Cooperation of LPA3 and LPA2 Is Essential for Photosystem II 1 Assembly in Arabidopsis

Photosystem II (PSII) is a multisubunit membrane protein complex that in we report the identification of a chloroplast protein, LPA3, which for the assembly of CP43 subunit in PSII complexes. LPA3 interacts with LPA2, a previously identified PSII CP43 assembly factor, and a double mutation of LPA2 and LPA3 is more deleterious for assembly than either single mutation, resulting in a seedling-lethal phenotype. Our results indicate that LPA3 and LPA2 have overlapping functions in assisting CP43 assembly, and that cooperation between LPA2 and LPA3 is essential for PSII assembly. In addition, we provide evidence that LPA2 and LPA3 interact with Alb3, which is essential for thylakoid protein biogenesis. Thus, the function of Alb3 in some PSII assembly processes is probably mediated through interactions with LPA2 and LPA3. of performed native gel electrophoresis analysis of thylakoid solubilized The results

1 0 essential for PSII assembly. In addition, we provide evidence that LPA2 and LPA3 Oxygenic photosynthesis, in which oxygen and organic carbon are produced from water 2 and carbon dioxide using sunlight, provides energy for nearly all living organisms on 3 Earth. Four major multi-protein complexes, located in thylakoid membranes, are 4 responsible for the capture of light and its conversion to chemical energy in eukaryotic the total protein or thylakoid samples from the lpa3 mutant. These findings indicate that 1 the gene mutation leads to the loss of LPA3, and that the lpa3 mutant phenotype is 2 likely due to the loss of the function of the At1g73060 gene. of LPA3 comprise the plastid target signals (Fig. S4).

2
To confirm the prediction that LPA3 is a chloroplast protein, we isolated intact that some LPA3 is present in the stroma, but some is also associated with the thylakoid 1 8 membrane fractions (Fig. 3D). The chloroplast stromal ribulose bisphosphate 1 9 carboxylase large subunit (RbcL) and the integral thylakoid membrane protein D2 of 2 0 PSII were used as controls for this experiment. To examine the strength of the 2 1 membrane association of LPA3, we treated thylakoid membranes with salt and 2 2 chaotropic agents, which are known to remove extrinsic proteins (Boudreau et al., 1997). CaCl 2 (Fig. 3E).

8
To assess if the association of LPA3 with the membranes is modulated by the light 2 9 conditions, we isolated thylakoid membrane and stroma fractions for immunoblot 3 0 analysis from plants that were illuminated at 20 µmol m -2 s -1 (low light intensity) and 3 1 1000 µmol m -2 s -1 (high light intensity) for 2 h after growth at 120 µmol m -2 s -1 for five  Naver et al., 2001;Ossenbühl et al., 2004;Peng et al., 2006;Ma et al., 2007). Since the 8 above findings showed that LPA3 is required for efficient PSII assembly (Fig. 2), we 9 postulated that it may interact with specific subunits of PSII. To test this possibility, we 1 0 first used a modified split-ubiquitin system to examine potential interactions between 1 1 LPA3 and PSII proteins. The results showed that LPA3 interacts with CP43, but not D1, 1 2 D2 or CP47 (Fig. 4A). However, in further co-immunoprecipitation analyses of these 1 3 interactions CP43 was not detected in immunoprecipitates by either anti-LPA2 or 1 4 anti-LPA3 antibodies under non-denaturing conditions (Fig. 4B). Interactions between LPA1, LPA2 and LPA3

7
We have previously demonstrated the involvement of LPA1 and LPA2 in PSII assembly, interact with each other, but there was no indication that either protein interacts with 2 2 LPA1 in the yeast two-hybrid system (Fig. 4C). In order to further examine whether the 2 3 interaction between these assembly factors occurs in vivo, we used bimolecular 2 4 fluorescence complementation (BiFC) analysis (Fig. 4D). The results confirmed that 2 5 LPA2 and LPA3 interact in planta, and that there was no interaction between either of 2 6 these proteins and LPA1. Further co-immunoprecipitation analysis corroborated the 2 7 observations regarding these interactions (Fig. 4B). The Alb3/Oxa1/YidC protein family plays essential roles in the assembly of membrane 3 1 protein complexes (Kuhn et al., 2003;van der Laan et al., 2005;Kiefer and Kuhn, 2007), and we have previously shown that LPA2 interacts with Alb3 (Ma et al., 2007). Thus, given the interaction between LPA3 and LPA2, we examined the possibility that LPA3 1 also interacts with Alb3. The results showed that LPA3 and Alb3 interact in the yeast 2 two-hybrid system (Fig. 4A). BiFC and co-immunoprecipitation analyses further 3 demonstrated that Alb3 directly interacts with LPA2 and LPA3, but not LPA1 (Figs. 4B 4 and D). The above results strongly indicated that LPA2 and LPA3 interact with each other, but Fv/Fm ratio was close to zero in the lpa2lpa3 double mutant (Fig. 5A).
1 5 Next we examined the effects of lpa2 and lpa3 double mutation on the steady state photobleaching. Immunoblot analysis showed that PSII proteins D1, D2, CP47 and 1 9 CP43 were hardly detectable in lpa2lpa3 (Fig. 5B). The amounts of LHCII, Cytf and 2 0 CF1 were not reduced, but the contents of PsaA/B were about 70% of wild type levels.

1
Since the assembly of newly synthesized CP43 into PSII was slowed in either lpa2 or 2 2 lpa3 mutants, we also performed in vivo protein labeling analysis to check the assembly of PSII complexes from CP43-free PSII complexes in the lpa2lpa3 double 2 8 mutant.

9
In subsequent analyses we investigated the effects of mutations affecting one of LPA2 was unaffected in the lpa3 mutant. In contrast, the total cellular LPA3 protein content of the lpa2 mutant was similar to wild-type levels (Fig. 5D). the cooperative function of LPA3 with LPA2 in PSII assembly is also presented. Several key regulatory processes, including transcription, transcriptional stabilization, post-transcriptional level.

0
One of the possible roles of the LPA3 protein is that it may be required for the 2 1 synthesis of one or more PSII proteins, since it has been shown that loss of any genes 2 2 coding for PSII core proteins will profoundly affect levels of other components (Jensen 2 3 et al., 1986;de Vitry et al., 1989;Yu and Vermaas, 1990;1993). Accordingly, our in vivo labeling experiments showed that the synthesis rates of the PSII protein CP43 were 2 5 dramatically decreased in the lpa3 mutant, although those of the PSII core proteins D1, complex assembly, as it has been well demonstrated that rapid degradation of core protein occurs when it cannot be efficiently assembled into the PSII protein complex 1 (Jensen et al., 1986;de Vitry et al., 1989;Yu and Vermaas, 1990). Indeed, the assembly 2 of PSII complexes was less efficient in lpa3 than in wild-type plants ( Interestingly, the PSII protein CP43 was found at ca. 25% of wild-type levels in 9 lpa3 ( Fig. 1), although in the synthesis analyses the amount of labeled CP43 was well 1 0 below 10% of wild-type levels (Fig. 3). This suggests that newly synthesized CP43 1 1 polypeptide is very rapidly degraded, by a co-or early post-translation degradation 1 2 process. The greatly reduced CP43 synthesis rate may also suggest that translation considering the similar protein labeling pattern in very young seedlings and mature 1 9 leaves (Fig. 3), it is possible that PSII repair rather than biogenesis is affected in the 2 0 mutant, and the immunoblots revealed rates of biogenesis while the labeling following 2 1 pulses primarily detected subunits that were destined to replace previously formed 2 2 subunits. Cooperation between LPA2 and LPA3 Is Essential for PSII Assembly 2 5 The analyses described above demonstrated that LPA3 is required for the efficient 2 6 assembly of PSII, but the mechanisms involved remained to be elucidated. One possibility we explored is that LPA3 may be an integral constituent of PSII, since the 2 8 assembly of PSII has been shown to be perturbed in several PSII mutants that lack provided direct evidence that LPA3 specifically interacts with CP43 (Fig. 4A). However, 5 such interactions between these proteins were not detected in co-immunoprecipitation 6 analysis (Fig. 4B). The possibility for the absence of co-immunprecipitation is that the 7 interactions might be less stable or shorter in duration during the 8 co-immunoprecipitation assays. Regardless of the reasons for the lack of stable 9 interactions, the apparent transience of the interaction between LPA3 and CP43 suggests 1 0 that LPA3 probably assists, through membrane association, the correct folding and 1 1 assembly of CP43 into PSII complexes containing D1, D2 and CP47 proteins. between LPA2 and LPA3 in the PSII assembly process. Our results showed that LPA2 1 5 directly interacts with LPA3 (Fig. 4), indicating that LPA2 and LPA3 may form a 1 6 complex that regulates the insertion of CP43 into PSII. In addition, the seedling-lethal 1 7 phenotype of the lpa2lpa3 double mutant (Fig. 5) suggests that LPA2 and LPA3 suggest that LPA2 and LPA3 are both essential for PSII assembly.

2
LPA2 is an intrinsic membrane protein, while LPA3 is an extrinsic protein 2 3 associated with the thylakoid membranes (Fig. 3). Tight association between LPA3 and 2 4 membranes may make it available for interactions with another PSII assembly factor, 2 5 LPA2, during the PSII assembly process. However, the association between LPA3 and 2 6 the thylakoid membrane was decreased in lpa2 mutants (Fig. 5), but LPA3 could still 2 7 assist PSII assembly by interacting with CP43, albeit at reduced efficiency, as suggesting that Alb3 plays an essential role in thylakoid biogenesis (Sundberg et al. Alb3 has been demonstrated to interact directly with the D1, D2 and CP43 PSII core has also been shown to be involved in the assembly of D1 into PSII (Bellafiore et al., process. We found that Alb3 directly interacts with LPA2 and LPA3 (Fig. 4), both of 2 4 which assist CP43 assembly within PSII. The interaction of LPA2 with CP43 was 2 5 similar to that of LPA3, as seen in the yeast two-hybrid analysis, but the stable 2 6 interaction between them was not detected in the co-immunoprecipitation analysis ( Fig.   2  7 4). It may be speculated that LPA2/LPA3 transiently interacts with CP43 and 2 8 subsequently pass it to Alb3. Thus, the function of Alb3 in some of the PSII assembly 2 9 processes may be mediated through interactions with LPA2 and LPA3, ensuring that grown at ~20 µmol m -2 s -1 to avoid photobleaching. pre-illuminated Arabidopsis leaves with a fully activated Calvin cycle to exclude any 2 7 acceptor-side limitation of P700 oxidation. PSI was selectively excited by for 10 min, and finally loaded onto a separation gel with a 6 to 12% acrylamide 1 3 gradient.

4
The proteins were separated in 15% SDS polyacrylamide gels containing 6 M urea, 1 5 prior to transference to a nitrocellular membrane. Immunodetection using specific 1 6 antibodies was performed with the enhanced chemiluminescence method, according to 1 7 standard techniques. X-ray films were scanned and analyzed using the AlphaImager TM into Escherichia coli strain BL21 cells, and the overexpressed proteins were induced by 2 2 0.6 mM isopropylthio-β-D-galactoside for 5 h. The proteins were purified on a Ni-NTA 2 3 resin matrix using a nickel affinity column, and polyclonal antibodies were raised in 2 4 rabbit using purified antigen. In vivo protein labeling was essentially performed according to Meurer et al. (1998).

8
Primary leaves from 12-d-old and 5-wk-old Arabidopsis seedlings were preincubated in  2006). For autoradiography, gels were stained, dried, and exposed to X-ray films. The lpa3 mutation was mapped using molecular markers based on a simple sequence 5 length polymorphism (Lukowitz et al., 2000). Genomic DNA was isolated from F2 6 plants derived from a cross between lpa3 (genetic background, Columbia) and plastid-targeting signals were then compared between lpa3 and wild type plants. Immunolocalization Analysis 2 7 Intact chloroplasts were isolated according to Munekage et al. (2002). Briefly,

6
Polysomes were isolated from leaf tissues under conditions that maintained 2 7 polysome integrity, according to Barkan (1988). Briefly, RNA was isolated, fractionated, 2 8 and transferred onto nylon membranes. The filters were hybridized with 32 P-labeled 2 9 cDNA probes, then exposed to an X-ray film for 1 to 3 d. The hybridization probes were Interactions between them were then determined by growing diploid yeast colonies on 8 SD-His-Leu-Trp plates prior to measuring their β -Gal activity using the X-Gal filter 9 assay described by Stagljar et al. (1998). Protoplasts were isolated from 5-wk-old Arabidopsis rosette leaves grown as previously and pSAT4A-cEYFP-N1, and plasmids were co-transformed into protoplasts (Citovsky proteins were subjected to immunoblot analyses. Supplemental Data 3 0 The following materials are available in the online version of this article.