as a control. Data points represent means ± sd of three experiments. Asterisks denote the statistical significance of the differences between the samples. **P ≤ 0.01.

(C) VIGS of Nb-CPP1 and Nb-POR1/2 both resulted in leaf yellowing phenotypes compared with the TRV control. VIGS of either Nb-POR1 or Nb-POR2 alone did not result in any visible effects. The plants were photographed 20 d after infiltration.

(D) The amounts of endogenous Nb-POR1 and Nb-POR2 mRNAs in the TRV:NbPOR1/2 double-silenced lines were measured by RT-PCR. Actin mRNA levels were used as a control.

(E) Chlorophyll autofluorescence of chloroplasts (pseudo-colored blue) in leaf protoplasts of the TRV control and TRV:NbCPP1 lines was visualized (left) and quantified (right) by confocal laser scanning microscopy. Differential interference contrast (DIC) images are also shown. Yellow/white sectors of TRV:NbCPP1 leaves were used for the analyses. Data are expressed as mean ± sd of 30 individual protoplasts.

(F) Quantification of total chlorophyll, chlorophyll a, and chlorophyll b in TRV and TRV:NbCPP1 leaves. FW, fresh weight.

(G) Light micrographs of leaf sections (a and e) and transmission electron micrographs of leaf mesophyll cells (b and f) and chloroplasts (c, d, g, and h) from the TRV control (a to d) and TRV:NbCPP1 lines (e to h). c, chloroplasts. Bars = 100 μm in (a) and (e), 5 μm in (b) and (f), 1 μm in (c), (d), and (g), and 2 μm in (h).

Silencing of CPP1 Results in Photobleaching in Arabidopsis

The DEX-inducible At-CPP1 RNAi seedlings were germinated and grown on medium containing ethanol (−DEX) or 10 μM DEX. Under low-light conditions, (+)DEX RNAi seedlings were photobleached in the cotyledons, accompanied by excessive accumulation of ROS, visualized by nitro blue tetrazolium (NBT) staining (Figures 2A and 2B; see Supplemental Figure 5 online). NBT forms a dark-blue formazan precipitate when in contact with superoxide radicals (Bielski et al., 1980). Under high-light conditions, growth of (+)DEX RNAi seedlings was stunted, with excessive production of anthocyanin pigments in cotyledons (Figure 2A; see Supplemental Figure 5 online). When the At-CPP1 RNAi plants were grown on soil and sprayed with DEX, the plants developed leaf yellowing with growth retardation (Figure 2C). Real-time quantitative RT-PCR revealed silencing of At-CPP1 upon DEX spraying (Figure 2D).

Figure 2.

Analyses of Arabidopsis DEX-Inducible At-CPP1 RNAi Lines.

(A) Phenotypes of DEX-inducible At-CPP1 RNAi seedlings. The RNAi seedlings (line #1) were germinated and grown under low light (20 μmol m⁻² s⁻¹) or high light (200 μmol m⁻² s⁻¹) on medium containing ethanol (−DEX) or DEX.

(B) NBT staining to visualize superoxide production in the RNAi seedlings (lines #1 and #4) grown under low light (20 μmol m⁻² s⁻¹). The cotyledons of the seedlings grown on medium with DEX exhibited a strong blue color indicating excessive production of superoxide.

(C) Leaf yellowing following DEX treatment. RNAi seedlings (line #1) were grown for 2 weeks on soil and then sprayed with ethanol (−DEX) or 30 μM DEX for 10 d. Leaf morphology of the RNAi plants following 10 d of ethanol or DEX spraying is also shown.

(D) Real-time quantitative RT-PCR analysis with the RNAi seedlings (line #1) confirmed silencing of At-CPP1 upon spraying with DEX (0, 4, and 8 d). Data points represent means ± sd of three experiments. Asterisks denote the statistical significance of the differences between the samples. **P ≤ 0.01.

CPP1 Is Localized in Chloroplast Membranes

To determine the subcellular localization of CPP1, green fluorescent protein (GFP) fusion constructs of Nb-CPP1 and At-CPP1 were transiently expressed in N. benthamiana leaves using agroinfiltration (Voinnet et al., 2003). Confocal laser scanning microscopy using protoplasts isolated from the infiltrated leaves revealed that green fluorescent signals of AtCPP1:GFP were detected in the chloroplasts, but not in mitochondria, visualized by tetramethyl rhodamine methyl ester (TMRM) staining (Figure 3A). In yeast, At-CPP1 (previously named Cdf1) was localized to mitochondria, and its overexpression caused Bax-like lethality (Kawai-Yamada et al., 2005). Green fluorescent signals of NbCPP1:GFP fusion proteins were detected in chloroplasts, and a GFP fusion protein of Nb-POR1 (NbPOR1:GFP) also showed a similar localization pattern in chloroplasts (Figure 3B). To determine the subplastidic localization of CPP1, immunoblotting was performed with chloroplast fractions purified from N. benthamiana leaves. Immunoblotting using anti-CPP1 and anti-POR antibodies detected CPP1 and POR proteins of N. benthamiana, respectively, in both thylakoid and envelope membrane fractions (Figure 3C). As a control for fractionation, the corresponding antibodies detected rbcL, Tic40, and D1 in the stromal, envelope, and thylakoid fractions, respectively. Localization of CPP1 in plastid membranes was confirmed by proteomics studies in Arabidopsis, maize (Zea mays), and cauliflower (Brassica oleracea ssp botrytis) (Bräutigam and Weber, 2009; Ferro et al., 2010; Huang et al., 2013). POR proteins were previously detected in both stromal thylakoid and envelope membranes in chloroplasts of light-adapted plants, similar to the dual localization of several enzymes catalyzing the final steps of chlorophyll biosynthesis (Barthélemy et al., 2000; Masuda and Fujita, 2008).

Figure 3.

Chloroplast Localization of CPP1.

(A) Chloroplast localization of AtCPP1:GFP. AtCPP1:GFP fusion protein was transiently expressed in N. benthamiana leaves by agroinfiltration and observed by confocal laser scanning microscopy. Mitochondria were visualized by TMRM staining. Chlorophyll autofluorescence is pseudo-colored blue.

(B) Chloroplast localization of NbCPP1:GFP and NbPOR1:GFP. NbCPP1:GFP fusion protein was transiently expressed in N. benthamiana leaves by agroinfiltration and observed by confocal laser scanning microscopy (top). After agroinfiltration of the N. benthamiana POR1 (NbPOR1):GFP construct, protoplasts were isolated and observed by confocal microscopy (bottom). Green fluorescent signals of NbCPP1:GFP and NbPOR1:GFP were detected in the chloroplasts.

(C) Immunoblotting showing localization of Nb-CPP1 in chloroplast envelope and thylakoid membranes. As a control, specific antibodies detected rbcL, Tic40, and D1 in the stromal, envelope, and thylakoid fractions, respectively.

CPP1 Interacts with POR Isoforms

To determine whether CPP1 and POR interact, we first used bimolecular fluorescence complementation (BiFC). CPP1 and POR isoforms of Arabidopsis and N. benthamiana were expressed in combination as yellow fluorescent protein (YFP)^N and YFP^C fusion proteins in N. benthamiana leaves by agroinfiltration. Confocal laser scanning microscopy of the mesophyll cells of the infiltrated leaves (Figure 4A) and their protoplasts (see Supplemental Figure 6A online) revealed strong YFP fluorescence within chloroplasts, indicating an interaction between CPP1 and POR proteins regardless of the species of origin. However, BiFC of At-CPP1 with Arabidopsis CHL27 (for Mg²⁺-protoporphyrin IX monomethylester oxidative cyclase), CHLM (for Mg²⁺-protoporphyrin IX methyl transferase), and GENOMES UNCOUPLED4 (GUN4) and BiFC of Nb-CPP1 with N. benthamiana CHLI2 (for Mg-chelatase subunit I isoform 2) all resulted in no fluorescence, suggesting a lack of interaction (Figure 4A; see Supplemental Figure 6B online). Next, we performed coimmunoprecipitation assays (Figure 4B). A fusion protein between Nb-POR1 and the hemagglutinin tag (NbPOR1:HA) and NbPOR2:HA were transiently expressed using agroinfiltration and then immunoprecipitated from n-dodesyl-β-d-maltoside–solubilized chloroplast membranes using anti-HA antibodies. The immunoprecipitates contained the coexpressed NbCPP1:Flag proteins, suggesting interaction of Nb-CPP1 with both Nb-POR1 and Nb-POR2 in vivo (Figure 4B). However, HA fusions of N. benthamiana GUN4 (Nb-GUN4) or CHLI2 (Nb-CHLI2) could not be coimmunoprecipitated by NbCPP1:Flag, consistent with the BiFC results (see Supplemental Figure 7 online). For in vitro binding assays, purified maltose binding protein (MBP)–fused Nb-CPP1 and slr1918 (the CPP1 homolog of Synechocystis sp PCC6803) were immobilized on MBP resin and incubated with lysates of Escherichia coli cells expressing His tag–fused Nb-POR1 or Nb-GUN4 proteins (Figure 4C). After extensive washing of the resin, bound proteins were eluted and subjected to immunoblotting with anti-MBP and anti-His antibodies. Resin-bound MBP:NbCPP1 and MBP:slr1918, but not MBP alone, could pull down NbPOR1:His in vitro, suggesting an interaction between the CPP1 homologs and POR (Figure 4C). However, MBP:NbCPP1 could not pull down NbGUN4:His (Figure 4C). Collectively, these results support that CPP1 and POR interact with each other in chloroplast membranes.

Figure 4.

Interactions of CPP1 with POR Isoforms in Chloroplasts.

(A) BiFC-visualized interactions between CPP1 and POR isoforms of N. benthamiana and Arabidopsis. CHL27 encodes Mg²⁺-protoporphyrin IX monomethylester oxidative cyclase involved in chlorophyll biosynthesis. Y^N, YFP^N; Y^C, YFP^C.

(B) Coimmunoprecipitation of NbCPP1:Flag and NbPOR:HA proteins. After agroinfiltration, the chloroplast membrane fraction was immunoprecipitated (IP) with anti-HA antibodies, and the coimmunoprecipitates were detected by immunoblotting (IB) using anti-Flag antibodies.

(C) In vitro binding assays showing interactions of MBP:NbCPP1 and MBP:slr1918 with NbPOR1:His, but not with NbGUN4:His. Slr1918 is the CPP1 homolog of Synechocystis sp PCC6803.

CPP1 Deficiency Leads to Reduced Accumulation of POR Proteins in Light-Grown Plants

Immunoblotting demonstrated that steady state POR protein levels in pale-green sectors of the leaves were significantly reduced in TRV:NbCPP1 VIGS N. benthamiana plants and TRV:NbPOR1/2 plants (for double silencing of POR1 and POR2) compared with control TRV or TRV:NbGyrB (for DNA gyrase subunit B) lines, whereas the levels of ATP synthase β-subunit (atpB), CAB, D1, and chloroplast HSP70 (cpHSP70) remained constant (Figure 5A). The VIGS control TRV:NbGyrB lines are silenced for chloroplast-targeted GyrB of N. benthamiana and show leaf yellowing (Cho et al., 2004). Thus, CPP1 deficiency resulted in reduced POR protein accumulation in N. benthamiana. In Arabidopsis, silencing of CPP1 by DEX spraying decreased POR protein levels in seedlings from four independent DEX-inducible CPP1 RNAi lines compared with control samples (see Supplemental Figure 8 online). Total POR protein levels and CPP1 protein levels progressively decreased in the Arabidopsis RNAi seedlings following 0 to 8 d of DEX spraying, whereas the levels of other control chloroplast proteins were unchanged (Figure 5C). The transcript levels of Arabidopsis PORB and PORC, which are expressed under light conditions (Frick et al., 2003; Masuda et al., 2003), as well as the levels of CHLH (for Mg-chelatase subunit H) and CHL27 remained unaffected by CPP1 silencing, whereas CPP1 was silenced by RNAi (Figure 5D). These results suggest that CPP1 is required for POR protein accumulation in chloroplasts.

Figure 5.

Reduced POR Stability in CPP1-Silenced Plants.

(A) Immunoblotting of leaf protein extracts from the N. benthamiana VIGS lines. Pale-green sectors were used for the VIGS lines that show leaf yellowing.

(B) Fluorescence emission spectra showing relative fluorescence of Pchlide (fluorescence emission maximum at 636 nm) in the VIGS plants.

(C) Immunoblotting to examine time-dependent regulation of CPP1 and POR protein levels in DEX-inducible At-CPP1 RNAi Arabidopsis seedlings (line #1) after 0 to 8 d of ethanol (−DEX) or DEX spraying. GluTR, glutamyl-tRNA reductase.

(D) RT-PCR in At-CPP1 RNAi seedlings (line #1) after 0, 4, or 8 d of DEX spraying. The transcript level of UBC10 was used as a control.

(E) Susceptibility of POR proteins to photooxidative stress. Wild-type (WT) N. benthamiana plants grown under normal light conditions (60 μmol m⁻² s⁻¹) were transferred to high-light intensity (500 μmol m⁻² s⁻¹) and incubated for 0 to 72 h. Phenotypes of the plants at 0 (HL-0 h) and 72 h of high-light treatment (HL-72 h) are shown. Leaf protein extracts prepared from the high-light-treated plants at various time points were subjected to immunoblotting using various antibodies. At 72 h of treatment, green (g), pale-green (pg), and yellow leaves (y) were collected for immunoblotting.

We next examined whether CPP1 depletion leads to accumulation of the POR substrate Pchlide caused by POR deficiency by measuring fluorescence emission spectrum (Figure 5B; see Supplemental Figure 9 online). In N. benthamiana, both TRV:NbCPP1 and TRV:NbPOR1/2 VIGS leaves accumulated high levels of the POR substrate Pchlide, consistent with POR deficiency in those lines, while the Pchlide peaks were much smaller in the TRV and TRV:NbGyrB control lines (Figure 5B). Pchlide accumulation was also observed in Arabidopsis CPP1 RNAi seedlings upon DEX spraying, albeit weakly (see Supplemental Figure 9 online).

To determine whether CPP1 is involved in the import of POR, we performed chloroplast import assays (see Supplemental Figure 10 online). First, NbPOR1:GFP was expressed in leaves of TRV control and TRV:NbCPP1 VIGS N. benthamiana plants by agroinfiltration. Confocal laser scanning microscopy revealed that NbPOR1:GFP was targeted normally to the chloroplasts in both TRV and TRV:NbCPP1 lines, suggesting that Nb-CPP1 deficiency did not significantly affect the import of Nb-POR1 into chloroplasts (see Supplemental Figure 10A online). Next, PORC:Flag was expressed in the leaves of TRV and TRV:NbCPP1 plants by agroinfiltration. Immunoblotting of the leaf protein extracts using anti-Flag antibodies detected only the proteolytically processed, mature form of PORC, but not its precursor form, in both TRV and TRV:NbCPP1 lines, suggesting normal import of PORC into chloroplasts (see Supplemental Figure 10B online). Finally, the PORC:Flag construct was transformed into Arabidopsis protoplasts isolated from the At-CPP1 RNAi plants that were sprayed with ethanol (−DEX) or DEX for 6 d. Immunoblotting of the protoplast protein extracts using anti-Flag antibodies detected only the mature form of PORC in both (−)DEX and (+)DEX samples, while the total POR protein level was reduced (see Supplemental Figure 10C online). These results suggest that chloroplast import of PORC was unaffected by At-CPP1 silencing. Taken together, these results suggest that CPP1 is required for POR stability rather than for import of POR into chloroplasts.

Consistent with the protective function of CPP1 for POR, we found that POR proteins were susceptible to photooxidative stress (Figure 5E). Wild-type N. benthamiana plants in the eight-leaf stage grown under normal light conditions (60 μmol m⁻² s⁻¹) were transferred to high-light intensity (500 μmol m⁻² s⁻¹) and incubated for 0 to 72 h. The plants did not exhibit any visible symptoms until 48 h of treatment, but after 72 h showed a leaf yellowing phenotype, particularly in older leaves (Figure 5E). Immunoblotting revealed that Nb-POR protein accumulation progressively decreased until 48 h, despite the lack of leaf yellowing, while N. benthamiana CPP1, cpHSP70, atpB, and D1 protein levels remained constant throughout the period (Figure 5E). When immunoblotting was performed with green, pale-green, and yellow leaves collected from the plants at 72 h of treatment, CPP1, cpHSP70, and atpB protein levels were higher in green leaves than in pale-green or yellow leaves. POR protein accumulation was greatly reduced in yellow leaves at 72 h of high-light treatment (Figure 5E). However, we did not observe any increases in POR protein amounts in insoluble fractions, suggesting that POR proteins are degraded before being aggregated under high-light conditions. Interestingly, we repeatedly observed that D1 proteins disappeared in green leaves but were present in pale-green and yellow leaves at 72 h. This may be due to the fact that green leaves were located in the upper part of the plants and thus were exposed to higher light intensity. It has been reported that D1 proteins are easily degraded under high light conditions (Aro et al., 1993). Taken together, these results suggest that Nb-POR protein is sensitive to photooxidative stress.

CPP1 Deficiency Leads to Reduced Accumulation of POR Proteins in Etioplasts and Delayed Greening upon Illumination

To study the effects of CPP1 silencing during transition from dark to light, At-CPP1 RNAi seedlings were grown in the dark for 5 d and then transferred to light for 4 or 24 h on medium with or without DEX. When grown on medium with DEX, RNAi seedlings exhibited delayed greening and anthocyanin accumulation upon illumination (Figure 6A). Real-time quantitative RT-PCR shows that CPP1 was silenced in the DEX-inducible At-CPP1 RNAi seedlings upon DEX treatment during dark-to-light transition (Figure 6B, bottom). This analysis also showed that the CPP1 transcript level increased upon illumination (Figure 6B, bottom), consistent with an increase in CPP1 protein levels upon illumination in wild-type seedlings (Figure 6B, top). Immunoblotting revealed that illumination caused rapid degradation of POR proteins, accompanied by accumulation of CAB, D1, and rbcL in wild-type, (−)DEX, and (+)DEX RNAi seedlings (Figure 6C; see Supplemental Figure 11 online), consistent with the previous reports (Masuda and Takamiya, 2004; Reinbothe et al., 2010). Compared with (−)DEX and wild-type seedlings, dark-grown seedlings treated with DEX had less POR accumulation in etioplasts and delayed accumulation of the photosynthesis-related proteins upon illumination, consistent with the delayed greening phenotype. Transcription profiles demonstrated no effect of DEX on expression profiles of PORA, PORB, PORC, CHLH, and CHL27 (Figure 6D). PORA and PORB were downregulated by light, while PORC, CHLH, and CHL27 were induced by light.

Figure 6.

Analyses of the DEX-Inducible At-CPP1 RNAi Lines during the Transition from Dark to Light Conditions.

(A) Phenotypes of the seedlings. RNAi seedlings were dark grown or grown under dark conditions and then transferred to light (80 μmol m⁻² s⁻¹) for 4 or 24 h on medium with or without DEX.

(B) Immunoblotting and real-time quantitative RT-PCR. Immunoblotting was performed with wild-type (WT) seedlings during the transition from dark (D) to light conditions (for 4 and 24 h) (top). At-CPP1 mRNA levels in the RNAi seedlings upon DEX treatment during the dark-to-light transition were measured by real-time quantitative RT-PCR and compared with the level in (−)DEX samples grown in the dark (bottom). Asterisks denote the statistical significance of the differences between the samples. **P ≤ 0.01.

(C) Immunoblotting during the transition from dark to light conditions (for 4 and 24 h). Due to a short exposure, only light-induced POR degradation was shown.

(D) RT-PCR. At-CPP1 was silenced in DEX-treated RNAi seedlings without affecting the transcript profiles of PORA, PORB, PORC, CHLH, and CHL27. UBC10 was used as a control.

(E) TEM of ultrastructural changes of the cotyledon etioplasts during the dark-to-light transition. The arrow indicates degenerating plastids. P, prolamellar bodies. Bars = 1 μm in (a), (c), and (d) and 0.5 μm in (b), (e), (f), (g), and (h).

TEM observation revealed that the cotyledons of etiolated (−)DEX RNAi seedlings displayed normal etioplasts with extensive PLBs and unstacked prothylakoids (Figure 6E). The PLB disappeared or shrank after 4 h of illumination, followed by formation of grana stacks after 24 h. Etioplasts of DEX-treated RNAi seedlings had either smaller PLBs or no PLB, and both phenotypes were simultaneously observed in a single cotyledon cell. Approximately 48.4% of (+)DEX etioplasts lacked a PLB, while only 10% of (−)DEX etioplasts lacked a PLB (see Supplemental Figure 12A online). Analyses of TEM pictures of the cotyledon sections using Image J software demonstrated that the average ratio of PLB area to etioplast area was 0.47 in (−)DEX and 0.22 in (+)DEX etioplasts, suggesting reduced PLB size in CPP1-deficient etioplasts (see Supplemental Figure 12B online). Furthermore, the thylakoid systems in DEX-treated etioplasts developed poorly even after 24 h of illumination and many plastids were degenerating by that time (Figure 6E). Dark-grown Arabidopsis seedlings that have defects in POR accumulation in etioplasts due to downregulation of POR gene expression show defective PLB formation (Sperling et al., 1997, 1998; Franck et al., 2000). These results suggest that CPP1 modulates POR protein accumulation and thereby affects PLB formation in etioplasts. Thus, CPP1 appears to play a crucial role in regulating POR function during dark-to-light transition as well as in light-adapted conditions.

CPP1 Can Stabilize POR Proteins with Its Chaperone Activity

We next examined whether CPP1 has chaperone activity for POR by assessing its ability to prevent heat-induced denaturation and aggregation of POR proteins in vitro. To use NbPOR1:His as a substrate for chaperone assay, we first demonstrated that purified NbPOR1:His and Nb-POR1 proteins without a His tag (prepared by treatment with tobacco etch virus [TEV] protease) exhibited a similar pattern of protein aggregation in response to heat, and aggregation of Nb-POR proteins (without tag) was almost completely inhibited by MBP:NbCPP1 (see Supplemental Figure 13A online). Addition of purified MBP:NbCPP1 proteins suppressed thermal aggregation of NbPOR1:His in a concentration-dependent manner: A molar ratio of 2:1 (chaperone to substrate) completely suppressed aggregation (Figure 7A), whereas addition of MBP, glutathione S-transferase (GST), or MBP:NbRae1 fusion protein as a control had no effect (Figure 7F; see Supplemental Figure 13B online). Nb-Rae1 is a nuclear envelope protein of N. benthamiana. MBP:slr1918 also exhibited chaperone activity for NbPOR1:His, albeit weakly (Figure 7B). Nb-CPP1 lacking the JL domain (MBP:NbCPP1ΔJL) failed to prevent NbPOR1:His aggregation, indicating that the JL domain is required for chaperone activity (Figure 7C). However, the slr1918 JL domain alone [20(Met)-81(Lys); slr1918-JL] exhibited no chaperone activity for NbPOR1:His, suggesting that other regions of CPP1 are required for the activity (Figure 7G). These regions may play a role in maintaining the proper conformation of CPP1 for its chaperone activity in vitro. The transmembrane domain deletion mutants of Nb-CPP1 and slr1918 could not be expressed in E. coli. GST-fused N. benthamiana chloroplast HSP70 (Schroda et al., 1999; Su and Li, 2010) (GST:NbcpHSP70) also exhibited chaperone activity for NbPOR1:His (Figure 7D), but there was no significant synergism when MBP:NbCPP1 and GST:NbcpHSP70 were combined in the presence or absence of ATP (Figure 7E; see Supplemental Figure 13C online). Thermal aggregation of NbPOR2:His was also suppressed by MBP:NbCPP1 (see Supplemental Figures 13D and 13E online). Taken together, these results indicate that CPP1 has holdase chaperone activity for POR proteins and that both the JL domain and the transmembrane domains are required for full activity.

Figure 7.

Chaperone Activity of CPP1 for POR.

Absorbance at 340 nm (A₃₄₀) was measured at 2-min intervals to quantify turbidity. Thermal aggregation of NbPOR1:His (1.6 μM) was examined for 20 min at 43°C with increasing amounts of MBP:NbCPP1 (A), MBP:slr1918 (the CPP1 homolog of Synechocystis sp PCC6803) (B), MBP:NbCPP1∆JL (C), and GST:NbcpHSP70 (D). In addition, indicated amounts of MBP:NbCPP1 were combined with GST:NbcpHSP70 (0.16 μM) (E). MBP (1.6 μM) or GST (1.6 μM) alone did not show any chaperone activity for NbPOR1:His (F). Thermal aggregation of NbPOR1:His (1.6 μM) was examined for 20 min at 43°C with MBP:slr1918 or with different concentrations of slr1918-JL (without tag), which contains only the JL domain (G). Aggregation of NbPOR1:His (1.6 μM) in response to 1 mM H₂O₂ was examined for 20 min at 30°C with different amounts of MBP:NbCPP1 (H) and MBP:slr1918 (I). Addition of water (DW) instead of 1 mM H₂O₂ did not result in aggregation of NbPOR1:His.

[See online article for color version of this figure.]

We next tested whether CPP1 can prevent POR denaturation/aggregation induced by oxidative stress in vitro. NbPOR1:His proteins were aggregated at 30°C upon exposure to 1 mM hydrogen peroxide (H₂O₂) but not to water as a control, and addition of purified MBP:NbCPP1 proteins suppressed aggregation of NbPOR1:His in a concentration-dependent manner (Figure 7H). Addition of MBP:slr1918 also protected NbPOR1:His from H₂O₂-induced denaturation and aggregation (Figure 7I). Addition of GST as a control did not affect aggregation of NbPOR1:His in response to H₂O₂ (see Supplemental Figure 13F online). In addition, real-time quantitative RT-PCR analysis suggested that Nb-CPP1 mRNAs in leaves of N. benthamiana plants (eight-leaf stage) were ∼5.0- and 1.9-fold more abundant than Nb-POR1 and Nb-POR2 mRNAs, respectively (see Supplemental Figure 14 online). Collectively, these results support our proposal that CPP1 stabilizes POR proteins against photooxidative stress in chloroplasts.

Conserved CPP1 Functions in Cyanobacteria

In cyanobacteria, Pchlide reduction in the dark exclusively depends on the light-independent DPOR (Muraki et al., 2010), but the contribution of POR increases with light intensity; under high intensity, Pchlide is exclusively reduced by POR (Fujita et al., 1998; Yamazaki et al., 2006). Cyanobacterial POR also seems to function as a scavenger for Pchlide, which can cause photooxidative damage upon illumination (Fujita et al., 1998). Synechocystis sp PCC6803 has a single CPP1 homolog, slr1918 (see Supplemental Figure 1C online). Wild-type and slr1918 knockout mutant Synechocystis strains grown under dim light (10 μmol m⁻² s⁻¹) were transferred to three different light conditions and incubated for 1 to 3 d under photoautotrophic conditions. Under low-light intensity (20 μmol m⁻² s⁻¹), both the wild type and slr1918 mutant strain grew well and synthesized chlorophyll normally (Figure 8A). However, under higher light intensity (100 and 150 μmol m⁻² s⁻¹), cell growth and chlorophyll synthesis were inhibited more strongly in the mutant than in the wild type. Immunoblotting revealed that the wild type and slr1918 mutant strains showed similar patterns of light-induced accumulation of POR proteins under low-light intensity (Figure 8B). However, under high-light intensity, much less POR protein accumulated in the mutant than in the wild type. Therefore, the POR-stabilizing function of slr1918 may be required for Pchlide reduction under high-light conditions when DPOR activity is suppressed by high oxygen evolution (Fujita et al., 1998; Masuda and Takamiya, 2004; Yamazaki et al., 2006). By contrast, protein expression patterns of atpB, GluTR, rbcL, D1, and HSP70 were similar in both strains under different light conditions.

Figure 8.

Conserved Function of slr1918, the CPP1 Homolog of Synechocystis sp PCC6803, for POR Protein Accumulation and Chlorophyll Synthesis.

(A) The wild-type (wt) and slr1918 knockout mutant (m) strains were grown under dim light (10 μmol m⁻² s⁻¹) and transferred to light with higher intensity for 3 d.

(B) The wild-type and slr1918 mutant strains were incubated under different light intensity for 1 to 3 d before being subjected to immunoblotting. Equal amounts of proteins were loaded in each lane.

DISCUSSION

In this study, we identified Cdf1/CPP1 as a critical regulator of POR stability in angiosperms and cyanobacteria. Most of the chlorophyll intermediates, including Pchlide, are strong photosensitizers that can produce ROS upon photoexcitation, causing oxidative damage and cell death when present in excess (Ledford and Niyogi, 2005; Li et al., 2009). Photoexcitation of Pchlide molecules produces singlet oxygen, H₂O₂, superoxide, and hydroxyl radical (Meskauskiene et al., 2001; Hideg et al., 2010). Oxidative stresses cause oxidation of amino acid residues, particularly Met and Cys, leading to protein destabilization and denaturation (Berlett and Stadtman, 1997; Stadtman, 2006). It is known that the POR enzyme is activated by light and that its substrate and product are photosensitizers, but it remains unclear how POR is protected from photooxidative damage. We demonstrated in this study that CPP1 deficiency results in diminished POR protein accumulation and defective chlorophyll synthesis and that CPP1 has holdase chaperone activity for POR proteins in vitro in response to heat or H₂O₂. We also demonstrated that POR proteins are indeed susceptible to photooxidative stress in vivo. These data suggest that CPP1 may regulate POR stability in chloroplasts by preventing denaturation and aggregation of POR with its chaperone activity. In the absence of CPP1, POR may be susceptible to denaturation due to photooxidation, leading to its degradation by chloroplast proteases. Depletion of POR may then result in accumulation of its substrate Pchlide, the photoexcitation of which causes excessive accumulation of ROS. Being localized in chloroplast membranes with multiple transmembrane domains, CPP1 may also contribute to anchoring of POR proteins on the thylakoid and envelope membranes through protein–protein interactions.

There are quite a few examples of protective mechanisms against photooxidative stress in chloroplasts. Chloroplast Hsp70 chaperone, its partner DnaJ/Hsp40, HSP21, DegP, and Filamentous temperature sensitive H proteases, and auxiliary proteins, such as LPA1/REP27 and TLP18.3, play roles in photoprotection and repair of photosystem II (Mulo et al., 2008; Nordhues et al., 2010). NADPH-dependent thioredoxin reductase and 2-Cys peroxiredoxins, which constitute a chloroplast H₂O₂-scavenging system, are involved in the protection of CHL27 (Stenbaek et al., 2008). Interestingly, LIL3, a light-harvesting-like protein, interacts with geranylgeranyl reductase (GGR) and is required for stable accumulation of GGR for chlorophyll biosynthesis, suggesting that LIL3 stabilizes GGR in plastid membranes (Tanaka et al., 2010). LIL3 was also proposed to function as a membrane anchor of GGR that lacks predictable transmembrane regions. In addition, CP12, a small protein present in the chloroplasts of most photosynthetic organisms, forms a stable complex with GAPDH and functions as its specific chaperone-like protein to prevent aggregation and inactivation of the enzyme (Erales et al., 2009). CPP1 may represent another example of a specific protein stabilization mechanism in chloroplasts. Unlike most chaperones that transiently bind to denatured proteins, CPP1 appears to form an interaction with the native form of POR isoforms. Thus, similar to the function of CP12 for GAPDH, CPP1 may closely associate with POR and stabilize POR protein structure using its chaperone activity in ROS-rich environments. Although we mainly tested the relationship between CPP1 and POR enzymes in this study, there is a possibility that CPP1 has other substrates/interaction partners within chloroplast, which will be a subject of our future study. In addition, we cannot exclude the possibility that CPP1 may be involved in normal maturation and activation of POR in addition to stabilizing the protein against oxidative stresses.

Interestingly, CPP1 deficiency resulted in defective greening of the etiolated seedlings upon illumination in addition to its effect in light-adapted plants. CPP1 depletion in etiolated seedlings led to reduced POR protein levels, reduced PLB formation, and delayed photomorphogenesis. These phenotypes are reminiscent of the phenotypes of POR-deficient plants generated by various means. When grown under continuous far-red light, Arabidopsis wild-type etiolated seedlings are depleted of POR proteins, accompanied by the loss of PLB and they fail to green upon transfer to white light due to photooxidative damage (Sperling et al., 1997). The Arabidopsis cop1 photomorphogenetic mutant also lacks PORA and PLB in etiolated seedlings, and the susceptibility of the mutant to photooxidative damage can be rescued by overexpression of either PORA or PORB (Sperling et al., 1998). Defective PLB formation is also observed in dark-grown Arabidopsis POR antisense seedlings, in which both PORA and PORB are downregulated (Franck et al., 2000). These results demonstrated that the total POR protein levels closely correlate with PLB formation in etioplasts and the subsequent light-induced chloroplast development. Our results show that CPP1 is required for POR accumulation and PLB formation in etioplasts, and its deficiency disrupts biogenesis of thylakoid membranes and chloroplasts upon illumination. Within etioplasts, a ternary complex of POR-NADPH-Pchlide is the main protein component of PLB that displays a regular lattice-like membrane structure (Masuda and Takamiya, 2004; Reinbothe et al., 2010). The proteomics during the transition of etioplasts to chloroplasts shows that a battery of proteins involved in antioxidation and detoxification, such as ascorbate peroxidases, glutathione reductase, and PsbS, are highly accumulated in etioplasts, suggesting the importance of defense against oxidative stress in etioplasts (Kanervo et al., 2008). Currently, it remains unclear how CPP1 contributes to POR accumulation in etioplasts. There is a possibility that CPP1 plays a role in POR protection in etioplasts from oxidative stress, similar to its role in light-adapted plants. Alternatively, CPP1 may play a role in the assembly and/or stability of the POR complex in etioplasts. Further study will be required to determine how CPP1 modulates POR accumulation in etioplasts.

POR catalyzes sequential hydride and proton transfers in the photoexcited and ground states, respectively (Menon et al., 2009; Heyes et al., 2011). The conserved Tyr and Lys residues in the active site of POR are involved in the formation of ternary enzyme-substrate complexes. Furthermore, the Tyr residue stabilizes the Pchlide excited state, promoting hydride transfer from NADPH to Pchlide, and participates in proton transfer (Menon et al., 2009). It has been generally thought that POR evolved from cyanobacteria, and the evolution of POR reflects the transformation from anoxygenic to oxygenic photosynthesis (Reinbothe et al., 1996, 2010; Masuda and Takamiya, 2004). When the catalytic mechanism was compared between cyanobacterial and plant POR, the light-driven hydride transfer step was found to be conserved among all POR enzymes, but the proton transfer step is variable, suggesting that functional adaptation and optimization occurred during the evolution of POR (Heyes et al., 2011). CPP1 homologs are ubiquitously found in oxygenic photosynthetic organisms that possess POR, including cyanobacteria, green algae, mosses, gymnosperms, and angiosperms. These CPP homologs possess the conserved protein structure that consists of a JL domain and multiple transmembrane domains. We demonstrated that the cyanobacterial CPP1 homolog (slr1918) critically regulates POR stability and chlorophyll synthesis in Synechocystis under high-light conditions when oxygen-sensitive DPOR activity is repressed. The results suggest that the chaperone-like function of CPP1 is required regardless of the differences in catalytic structure between cyanobacterial and eukaryotic POR. The conserved CPP1 function in cyanobacteria and the ubiquitous presence of CPP1 in oxygenic photosynthetic organisms suggest the essential role of CPP1 for POR function in chlorophyll synthesis.

METHODS

Plant Material and Growth Conditions

Arabidopsis thaliana (ecotype Columbia-0) and Nicotiana benthamiana plants were grown in a growth room at 23°C under a 16-h-light/8-h-dark cycle. The T-DNA insertion mutant of Arabidopsis CPP1 (GK-852F12-025757) was obtained from GABI-Kat (www.gabi-kat.de/).

Cyanobacterial Strains and Culture Conditions

The motile Synechocystis sp PCC 6803 wild-type strain and the slr1918 knockout mutant strain were grown on 1% BlueGreen-11 agar plates buffered with 10 mM TES-potassium hydroxide (KOH), pH 8.0, at 28°C under continuous white light (10 μmol m⁻² s⁻¹) as described (Song et al., 2011), before exposure to increased irradiance (20, 100, and 150 μmol m⁻² s⁻¹) for 3 d. Chloramphenicol (20 μg mL⁻¹) was added to the growth medium for the slr1918 mutant that contains a gene conferring chloramphenicol resistance.

Phylogenetic Analysis

The CPP1 homologous sequences from N. benthamiana, grape (Vitis vinifera), soybean (Glycine max), Populus trichocarpa, Arabidopsis, rice (Oryza sativa), maize (Zea mays), Physcomitrella patens, wheat (Triticum aestivum), Synechocystis sp PCC 6803, Chlamydomonas reinhardtii, and Picea sitchensis were aligned using the ClustalW program (http://www.ch.embnet.org/software/ClustalW.html). Mega phylogenetic trees for these data were analyzed using the MEGA 5.2.2 program (http://www.megasoftware.net/). The maximum likelihood tree was generated with the following parameters: Poisson model, uniform rates, complete deletion, and bootstrap (2000 replicates).

Gene Cloning

The initial VIGS screening was performed with a partial Nb-CPP1 cDNA of 726 bp. To obtain the full-length Nb-CPP1 cDNA, 5′-rapid amplification of cDNA ends (RACE) PCR was performed with mRNAs isolated from N. benthamiana seedlings using the SMART RACE cDNA amplification kit (BD Biosciences) according to the manufacturer’s manual. The gene-specific primers used for 5′-RACE PCR were NbCPP1-Gsp(1)/(2). Lists of primers and accession numbers of the genes used in this study are shown in Supplemental Tables 1 and 2 online, respectively. Information on various constructs and cloning methods is available in Supplemental Table 3 online.

VIGS in N. benthamiana

A 726-bp Nb-CPP1 cDNA fragment was amplified by PCR and cloned into the pTV00 vector containing part of the TRV genome, using the BamHI and ApaI sites. The Nb-POR1 and Nb-POR2 cDNA fragments were amplified by PCR with NbPOR1-VIGS(F)/(R) and NbPOR2-VIGS(F)/(R) primers, respectively. To achieve cosilencing of Nb-POR1 and Nb-POR2, the PCR product of Nb-POR1 (402 bp) was first cloned into pTV00 vector using the BamHI and ApaI sites, followed by the cloning of the Nb-POR2 fragment (400 bp) into the same plasmid using BamHI and SpeI sites. VIGS was performed as described (Cho et al., 2004; Ahn et al., 2011; Jeon et al., 2012). All primers used for cloning are listed in Supplemental Table 1 online.

Generation of Arabidopsis DEX-Inducible At-CPP1 RNAi Lines

A 400-bp At-CPP1 cDNA fragment was amplified by PCR using AtCPP1-sense(F)/(R) primers containing XhoI and ClaI sites for the sense construct and AtCPP1-antisense(F)/(R) primers containing SpeI and EcoRI sites for the antisense construct. Using these constructs, DEX-inducible At-CPP1 RNAi transgenic Arabidopsis lines were generated as described (Ahn et al., 2011). For induction of RNAi, the transgenic seedlings were grown on medium containing 10 μM DEX in ethanol (0.033%). Alternatively, the RNAi seedlings were sprayed with 30 μM DEX in ethanol (0.033%) and Tween 20 (0.01% [w/v]).

Agrobacterium tumefaciens–Mediated Transient Expression

Agroinfiltration was performed as described (Voinnet et al., 2003). Agrobacterium C58C1 cultures containing expression constructs were adjusted to OD₆₀₀ = 0.6 in MES buffer (10 mM MES, pH 7.5, and 10 mM MgSO₄). The suspension was incubated with acetosyringone for 2 to 3 h at a final concentration of 150 μM and infiltrated into leaves of N. benthamiana plants. In all infiltration experiments, Agrobacterium C58C1 carrying the 35S-p19 construct was coinfiltrated to achieve maximum levels of protein expression as described (Voinnet et al., 2003). Expressed proteins were analyzed 24 to 48 h after infiltration.

RT-PCR Analyses

RT-PCR was performed with total RNA isolated from leaves of VIGS and RNAi lines with 15 to 35 cycles of amplification as described (Ahn et al., 2011). To measure the transcript levels in N. benthamiana, the following primers were used: CPP1, NbCPP1(F)/(R); POR1, NbPOR1(F)/(R); POR2, NbPOR2(F)/(R); and actin, Nbactin(F)/(R). To measure the transcript levels in Arabidopsis, the following primers were used: CPP1, AtCPP1(F)/(R); PORA, PORA(F)/(R); PORB, PORB(F)/(R); PORC, PORC(F)/(R); CHLH, CHLH(F)/(R); CHL27, CHL27(F)/(R); and UBC10, UBC10(F)/(R).

Real-Time Quantitative RT-PCR

Real-time quantitative PCR was performed as described (Ahn et al., 2011) using the following primers: Nb-CPP1, NbCPP1-qRT-PCR(F)/(R); Nb-β-tubulin, β-tubulin-qRT-PCR(F)/(R); At-CPP1, AtCPP1-qRT-PCR(F)/(R); and At-UBC10, UBC10-qRT-PCR(F)/(R).

Analyses of Chlorophyll Precursors

Chlorophyll precursors were extracted from N. benthamiana VIGS lines as described (Terry and Kendrick, 1999). Spectrofluorometry was performed using a fluorescence spectrophotometer (Hitachi F-2000) at an excitation wavelength of 440 nm and an emission wavelength of 600 to 700 nm as described (Terry and Kendrick, 1999).

Measurement of Chlorophyll Contents

Yellow sectors of the fourth leaf above the infiltrated leaf were collected from the VIGS plants. Chlorophyll concentration per unit fresh weight was calculated as previously described (Pattanayak et al., 2005).

TEM

Tissue sectioning and TEM were performed as described (Jeon et al., 2012), using the fourth or fifth leaf above the infiltrated leaf of the N. benthamiana VIGS lines 20 d after infiltration and the cotyledons of the etiolated Arabidopsis At-CPP1 RNAi seedlings grown on medium with or without DEX for 5 d.

NBT Staining

To detect the production of superoxide radicals in situ, NBT staining was performed as described (Bielski et al., 1980; Wohlgemuth et al., 2002).

Preparation of Chloroplast Fractions

From chloroplasts isolated from N. benthamiana leaves, stroma, envelope, and thylakoid membranes were fractionated as described (Kwon and Cho, 2008).

Subcellular Localization and BiFC

cDNAs of Nb-CPP1, Nb-POR1, and At-CPP1 corresponding to the full-length protein were cloned into the pCambia-GFP plasmid to generate GFP fusion proteins. EcoRI and NcoI sites were used for Nb-CPP1 and Nb-POR1 cDNAs, and BamHI and BglII sites were used for the At-CPP1 cDNA. The recombinant plasmids or the control vector were introduced into N. benthamiana leaves by agroinfiltration. Expression of the GFP fusion proteins was monitored 24 h after transformation using a confocal laser scanning microscope (Zeiss LSM510) as previously described (Cho et al., 2004). For visualization of mitochondria, TMRM staining was performed, and TMRM fluorescence was detected by confocal laser scanning microscopy as described (Cho et al., 2004). For BiFC, the coding regions of Nb-CPP1 and At-CPP1 were amplified by PCR and cloned into the pSPYNE vector containing the N-terminal region (amino acid residues 1 to 155) of the YFP, resulting in pSPYNE-NbCPP1 and pSPYNE-AtCPP1, respectively. N. benthamiana POR1, POR2, PORA, PORB, and PORC cDNAs were cloned into pSPYCE vector containing the C-terminal region of YFP (residues 156 to 239). Agrobacteria containing the pSPYNE and pSPYCE fusion constructs were agroinfiltrated together into leaves of 3-week-old N. benthamiana plants as described (Walter et al., 2004). After 48 h, leaf discs or protoplasts were generated from the leaves, and the YFP signal was detected by confocal laser scanning microscopy and fluorescent microscopy.

Coimmunoprecipitation

NbCPP1:Flag and NbPOR1:HA or NbCPP1:Flag and NbPOR2:HA fusion proteins were coexpressed in N. benthamiana leaves by agroinfiltration. In addition, NbCPP1:Flag and NbGUN4:HA or NbCPP1:Flag and NbCHLI2:HA fusion proteins were coexpressed by agroinfiltration as a control. After 48 h of incubation, chloroplast membranes were prepared from the leaves and solubilized using the solubilization buffer containing n-dodesyl β-d-maltoside as described (Sun et al., 2010). The solubilized membrane fraction was incubated with 2 μg of anti-HA antibody (Applied Biological Materials) for 4 h at 4°C, followed by addition of 10 μL of Protein A-Sepharose beads (Fast Flow; Amersham Pharmacia) for a further 3-h incubation. After washing, the sepharose beads were resuspended in 2× SDS sample buffer and boiled, and the resin was precipitated by brief centrifugation. The supernatant was subjected to SDS-PAGE and immunoblotting using the monoclonal anti-Flag antibody (1:10,000; Sigma-Aldrich) and monoclonal anti-HA antibody (1:5000; Applied Biological Materials).

Purification of Recombinant Proteins

To purify MBP:NbCPP1 and MBP:NbCPP1ΔJL, the Nb-CPP1 cDNA fragments corresponding to the amino acid residues 61(Trp)-252(Lys) and 112(Lys)-252(Lys), respectively, were amplified by PCR and cloned into pMAL c2 vector (New England Biolabs). To purify MBP:slr1918, the full-length coding region of the slr1918 gene was cloned into pMAL c2 vector. To purify GST:NbcpHSP70, the Nb-cpHSP70 cDNA fragment corresponding to the amino acid residues 62(Pro)-707(Lys) was cloned into pGEX-4T-1 vector (Amersham Pharmacia). The MBP and GST fusion proteins were purified following the manufacturer’s instructions using MBP Excellose resin and Glutathione Excellose resin, respectively (Bioprogen). To purify NbPOR1:His and NbPOR2:His, the Nb-POR1 and Nb-POR2 cDNA fragments corresponding to the amino acid residues 63(Ala)-396(Leu) and 66(Ile)-399(Ala), respectively, were amplified by PCR and cloned into the pET-29a vector (Novagen). The 6× His fusion proteins were purified using Nickel-nitrilotriacetic acid agarose (Qiagen) following the manufacturer’s instructions. To purify Nb-POR1 without epitope tag, the Nb-POR1 cDNA fragment described above was cloned into a modified pET-28a vector carrying six N-terminal His residues and a TEV cleavage site (Ko et al., 2008). After purification of the protein using Nickel-nitrilotriacetic acid agarose (Qiagen), His tag was removed by TEV protease as described (Ko et al., 2008). To purify slr1918-JL without tag, the slr1918 gene fragment corresponding to the amino acid residues 20(Met)-81(Lys) was cloned into the modified pET-28a vector, and protein purification and His tag removal were performed as described above.

In Vitro Binding Assay

MBP, MBP:NbCPP1, and MBP:slr1918 proteins immobilized on MBP Excellose resin (Takara) were incubated with Escherichia coli lysates expressing NbPOR1:His or NbGUN4:His in buffer (50 mM HEPES, pH 7.5, 200 mM NaCl, and 10% glycerol) for 2 h at room temperature. After extensive washing of the resin, bound proteins were eluted with SDS sample buffer, and 25% of the eluted proteins were subjected to immunoblotting.

Immunoblot Analyses

Protein extracts (30 µg) were subjected to SDS-PAGE and immunoblotting as described (Ahn et al., 2011; Jeon et al., 2012). Anti-CPP1 antibodies were generated in rabbits against an oligopeptide, ASEEEIWASRNFLL, that corresponds to amino acid residues of At-CPP1 at positions 79 to 92 using the antibody production services of Cosmogenetech. Immunoblotting was performed using mouse monoclonal antibodies against the HA tag (1:10,000 dilution; Applied Biological Materials), the His tag (1:5000 dilution; Applied Biological Materials), and the Flag tag (1:10,000; Sigma-Aldrich), using rabbit polyclonal antibodies against CPP1 (1:2000; Cosmogenetech), using rabbit polyclonal antibodies against POR, rbcL, D1, atpB, psaA, cpHsp70, and GluTR (1:5000, 1:10,000, 1:5000, 1:10,000, 1:5000, 1:10,000, and 1:5000 dilution, respectively; Agrisera), or using goat polyclonal antibodies against CAB (1:5000; Santa Cruz Biotechnology). The membranes were then treated with horseradish peroxidase–conjugated goat anti-mouse IgG antibodies (1:10,000; Invitrogen), goat anti-rabbit antibodies (1:10,000; Invitrogen), or donkey anti-goat antibodies (1:10,000; Santa Cruz Biotechnology), respectively. Signals were detected on x-ray film (Kodak) using an ECL chemiluminescence kit (ELPIS-Biotech).

POR Import into Chloroplasts

POR import into chloroplasts was analyzed as described (Lee et al., 2008). Flag-tagged PORC proteins were transiently expressed in leaves of TRV control and TRV:NbCPP1 VIGS lines. After 16 h of incubation, total proteins were extracted from the leaves and subjected to SDS-PAGE and immunoblotting with anti-Flag antibody (1:10,000; Sigma-Aldrich) to detect expressed PORC proteins. In vitro transcription and translation of the precursor form of PORC was performed using a wheat germ extract system (Promega) according to the manufacturer’s instructions.

Measurement of Chaperone Activity

NbPOR1:His (1.6 μM) was incubated in 50 mM HEPES-KOH, pH 8.0, buffer at 43°C with various concentrations of MBP:NbCPP1, MBP:NbCPP1∆JL, MBP:slr1918, slr1918∆TM, or GST:NbcpHSP70. In addition, NbPOR1:His (1.6 μM) was incubated with both MBP:NbCPP1 (at various concentrations) and GST:NbcpHSP70 (0.16 μM) to observe any synergistic effects on chaperone activity. NbPOR2:His (1.6 μM) was incubated in 50 mM HEPES-KOH, pH 8.0, buffer at 45°C with various concentrations of MBP:NbCPP1. Aggregation of NbPOR1:His (1.6 μM) in response to 1 mM H₂O₂ or water was examined for 20 min at 30°C with different amounts of MBP:NbCPP1 and MBP:slr1918. Aggregation of NbPOR1:His or NbPOR2:His induced by heat or H₂O₂ was determined by monitoring the turbidity increase (A₃₄₀) at 2-min intervals using a temperature-controlled spectrophotometer (DU800; Beckman) as described (Park et al., 2009).

Generation of the slr1918 Knockout Mutant Strain of Cyanobacteria

DNA constructs for generating the slr1918 knockout mutant were prepared using a fusion PCR cloning strategy (Wang et al., 2002). Briefly, fragment A including the slr1918 5′-upstream flanking region and the slr1918 start codon was amplified by PCR with slr1918-fA(F)/(R) primers using genomic DNA as template. Fragment B, including the slr1918 stop codon and the 3′-downstream region, was amplified by PCR using slr1918-fB(F)/(R) primers. The overlapping ends of fragment A, the chloramphenicol resistance cassette, and fragment B were annealed at 55°C for three-piece PCR fusion as described (Lee et al., 2007). The amplified DNA fragment was then ligated into the blunt-end cloning vector TOP Cloner (Enzynomics). The recombinant plasmid was transformed into Synechocystis wild-type strains as described previously (Lee et al., 2007).

Statistical Analyses

Two-tailed Student’s t tests were performed using the Minitab 16 program to determine the statistical differences between the samples.

Accession Numbers

Accession numbers for the genes used in this study are shown in Supplemental Table 2 online.

Supplemental Data

The following materials are available in the online version of this article.

• Supplemental Figure 1. Protein Structure, Phylogenetic Tree, and Sequence Alignment of CPP1.

• Supplemental Figure 2. Computational Modeling of the CPP1 J-Like Domain.

• Supplemental Figure 3. An Embryo-Lethal Phenotype of the Arabidopsis cpp1 Null Mutant.

• Supplemental Figure 4. Chloroplast Defects in TRV:NbCPP1 VIGS Lines.

• Supplemental Figure 5. Light Sensitivity of the DEX-Inducible At-CPP1 RNAi Seedlings.

• Supplemental Figure 6. Bimolecular Fluorescence Complementation Analyses.

• Supplemental Figure 7. Control Coimmunoprecipitation Experiments.

• Supplemental Figure 8. Stability of POR in Different Arabidopsis DEX-Inducible At-CPP1 RNAi Lines.

• Supplemental Figure 9. Fluorescence Emission Spectrum Showing Relative Fluorescence of Pchlide (Fluorescence Emission Maximum at 636 nm).

• Supplemental Figure 10. Analyses of POR Import into Chloroplasts.

• Supplemental Figure 11. POR Protein Levels during Dark-to-Light Transition in the Wild-Type Arabidopsis Plants.

• Supplemental Figure 12. Quantification of Prolamellar Bodies in Etioplasts of DEX-Inducible At-CPP1 RNAi Seedlings.

• Supplemental Figure 13. Control Experiments for Chaperone Assay.

• Supplemental Figure 14. Relative mRNA Levels of Nb-CPP1, Nb-POR1, and Nb-POR2.

• Supplemental Table 1. List of Primers Used in This Study.

• Supplemental Table 2. Accession Numbers.

• Supplemental Table 3. Constructs.

• Supplemental Data Set 1. Sequence Alignments Used for the Phylogenetic Analysis Shown in Supplemental Figure 1B.

Glossary

GAPDH

glyceraldehyde-3-phosphate dehydrogenase

Pchlide

protochlorophyllide

POR

Pchlide oxidoreductase

DPOR

dark-operative Pchlide oxidoreductase

PLB

prolamellar body

VIGS

virus-induced gene silencing

JL

J-like

DEX

dexamethasone

RNAi

RNA interference

TEM

transmission electron microscopy

NBT

nitro blue tetrazolium

GFP

green fluorescent protein

TMRM

tetramethyl rhodamine methyl ester

BiFC

bimolecular fluorescence complementation

YFP

yellow fluorescent protein

HA

hemagglutinin

GST

glutathione S-transferase

H₂O₂

hydrogen peroxide

ROS

reactive oxygen species

GGR

geranylgeranyl reductase

RACE

rapid amplification of cDNA ends

TRV

tobacco rattle virus

TEV

to be defined