PALE-GREEN LEAF 1, a rice cpSRP54 protein, is essential for the assembly of the PSI-LHCI supercomplex.

Although photosynthetic multiprotein complexes have received major attention, our knowledge about the assembly of these proteins into functional complexes in plants is still limited. In the present study, we have identified a chlorophyll-deficient mutant, pale-green leaf 1 (pgl1), in rice that displays abnormally developed chloroplasts. Map-based cloning of this gene revealed that OsPGL1 encodes a chloroplast targeted protein homologous to the 54-kDa subunit of the signal recognition particle (cpSRP54). Immunoblot analysis revealed that the accumulation of the PSI core proteins PsaA and PsaB, subunits from the ATP synthase, cytochrome, and light-harvesting complex (LHC) is dramatically reduced in pgl1. Blue native gel analysis of thylakoid membrane proteins showed the existence of an extra band in the pgl1 mutant, which located between the dimeric PSII/PSI-LHCI and the monomeric PSII. Immunodetection after 2D separation indicated that the extra band consists of the proteins from the PSI core complex. Measurements of chlorophyll fluorescence at 77 K further confirmed that PSI, rather than PSII, was primarily impaired in the pgl1 mutant. These results suggest that OsPGL1 might act as a molecular chaperone that is required for the efficient assembly and specific integration of the peripheral LHCI proteins into the PSI core complex in rice.

INTRODUCTION

Photosynthesis is a fundamental and critical process for plants to produce energy and maintain cellular functions. In higher plants, oxygenic photosynthesis is accomplished by an array of light‐driven reactions that occurs predominantly within the thylakoid membranes of the chloroplast, which converts light energy into chemical energy (ATP and NADPH) to sustain all life forms on earth by providing food and oxygen. The proteins involved in the photosynthetic light reactions can be classified into four major protein complexes, which are in order of photosystem II (PS II), the cytochrome b f (Cytb6f), photosystem I (PSI), and ATP synthase (F‐ATPase) (Nelson & Ben‐Shem, 2004). The initial and fundamental steps of oxygenic photosynthesis are catalyzed by the two photosystems, PSI and PSII (Albertsson, 2001; Dekker & Boekema, 2005). PSII is localized in the stacked grana compartment of thylakoids and is responsible for the oxidation of water and subsequent evolution of oxygen, while PSI functions as a light‐driven plastocyanin‐ferredoxin oxidoreductase that is mainly located in the stromal lamella of thylakoids (McCarty et al., 2000). In addition, other membrane protein complexes, such as light‐harvesting proteins (LHCI for PSI and LHCII for PSII), some soluble proteins, and cofactors, are associated with the two photosystems and mediate electron transport (Nelson & Yocum, 2006). Remarkable progress has been achieved in determining the composition and structure of these thylakoid protein complexes (Albanese et al., 2020; Kawashima et al., 2018; Sheng et al., 2019; Su et al., 2017; Suga et al., 2019; van Bezouwen et al., 2017; Wei et al., 2016; Zhao et al., 2020). The assembly of these complexes requires a host of protein–protein interactions and inter‐organelle trafficking; however, much of which is still poorly understood.

PSI is one of the largest multiprotein machineries known to reside in biological membranes (Amunts & Nelson, 2008; Chitnis, 2001; Nelson & Ben‐Shem, 2004; Schöttler et al., 2011). In higher plants, PSI is composed of a reaction center (RC; also known as core complex) and the peripheral antenna system responsible for light harvesting (LHCI), which is also called as PSI‐LHCI supercomplex (Nelson & Ben‐Shem, 2004; Saenger et al., 2002). The RC complex comprises at least 15 protein subunits, named alphabetically from PSI‐A to PSI‐P, except the PSI‐M, which is exclusively present in cyanobacteria (Chitnis, 1996, 2001; Inoue et al., 2004). Of these, two large homologous subunits, PsaA and PsaB, form the central heterodimer that binds the primary electron donor, P700, and the intermediate electron acceptors, A0, A1, and FX, and therefore constitute the inner core of PSI (Boudreau et al., 1997; Dühring et al., 2007; Nelson & Ben‐Shem, 2004). The peripheral antenna complex, LHCI, consists of four light‐harvesting chlorophyll (Chl)‐containing proteins (Lhca1–Lhca4). These proteins are assembled into two dimers that form a crescent‐shaped belt which binds tightly to the RC mainly through the interactions with the PsaG subunit (Ben‐Shem et al., 2003; Nelson & Ben‐Shem, 2004). PSI also harbors a large complement of pigments and cofactors such as Chl molecules, carotenes, phylloquinones, luteins, violaxanthins, and ion sulfur clusters (Amunts et al., 2010; Ganeteg et al., 2004). These molecules are associated with the core complex and LHCI proteins and contribute to light absorption and energy transfer (Golbeck, 2003; Saenger et al., 2002). Thus, it is critical that a large number of subunits must be integrated properly into a functional PSI‐LHCI supercomplex.

Successive integration of PSI subunits into a functionally active complex is assisted by many protein factors that may act as molecular chaperones (Schöttler et al., 2011; Yang et al., 2015). For example, the chloroplast‐encoded protein, putative chloroplast open reading frame 3 (Ycf3), contains a tetratricopeptide repeat and acts as a chaperone that interacts directly with PsaA and PsaD during the assembly of stable PSI units (Boudreau et al., 1997; Naver et al., 2001). Moreover, a Ycf3‐interacting protein, Y3IP1, is a nuclear‐encoded protein that targets thylakoid and cooperates with Ycf3 to ensure the assembly of stable PSI in the thylakoid membrane of tobacco and Arabidopsis (Albus et al., 2010). Ycf4 is another plastid genome‐encoded cofactor involved in PSI biogenesis, and disruption of the Ycf4 gene resulted in a reduction of PSI level and loss of PSI activity (Boudreau et al., 1997). PSBP‐Domain Protein 1 (PPD1), Photosystem I Assembly 2 (PSA2), and Photosystem I Assembly 3 (PSA3) are also reported to post‐translationally assist PSI assembly (Fristedt et al., 2014; Liu et al., 2012; Shen et al., 2017). However, the characterization and molecular functions of many other potential protein factors involved in integration of PSI‐LHCI supercomplex assembly are still poorly understood.

The stroma‐located chloroplast signal recognition particle (cpSRP) consists of two subunits, the conserved 54‐kDa GTPase cpSRP54 and a chloroplast‐specific 43‐kDa protein, cpSRP43, and both are found to be essential for assisting the mature but unfolded light‐harvesting Chl a/b binding proteins (LHCP) integration into thylakoid membrane (Akopian et al., 2013; Ziehe et al., 2017). After binding the LHCP to the heterodimeric cpSRP43/cpSRP54 complex, they form a soluble transit complex and dock to the SRP receptor cpFtsY and the Alb3 translocase at the membrane followed by the release and integration of the LHCP into the thylakoid membrane in a GTP‐dependent manner (Chandrasekar et al., 2017). The cpSRP54 protein contains an N‐terminal N domain, a central G domain with GTPase activity, and a methionine‐rich M domain in the C‐terminus (Franklin & Hoffman, 1993). A 10‐residue segment within the C‐terminal tail region of the M domain, which contains the conserved positively charged cpSRP43 binding motif ARRKR, was shown to be important for cpSRP43 binding in Arabidopsis (Dünschede et al., 2015; Funke et al., 2005). In contrast to the Arabidopsis proteins, cpSRP54 neither interacted with full‐length cpSRP32 nor is involved in LHCP recognition in Chlamydomonas reinhardtii, indicating that the molecular function of cpSRP54 developed and evolved diversely in different species (Dünschede et al., 2015). The interaction between cpSRP43 and LHCP has been unambiguously proven in land plants; however, whether cpSRP54 contacts LHCP directly and its precise contribution to the PSI‐LHCI supercomplex assembly in land plants, especially in the important crops, remains unclear.

As an easily observed phenotypic mutation in plants, leaf‐color mutants have played an important role in the functional investigation of Chl biogenesis and photosynthetic systems. The most common leaf‐color mutants are Chl‐deficient mutants with pale, yellow, or yellow‐white leaves. Chl‐deficient mutants usually have defects in photosynthetic carbohydrate production, which is required for plant growth and development. Therefore, these broad characteristic features of Chl‐deficient mutants are widely applied for the study of photosynthesis in plants (Chen et al., 2019; Yang et al., 2016; Zhao et al., 2017). Stöckel et al. (2006) reported the identification of Pale Yellow Green 7 (PYG7), which encodes a tetratricopeptide repeat protein involved in PSI assembly via analysis of a Chl‐deficient mutant. In the present work, a new Chl‐deficient mutant, pale‐green leaf 1 (pgl1), was isolated from a rice (Oryza sativa subsp. indica cv. Shuangkezao) mutant library created with the Co60 γ‐ray treatment. Functional characterization of the pgl1 mutant via a series of molecular and biochemical assays revealed that OsPGL1 encodes a chloroplast protein homologous to cpSRP54 and the findings from the present study suggest that this protein may play a pivotal role in assisting the efficient assembly of the PSI‐LHCI supercomplex in rice.

RESULTS

Phenotypic characterization of the pgl1 mutant

The pgl1 mutant was identified by its virescent leaf phenotype from a γ‐ray‐generated mutant library of the indica rice cv. Shuangkezao. Compared with the wild type (WT), the pgl1 mutant displayed pale‐green leaves and retarded growth (Figure 1a,b). Chl content analysis showed that both Chl a and b amounts were reduced in the mutant whereas the ratio of Chl a/b content was increased (Figure 1c). The total Chl content was therefore decreased by approximately 25% in comparison with that of the WT (Figure 1c).

FIGURE 1

Comparison of the phenotypes of the WT and pgl1. (a) Phenotype of aerial seedlings of the WT and pgl1. (b) Fully expanded leaves of the WT and pgl1. (c) Content of chlorophyll a, b, a + b, and the ratio of chlorophyll a/b content between the WT and pgl1. Measurements were based on five independent lines, the differences between the WT and pgl1 for each comparison are significant (*p value <.05; **p value <.01; ***p value <.001). (d–i) Ultrastructure of chloroplasts in seedlings of the WT and pgl1. (d), (f), and (h) are the WT; (e), (g), and (i) are pgl1. (d,e) Overall ultrastructure of chloroplasts. (f,g) Stroma lamellae. (h,i) Grana stacks and discs per grana stack. Red arrows indicate the grana stacks. Bar: a = 1 cm, b = 0.2 cm, d = 50 μm, e = 5 μm, f = 2.5 μm, g = 50 μm, h = 5 μm, i = 2.5 μm.

To evaluate the photosynthetic performance of the pgl1 mutation, the activity of photosynthetic oxygen evolution in thylakoid membranes was assessed using the presence of an artificial electron acceptor p‐benzoquinone (p‐BQ). No obvious difference in the activities of oxygen evolution was observed between the pgl1 mutant and the WT (Figure S1a), which suggested that the function of PSII in the core complexes might not be impaired in the pgl1 mutant. Photophosphorylation activity of pgl1 was also compared with that of the WT under both light and dark conditions, and similar photophosphorylation activities were observed in both genotypes (Figure S1b). In the dark‐adapted seedlings, the maximal PSII quantum yield (Fv/Fm) of pgl1 was not altered (Figure S1c), while the light intensity dependence of the PSII electron transport rate (ETR) and PSII quantum yield (YII) were attenuated (Figure S2). These results indicated that the mutation in OsPGL1 might lead to altered sensitivity to light.

Ultrastructure of the pgl1 chloroplasts

To assess the anatomical and cellular defects in chloroplasts of pgl1 mutants, the leaves of 2‐week‐old seedlings were examined by transmission electron microscopy (TEM). Irregularly shaped and abnormally developed chloroplasts were observed in leaves of pgl1 mutants. Chloroplasts in pgl1 were found to be smaller and contained a lower number of stroma lamellae, grana stacks, and discs per grana stack than those of WT (Figure 1d–i). In addition, thylakoids tended to be regularly arranged parallel to the long axis of the chloroplast in the WT, whereas in the pgl1 mutant the thylakoid distribution is irregular. These observations suggest that in the pgl1 mutant the organization and structural aspects of stroma and grana are affected.

Alterations in the content of chloroplast proteins

To further investigate the biochemical impacts on photosynthetic complexes in pgl1 mutants, immunoblot analyses were performed with antibodies against specific subunits of the photosynthetic thylakoid membrane complexes. Both PsaA and PsaB, the major PSI RC proteins, were dramatically decreased in pgl1 compared with the WT (Figure 2a). The content of the PSII core protein, PsbB (CP47), was also significantly lower in pgl1 than the WT (Figure 2a). However, the amounts of the other three PSII core proteins, D1, D2, and CP43, were not affected in pgl1 (Figure 2b). In addition, a significant decrease was detected in the accumulation of the ATP synthase subunit AtpB and AtpC, the cytochrome protein Cytb6 and Cytf, and the light‐harvesting complex LHCa1 and LHCb1, in the pgl1 mutant (Figure 2a).

FIGURE 2

Expression level of chloroplast thylakoid membrane proteins in the WT and pgl1. Western blot was performed with the specific antibodies of thylakoid membrane proteins using fully expanded leaves from the WT and pgl1. Each lane was loaded with the same total protein contents (20 μg) determined by BCA protein assay. (a) Proteins with a decreased content in pgl1 compared with the WT. (b) Proteins with a consistent content between the WT and pgl1. The names of primary antibodies used for each assay were listed on the right of blotting images and the corresponding categories of these photosynthetic proteins were listed on the left.

Map‐based cloning of the OsPGL1

To identify the OsPGL1 gene in rice, an F2 population derived from a cross between the pgl1 mutant and cv. Nipponbare was generated. A nearly 3:1 segregation ratio of WT to mutant phenotypes in the F2 population suggested that pgl1 is conferred by a single recessive mutation. Further fine mapping was conducted through a large F2 mapping population consisting of 1500 F2 plants based on a high‐density simple‐sequence repeat (SSR) markers. Linkage analysis delineated the OsPGL1 gene within a 22‐kb genomic interval flanked by the SSR markers RM26077 and RM26079 on chromosome 11 (Figure 3a). This region is covered by two BAC clones, AC116949 and AC138196, and is predicted to contain three putative genes. Of these, one gene contains six exons and five introns within its 606‐bp open reading frame (ORF), which encodes a chloroplast targeted protein homologous to the 54‐kDa subunit of the signal recognition particle (cpSRP54), was considered as the most possible candidate for OsPGL1 based on the observed structural defects in the chloroplast of pgl1 (Figure 3a, GenBank accession number KP119616). Sequence analysis revealed a single nucleotide mutation from G to A occurred at the first exon‐intron splicing donor sites in the pgl1 mutant compared with the WT, resulting in premature termination of translation (Figure 3a).

FIGURE 3

Map‐based cloning of the OsPGL1 gene. (a) Physical map of preliminary mapping (upper) and fine mapping (lower). Molecular markers, number of reconstituting lines, and BAC clones are indicated. One point mutation at the exon‐intron splicing donor site G to A has been marked which causes a premature stop codon in pgl1. (b) Expression of the OsPGL1 transcript in different tissues of the WT and pgl1. Three independent lines from each genotype were used for qRT‐PCR assay, and the differences between the WT and pgl1 in leaf sheath, young leaf, and mature leaf are significant (***p value <.001). (c) Leaf phenotypes of representative OsPGL1 RNAi transgenic plants. A transformation using the original pTCK303 vector without RNAi hairpin structure was used as a negative control. (d) Expression of OsPGL1 transcript in the RNAi lines. Three biological replicates were performed for each T1 transgenic line, and the differences of expression of OsPGL1 between RNAi lines (1–3) and the negative control (vector) are significant (***p value <.001).

Protein sequence analysis for OsPGL1 by BLAST suggests that it is an allele of Os11g0153600 in indica groups. Besides OsPGL1, there are three other SRP54 proteins in the genome of O. sativa, including Os11g0153700, Os01g0772800, and Os05g0509500. The latter two are more similar to SRP54 proteins in human and yeast, while OsPGL1 is more similar to SRP54 proteins in Arabidopsis and Brachypodium distachyon based on phylogenetic analysis of these homologs (Figure S3).

Real‐time RT‐PCR analysis was performed to investigate the expression pattern of OsPGL1 in different tissues of the WT plants. OsPGL1 transcript is highly expressed in leave blades and leaf sheaths, but rarely detected in roots or panicles (Figure 3b). A comparison of OsPGL1 expression also showed that it is dramatically reduced in leaf blades and leaf sheaths of the pgl1 mutant (Figure 3b). These results suggest that OsPGL1 is primarily expressed in photosynthetic tissues.

To confirm that the mutant phenotype is attributed to the functional loss of the OsPGL1 gene, a loss‐of‐function reverse complementation test was performed. An RNAi construct containing a 317‐bp coding region sequence of OsPGL1 gene was introduced into the japonica rice cv. Zhonghua 11 using the Agrobacterium‐mediated transformation method. In total, 27 independent transgenic lines exhibited virescent phenotypes like the pgl1 mutant (Figure 3c). Real‐time RT‐PCR further revealed the transcript levels of OsPGL1 in the RNAi lines were significantly reduced (Figure 3d). Thus, we concluded that the splicing point mutation in OsPGL1 resulted in the pgl1 mutant phenotype.

Disruption of PSI assembly in the pgl1 mutant

To investigate any structural and assembly alterations of chloroplast proteins and their complexes, photosynthetic protein complexes were solubilized from thylakoid membrane proteins using n‐dodecyl‐b‐D‐maltoside (DM) and separated by blue native PAGE (BN‐PAGE). Compared with the WT, the pgl1 mutant exhibited decreased amounts of all the supercomplexes (Figure 4a). Interestingly, one extra band (band X) of approximately 590 kDa located between the dimeric PSII/PSI‐LHCI (band I) and the monomeric PSII (band II) was observed in the pgl1 mutant (Figure 4a). Analyses of the 2‐D SDS‐PAGE gels after Coomassie blue staining displayed similar complex subunit composition between band X and I (Figures 4b). Further, western blot assays revealed that the extra band can only be detected by the antibodies against the proteins in the PSI core complex, but not in the LHCI, ATPase, or the PSII core complexes (Figure 5). These results indicated that the extra band comprises proteins from the PSI core complex, and the representing protein complex of this extra band X might be an incomplete dimers of PSI‐LHCI complex with the absence of the LHCI subunits.

FIGURE 4

Blue native and 2D analysis of thylakoid membrane proteins. (a) Blue native PAGE separation of thylakoid membrane protein complexes in the WT and pgl1. Typical complex bands of thylakoid are marked as I, II, III, and IV and the extra band in pgl1 is marked by X. (b) 2D SDS‐PAGE gels showing the potential PSI‐LHCI core complex compounds in the WT and pgl1. The blue native gels from the WT and pgl1 were excised separately for loading and indicated on the top of each 2D gel. The proteins from band X were marked by the red arrows. The sizes of protein markers were indicated on the left of gels.

FIGURE 5

Detection of the thylakoid membrane protein components in the extra band X. Western blot was performed on the 2D gels from Figure 4 in both the WT and pgl1 by the specific antibodies of thylakoid membrane proteins. (a) PsaA. (b) PsaB. (c) PsaD. (d) Lhca1. (e) PsbB. (f) ATPB. Positions of bands I, II, III, IV, and X from blue native gels were marked on the top of 2D gels, and the sizes of the protein markers were indicated on the left. Red and blue arrows stand for signals from band I and X separately.

Because PSI and PSII display distinct fluorescence emission bands at 77 K, the relative amounts of these two photosystems can be estimated by 77‐K fluorescence emission spectroscopy. To investigate the impact of the pgl1 mutation on the stoichiometry of photosystems, the 77‐K Chl fluorescence of thylakoid membranes from seedlings of pgl1 was compared with that of the WT. After excitation of Chls at 430 nm, the thylakoid membranes of the WT showed two major peaks at 740 and 685 nm, which correspond to the PSI and PSII complexes, respectively. In contrast, pgl1 showed an 8‐nm blue shift in fluorescence emission maximum, probably due to the abnormal Chls of the PSI light antennae. In addition, the PSI peak of the mutant was significantly lower than that of the WT when normalized at 685 nm (Figure 6). This observation is consistent with the drastic reduction of light‐harvesting Chl proteins associated with photosystem I in comparison with the PSII internal antennae (e.g., CP43). These results further confirmed that the pgl1 mutation directly affects the composition and accumulation of the PSI complex, rather than the PSII complex.

FIGURE 6

Seventy‐seven‐Kelvin fluorescence spectra of thylakoids from leaves of seedlings. The excitation wavelength was set at 430 nm, and the emission spectra were recorded from 650 to 800 nm using a Hitachi F‐4500 spectrofluorometer. The spectra were normalized at 685 nm. The major emission peaks at about 740 and 685 nm were assigned to the PSI and PSII complexes, respectively. Eight‐nanometer blue shift of fluorescence emission maximum was observed in pgl1. X axis, wavelength (nm); Y axis, fluorescence intensity (rel. units)

DISCUSSION

Here, we report the identification and functional characterization of a new rice Chl‐deficient mutant, pgl1. Map‐based cloning of this locus and the supporting independent transgenic RNAi studies have established the molecular identify of OsPGl1 and that it encodes a chloroplast targeted protein similar to one of uncharacterized cpSRP54 homologs in rice. Our studies show that the mutant plants displayed a pale‐green leaf phenotype, significant reduction of growth, and a decreased level of Chl (Figure 1). We further assessed the effects of OSPGL1 on the accumulation of protein subunits from different photosynthetic complexes (Figure 2). Generally, the amounts of the PsaA and PsaB subunits of PSI were severely decreased in pgl1. Moreover, significant decreases were detected in the accumulation of one PSII core protein, CP47, the ATP synthase subunits, AtpB and AtpC, the cytochrome proteins, Cytb6 and Cytf, as well as the LHC proteins, LHCa1 and LHCb1, in the mutant. Our studies further revealed that in addition to the significant reduction in the levels of all the protein complexes, structural alteration was observed in the PSI‐LHCI supercomplex in pgl1 (Figures 4 and 5), in accordance with the fluorescence data which strongly indicated that PSI complex assembly was primarily affected in the pgl1 mutant (Figure 6). These results are reminiscent of previous findings in several other PSI mutants where reductions were also observed in PSII and other thylakoid complexes owing to the secondary damage of the mutation (Haldrup et al., 2000, 2003; Landau et al., 2009). Hence, the pgl1 mutation was identified as primarily affecting PSI assembly and functions.

PSI is a thylakoid‐embedded pigment‐protein complex that participates in the light reaction of photosynthesis (Nelson & Yocum, 2006; Wostrikoff et al., 2004). PSI harvests light, and then catalyzes electron transfer from plastocyanin to ferredoxin, ultimately leading to the reduction of NADP+ (Chitnis, 2001; Landau et al., 2009). The biogenesis of the PSI complex requires the coordinated expression of nuclear and chloroplast genes, the targeting of the subunits to their proper localization within the chloroplast, as well as the correct integration of multiple subunits into a functional complex (Rochaix, 2002; Rochaix et al., 2004). The proper docking of LHCI to PSI is considered to be a crucial step in the process of PSI assembly because a faulty connection between these two complexes would prevent the transfer of excitation energy from LHCI to the PSI RC (Barber & Anderson, 1992; Hippler et al., 2000). So far, several proteins that are required for efficient PSI biogenesis of higher plants, Chlamydomonas and Synechocystis, have been identified by mutational analyses. However, most of these proteins seem to be implicated specifically in the assembly of the PSI core complex. For instance, Ycf3 appears to act as a chaperone that interacts directly and specifically with the two core subunits, PsaA and PsaD, during PSI assembly (Naver et al., 2001). Ycf4 is associated with a high molecular mass complex containing several core subunits, therefore acting as a scaffold for PSI assembly. It is proposed that PsaF would be transferred along with PsaJ from the Ycf4 complex to the PSI RC (Ozawa et al., 2009). In the pyg7 mutant, PSI core subunits fail to assemble into a stable complex, implying that PYG7 is required for the proper accumulation of PSI (Stöckel et al., 2006). In the current study, a novel gene, OsPGL1, appears to be required for the efficient integration of the peripheral LHCI proteins to the PSI core complex in rice.

Intriguingly, OsPGL1 is predicted to encode a cpSRP54 protein in rice. The cpSRP54 protein has been implicated in the targeting of the nuclear‐encoded LHCPs to the thylakoid membrane (Eichacker & Henry, 2001; Li et al., 1995; Schünemann, 2004). So far, cpSRP54 has been extensively studied in vitro (Li et al., 1995; Schünemann et al., 1999; Tu et al., 1999), while information on its function in vivo is still limited. Pilgrim et al. (1998) first reported the exploration of cpSRP54 function in vivo by mutational analysis in Arabidopsis. Expression of a dominant negative form of cpSRP54 in Arabidopsis inhibits the biogenesis of LHCP and other chloroplast proteins, which are encoded in the chloroplast and co‐translationally targeted to the thylakoid membrane, implying that cpSRP54 plays a pleiotropic role in chloroplast biogenesis (Pilgrim et al., 1998). Recent studies revealed that loss of function of cpSRP54 or mutations within the M domain of cpSRP54 disrupt the formation of the cpSRP/LHCP transit complex (Dünschede et al., 2015; Henderson et al., 2016). In the pgl1 mutant, a significant reduction was also detected in the accumulation of many chloroplast proteins including those belonging to LHC, the PSI core complex, ATP synthase, and the cytochrome b6f complex (Figure 2). However, most of the core proteins in the PSII core complex still maintain similar expression levels as the WT, except CP47. CP47 is the supposedly most stable protein in the PSII complex and has the slowest relative turnover rates compared with D1, D2, and CP43 (Mattoo et al., 1999). Moreover, the activities of PSII are not affected by the decrease of CP47 (Figures 6 and S1c), indicating that the requirement of CP47 for the assembly of the PSII complex is sufficed in pgl1. Altogether, the potential contribution of OsPGL1 to the distinctive photoinactivation and protein dynamics of the PSII complex might be indirect as suggested by the previous studies with other PSI‐related mutants (Haldrup et al., 2000, 2003; Landau et al., 2009). On the other hand, it seems that there are also differences between Arabidopsis and rice in the absence of cpSRP54. Though mutations in cpSRP54 caused the reduction of the Chl content in the two species, the ratio of Chl a/b content was found to be unaffected in Arabidopsis while elevated in rice. In addition, increased expression of certain proteins such as the 70‐kDa heat shock protein (HSP70) was detected in the Arabidopsis mutant, suggesting that cpSRP54 might be nonessential and the mutant plant could compensate for the defects through alternative thylakoid targeting pathways. On the contrary, we did not observe a detectable elevation in the expression of HSP70 in rice. Therefore, it is tempting to speculate that in contrast to the findings in dicots, the cpSRP54 pathway is indispensable for thylakoid targeting in monocots.

The significantly decreased expression levels of many chloroplast proteins in pgl1 (Figure 2) further affect the activities of photosystems, especially PSI (Figures 4 and 6), which is predicted as the primary cause resulting in the pale‐green phenotype and reduced photosynthetic performance (Figures 1 and S2). Interestingly, although the expression levels of some PSII‐LHCII subunits, CP47 and Lhcb, are severely reduced (Figure 2), the activities of PSII still remain at a similar level in pgl1 in comparison with the WT (Figure S1c), indicating that the intact structure and function of PSII‐LHCII complex are not affected in pgl1. This conclusion was further validated by 77‐K fluorescence spectra analysis (Figure 6), suggesting the function of OsPGL1 is specific to the assembly of PSI, but not PSII, associated complexes. Moreover, the extra band from the blue native gel analysis contains mostly core subunits from the PSI complex rather than PSII complex (Figure 5), supporting the characteristics of specific targeting of PSI for OsPGL1. In rice, we found three other homologs of OsPGL1, the molecular functions of this gene family probably differentiated and specialized during the evolution and domestication processes in rice, and OsPGL1 evolved to specially recognize and assemble PSI‐LHCI supercomplex. Further investigation on the functions of other members of this gene family will provide new insights into the complex regulatory mechanism of photosystem assembly.

It is proposed that LHCPs are synthesized in the cytoplasm, translocated across the chloroplast envelope membrane into the stroma via an N‐terminal transit peptide, and then targeted to the thylakoid membrane (Keegstra & Cline, 1999; Payan & Cline, 1991). In the stroma, LHCPs bind post‐translationally to a heterodimer consisting of cpSRP54 and cpSRP43, forming a soluble targeting intermediate, designated transit complex (Li et al., 1995; Payan & Cline, 1991). The maintenance of the solubility of LHCPs by cpSRP ensures their transport through the aqueous stroma to the thylakoid membrane in Arabidopsis (Hutin et al., 2002). Although the heterodimer formed by cpSRP54 and cpSRP43 is different between Arabidopsis and C. reinhardtii, our results showed a similar heterodimer formed by cpSRP54 and cpSRP43 in rice as in the Arabidopsis. This finding is further supported by the 77‐K fluorescence spectra analysis with noticeably decreased PSI peak relative to the PSII fluorescence peak in pgl1 mutant, which is consistent with the analysis of cpSRP54 and cpSRP43 mutants in Arabidopsis (Hutin et al., 2002). Based on blue native gel analysis, we detected an extra band in the thylakoid membranes of pgl1 (Figure 4). This extra protein band corresponds to a molecular mass of approximately 590 kDa and is located between the dimeric PSII/PSI‐LHCI and the monomeric PSII. Further immunodetection after a second‐dimension separation indicated that the extra band comprised the proteins from the PSI core complex, but not from the LHCI or the PSII core complexes. Considering that the PSI‐LHCI supercomplex is with a molecular mass of about 670 kDa while LHCI contains four polypeptides (Lhca1–Lhca4) of about 22, 25, 26, and 27 kDa, we speculated that due to the mutation in OsPGL1, LHCI failed to integrate efficiently to the PSI core complex. Thus, the OsPGL1 encoded cpSRP54 in rice might be involved specifically in the integration of LHCI, ultimately leading to the efficient assembly of the PSI‐LHCI supercomplex.

MATERIALS AND METHODS

Plant material and growth conditions

The rice mutant pgl1 was screened from a mutagenic indica cv. Shuangkezao with the Co60 γ‐ray. All the plants for experiments were grown in a growth chamber under a photon flux density of 400 μmol photons m−2 s−1 at 28°C with a 16‐h light/8‐h dark photoperiod. Relative humidity was controlled at approximately 50%.

Chl analysis

The fully expanded leaves of three‐leaf‐stage seedlings of the WT and pgl1 were harvested, weighed, and ground in liquid nitrogen for Chl extraction with 80% acetone. The amounts of the Chl a and Chl b were determined spectrophotometrically, as described by Porra et al. (1989). Five replicates were performed for each genotype.

Thylakoid membrane isolation, oxygen evolution and estimation of photophosphorylation activities

Leaf tissue was ground in cold STN medium (0.4‐M sucrose, 50‐mM Tris–HCl, pH 7.6, 10‐mM NaCl). The homogenate was filtered through two layers of gauze and then centrifuged at 200  for 3 min. The supernatant was collected and centrifuged at 6000  for 10 min. The pellet was washed twice with the homogenizing medium and then resuspended in the same medium. All operations were carried out at 4°C. The Chl was determined as described above, and the total protein content was determined with the protein assay kit (Bio‐Rad, USA) after the thylakoid membranes were treated with 5% Triton X−100 shaking on ice for 1 h.

The thylakoid membranes (5 μg Chl ml−1) were stirred with the reaction buffer (0.4‐M sucrose, 50‐mM Tris–HCl, pH 7.6, 10‐mM NaCl, 2‐mM MgCl2, 2‐mM EDTA, 0.5‐mM p‐BQ, and 2‐mM NH4Cl) in the thermostated glass vessel of a Clark‐type oxygen electrode at 25°C. Oxygen evolution was normally detected 2 min after the start of illumination (800 μmol photons m−2 s−1). The solubility of O2 in water is 253 μM at 25°C.

Measurements of photophosphorylation activities of the WT and pgl1 were estimated by measuring light‐induced ATP synthesis in chloroplasts using isolated thylakoid membranes. ATP production was measured by comparing the ATP level in the dark against the level of 1.5 min after illumination (45 μmol photons m−2 s−1) at 28°C. ATP content was then analyzed by the luciferin‐luciferase method using a luminometer (RS 9901 luminometer). Five replicates were performed for each analysis described above.

Measurement of Chl fluorescence parameters

The leaves of both the WT and pgl1 were adapted in the dark overnight before the measurements. The parameters of Chl fluorescence were measured by using a Dual‐PAM−100 (Heinz Walz, Germany), including ETR, PSII quantum yield [Y (II)], coefficient of photochemical quenching (qP), and coefficient of nonphotochemical quenching (qN). Five replicates were performed for each genotype.

Electrophoresis and immunoblotting

For total protein extraction, leaf samples were homogenized with lysis buffer (50‐mM Tris–HCl, pH 7.6, 150‐mM NaCl, 2% SDS, 0.01% 2‐mercaptoethanol) and centrifuged at 15,000 rpm for 30 min at 4°C. The total protein content was determined using a BCA protein assay kit (Pierce, USA). After denaturing, each sample with 20‐μg proteins was separated through a 12% SDS‐PAGE gel and transferred to a PVDF membrane (Millopore, Germany). Protein blots were blocked with 3% skimmed milk and then probed with specific antibodies against the PSI core complex (PsaA and PsaB), the PSII core complex (PsbA, PsbB, PsbC, and PsbD), the light‐harvesting complex (Lhca1 and Lhcb1), ATPase (ATPB and ATPC), the cytochrome b6f complex (Cytb6 and Cytf), and Hsp70 (chloroplastic). All the primary antibodies were purchased from the Agrisera Company (Sweden). Anti‐rabbit IgG (GE Healthcare, UK) was used as a secondary antibody for subsequent detection via enhanced chemiluminescence (GE Healthcare). The response of antibody PsbA, PsbC, PsbD, and HSP70 was confirmed by proteins with different dilutions.

BN‐PAGE was carried out according to Peng et al. (2006). The thylakoid membranes were washed with the buffer containing 330‐mM sorbitol and 50‐mM BisTris–HCl (pH 7.0) and resuspended in suspension buffer (20% glycerol and 25‐mM BisTris–HCl, pH 7.0) at a protein concentration of 1 mg ml−1. An equal volume of resuspension buffer containing 2% (w/v) DM was added to the thylakoid suspension gently. After incubation at 4°C for 1 h, the mixture was centrifuged at 12,000  for 10 min to remove insoluble material. The supernatant was combined with 1:10 volume of the sample buffer (5% Serva blue G in 100‐mM BisTris–HCl, pH 7.0, 0.5‐M 6‐amino‐n‐caproic acid, and 30% [w/v] glycerol) and applied to 0.75‐mm‐thick 5% to 13% acrylamide gradient gels.

For 2D analysis, excised BN‐PAGE lanes were soaked in 1% SDS for 1 h and then subjected to the second‐dimension SDS‐PAGE. After electrophoresis, SDS‐PAGE gels containing separate pigment proteins from the WT and the mutant, respectively, were juxtaposed and subjected to electroblotting together for western blot analysis.

TEM

Leaf tissue was cut into small pieces and fixed overnight in 2.5% glutaraldehyde in 0.1‐M phosphate buffer (pH 7.2) at 4°C. After fixation, the samples were washed with 0.1‐M phosphate‐buffered saline (PBS) and postfixed in 1% OsO4. Followed by rinsing in 0.1‐M PBS, the samples were dehydrated in a graded ethanol‐acetone series and embedded in Epon 812 resin. Ultrathin sections were obtained using an ultramicrotome, stained with 10‐mM lead citrate followed by 2% uranyl acetate, and examined with an H7650 transmission electron microscope.

Mapping of OsPGL1

To map OsPGL1, linkage analyses based on a series of SSR markers distributed evenly across all 12 rice chromosomes were performed using an F2 population (consisting of 350 F2 individuals) derived from a cross between the pgl1 mutant (indica, cv. Shuangkezao) and a WT rice Nipponbare (japonica). For fine mapping, an additional set of SSR markers, located in the genomic region defined by the initial linkage analysis, was then applied to a large F2 mapping population (consisting of 1500 F2 plants) developed from the same cross. The SSR markers used to analyze the polymorphisms between Ospgl1 and Nipponbare were obtained from the Gramene web (http://www.gramene.org).

Phylogenetic analysis

The BLAST search program (http://www.ncbi.nlm.nih.gov/BLAST/) was used to identify homologous protein sequences to OsPGL1. The obtained homologous sequences were aligned using MEGA version 5.1 software, and the neighbor‐joining tree was generated with the Poisson correction method using the same software. Bootstrap replication (1000 replications) was used for a statistical P support for the nodes in the phylogenetic tree.

Quantitative real‐time PCR

Total RNA was extracted from different tissues (roots, young leaves, mature leaves, leaf sheath, and young panicles) of the WT and pgl1 by using the Trizol reagent (Invitrogen) following the manufacturer's instructions. Samples were then treated with RNase‐free DNase (Promega) for 30 min. For real‐time PCR, 2 μg of total RNA was used to synthesize first‐strand cDNA using First‐Strand cDNA Synthesis Kit (Fermentas) following the manufacturer's protocols.

Real‐time RT‐PCR analysis was performed on three independent biological replicates using SYBR® Premix Ex Taq™ (Takara) on a Bio‐Rad MyIQ machine according to the manufacturer's protocol (Applied Biosystems). The relative quantification method (Delta–Delta Ct) was used to evaluate quantitative variation between the replicates examined. The amplification of ACTIN was used as an internal control to normalize all data. The OsPGL1 gene‐specific primers used for real‐time RT‐PCR were qPGL1‐F 5′‐GAACACGAGGTGAGCCAGTT‐3′ and qPGL1‐R 5′‐CCTGGAAGTGCTTCTTGTCC‐3′.

RNAi construct and transformation

The 317‐bp coding sequence of OsPGL1 was amplified using primers 5′‐ACTAGTCTCAGAAGGAACAGAAGCCAAAC‐3′ and 5′‐GAGCTCGGTCCTCCATTCGTTCTCCAC‐3′ and then ligated to the pTA2 vector. The positive transformant was further digested by BamHI and KpnI and then SacI and SpeI and ligated to the pTCK303 vector, respectively, to form a hairpin structure (Wang et al., 2004). The resultant construct was transformed into the japonica rice cultivar Zhonghua 11 via the Agrobacterium‐mediated transformation procedure. Three independent lines from the T1 generation were used for phenotyping and gene expression detection by qRT‐PCR.

Seventy‐seven‐Kelvin fluorescence spectroscopy

Fluorescence emission spectra at 77 K of the WT and pgl1 were recorded with excitation and emission slit widths of 5 and 1.5 nm, respectively, using a Hitachi F‐4500 spectrofluorometer. The excitation wavelength used was 430 nm, and the emission spectra were recorded from 650 to 800 nm. All samples were harvested by centrifugation (5000  for 5 min at 25°C), washed, resuspended in fresh BG−11 medium buffered with Tris–HCl (5 mM, pH 8.0), and then adjusted to a Chl content of 2 μg ml−1. The resuspended cultures were frozen in liquid nitrogen after the dark adaptation for 10 min. The fluorescence spectrum was obtained by averaging six scans for each sample in different tubes.

Statistical analysis

Statistical comparison between two samples was carried out using two‐tailed Student's t test with two‐sample unequal variance (*p < .05; **p < .01; ***p < .001). Error bars in all figures represent standard deviations. The number of replicates is reported in the figure legends.

CONFLICT OF INTEREST

The authors declare no conflicts of interest associated with the work described in this manuscript.

AUTHOR CONTRIBUTIONS

P.G., H.X., S.Y., H.W., and J.Y. conceived the project. P.G., H.X., H.M., S.Y., H.W., and J.Y. designed the experiments. P.G., H.X., Z.L., C.Z., and L.W. carried out the experiments. P.G., H.X., H.M., S.Y., R.D., H.W., and J.Y. analyzed and interpreted the data. P.G., H.X., R.D., H.W., and J.Y. wrote the manuscript. All authors read and approved of the manuscript.

Figure S1. Photosynthetic performance of the (a) Activity of photosynthetic oxygen evolution of thylakoid membranes in the the WT and pgl1 with the presence of an artificial electron acceptor p‐BQ. No significant difference can be found between the WT and pgl1. (b) Photophosphorylation activity of the WT and pgl1 in light and dark conditions. Significant difference can be found between light and dark treatment (** p‐value < 0.001), but no significantly different response can be found between the WT and pgl1. (c) Maximal PSII quantum yield (Fv/Fm) of the WT and pgl1 in the dark‐adapted seedlings. No significant difference can be found between the WT and pgl1.

Figure S2. Comparisons of fluorescence parameters between the WT and (a) electron transport rate (ETR). (b) PSII quantum yield [Y (II)]. (c) coefficient of photochemical quenching (qP). (d) coefficient of non‐photochemical quenching (qN). The means from three independent measurements were used for plotting. X‐axis, light intensity (μmol m−2 s−1); Y‐axis, the values of ETR (μmol s−1), Y (II) (rel. unit), qP (rel. unit) and qN (rel. unit).

Figure S3. Phylogenetic analysis of (a) Alignment of protein sequences of OsPGL1 (cv. Shuangkezao), Os11g0153700 (cv. Nipponbare), Os05g0509500 (cv. Nipponbare), Os01g0772800 (cv. Nipponbare), AtCPSRP54 (Arabidopsis), BdCPSRP54 ( ), ScSRP54 ( ), HsSRP54 isoform 1 and isoform 2 ( ). The conserved SRP_SPB domain has been indicated by the blue box. The N‐terminal (N), the central G domain (G) and the methionine‐rich M domain have been circled by the red, green and orange dot boxes, respectively. (b) Phylogenetic analysis of OsPGL1 and related genes from rice, Arabidopsis, , human and yeast. Bootstrap values are indicated for each clade.

Click here for additional data file.