Mechanism of REP27 Protein Action in the D1 Protein Turnover and Photosystem-II Repair from Photodamage

The function of the REP27 protein in the photosystem-II repair process was elucidated. REP27 is a nuclear-encoded and chloroplast-targeted protein containing two tetratricopeptide repeated (TPR) motifs, two putative transmembrane domains, and an extended C-terminal region. Cell fractionation and Western blot analysis localized the REP27 protein in the C. reinhardtii chloroplast thylakoids. A folding model for REP27 suggested chloroplast stroma localization for N- and C-termini regions as well as the two tetratricopeptide repeats. A REP27 gene knockout strain of C. reinhardtii , termed rep27 mutant, was employed for complementation studies. The rep27 mutant was aberrant in the PSII-repair process and had substantially lower than wild type levels of D1 protein. Truncated REP27 cDNA constructs were made for complementation of the rep27 , whereby TPR1, TPR2, TPR1+TPR2, or the C-terminal domains were deleted. rep27 -complemented strains minus the TPR motifs showed elevated levels of D1 in thylakoids, comparable to those in the wild-type, but the PSII photochemical efficiency of these strains was not restored, suggesting that the functionality of the PSII reaction center could not be recovered in the absence of the TPR motifs. It is suggested that TPR motifs play a role in the functional activation of the newly integrated D1 protein in the PSII reaction center. rep27 complemented strains missing the C-terminal domain showed low levels of D1 protein in thylakoids, as well as low PSII photochemical efficiency, comparable to those in the rep27 mutant. Therefore, the C-terminal domain is needed for a de novo biosynthesis and/or assembly of D1 in the photodamaged PSII template. We conclude that REP27 plays a dual role in the regulation of D1 protein turnover by facilitating co-translational biosynthesis-insertion (C-terminal domain) and activation (TPR motifs) of the nascent D1 during the PSII repair process. The comparative thylakoid membrane functional and protein analysis of wild type and rep27 mutant in this work confirms and strengthens this hypothesis. the soluble fraction. Appressed and non-appressed thylakoid membrane regions were isolated upon mechanical fractionation differential centrifugation, as described Thylakoid membrane vesicles were precipitated by differential centrifugation at 10,000 × g (10K fraction), 40,000 × g (40K fraction) and 140,000 × g (140K fraction). The Chl a / b ratio was determined in all fractions to provide a measure of the differential enrichment of grana and stroma-exposed thylakoid membranes, and an indication of the PSII/PSI ratio. The supernatant of the 140K fraction was used as the soluble fraction. Proteins of all subfractions were separated in SDS-PAGE gel and probed with specific polyclonal antibodies. of ( ∆ psbA ) and D2-less ( ∆ psbD ) mutant strains, probed with specific polyclonal antibodies raised against the REP27 protein. B, SDS-PAGE of proteins (2 μ g Chl loaded) separated onto 12% acrylamide and visualized with Coomassie Brilliant blue. The slightly slower electrophoretic mobility of the REP27 protein in the complemented strain ( rep27 comp) is due to the presence of the MYC tag, introduced into the plasmid-construct used for complementation.


INTRODUCTION
The unicellular green alga Chlamydomonas reinhardtii is a good model system to study the regulation of photosynthesis at the molecular level since the chloroplast development and differentiation can take place either under autotrophic, photo-heterotrophic, or dark-heterotrophic conditions. The chloroplast biogenesis and most of the photosynthetic apparatus assembly can occur in the dark, when cells are supplied with organic carbon such as acetate (Guenther et al., 3 1990). Vegetative cells are haploid, permitting ready phenotypic manifestation of mutations, or genetic lesions. Photosynthesis deficient mutants can thus be isolated and investigated, conferring to Chlamydomonas a significant advantage over other model plant systems.
Measurement of the chlorophyll fluorescence with intact cells offers a non-invasive approach to assessing the functionality of photosystem-II (PSII) and of the electron transport process in the thylakoid membrane of photosynthesis. Thus, a number of photosynthesis mutants with defects in the biogenesis and assembly of thylakoid membrane complexes were generated and isolated (Zhang et al., 1997;Wollman et al., 1999;Minai et al., 2006), providing valuable information about the corresponding processes and leading to the isolation and characterization of genes and proteins.
The PSII repair cycle (Guenther and Melis, 1990) is a process essential to photosynthesis and plant growth, occurring in all organisms of oxygenic photosynthesis, and serving to restore the functional status of PSII from a frequently occurring photodamage. Repair entails the unique selective degradation and replacement of the D1/32 kD PSII reaction center protein from the massive (>1,000 kD) PSII holocomplex (Mattoo and Edelman, 1987). The PSII damage and repair mechanism is highly conserved in all organisms of oxygenic photosynthesis, as it maintains the activity of photosynthesis by selectively degrading and replacing the PSII D1/32 kD reaction center protein (Melis, 1991;Aro et al., 1993). The rate constant of photodamage is a linear function of light intensity (Baroli and Melis, 1996;Tyystjärvi and Aro 1996), ranging between 0 in the dark to about 1.2 h -1 under bright sunlight and physiological growth conditions. In contrast, the enzymatic repair process occurs with a light intensity-independent rate constant, equal to about 0.7 h -1 (Neidhardt et al., 1998;Ohnishi et al., 2005;Yokthongwattana and Melis, 2006). Under bright sunlight conditions, the rate of photodamage can be faster than the rate of repair, resulting in accumulation of inactive D1 proteins, loss of photosynthetic yield and chloroplast productivity (Adir et al., 1990;Bailey et al., 2002). The repair entails D1 activation (Guenther et al., 1990;Neale and Melis, 1991) and post-translational modifications to restore the PSII water-splitting activity (Diner et al. 1988;Bowyer et al., 1992).
Biogenesis of the photosynthetic apparatus is a process involving the coordinated expression of genes leading to the biosynthesis and assembly of both chloroplast-and nuclearencoded proteins. The chloroplast genome of the unicellular green alga Chlamydomonas reinhardtii encodes approximately 100 genes, required for protein synthesis of the 4 photosynthetic apparatus and carbon fixing machinery (Maul et al., 2002). Genetic and biochemical studies of Chlamydomonas revealed the involvement of numerous nucleus-encoded factors acting in the transcription/translation or in several post-transcriptional events of chloroplast gene expression, such as mRNA processing, stability and translation into proteins (Barkan and Goldschmidt-Clermont, 2000;Somanchi and Mayfield, 2001). Compared to the present information available on the rapid light-dependent turnover of the D1 protein in PSII (Aro et al., 1993;Yokthongwattana and Melis, 2006), our understanding of the regulation of the PSII repair mechanism is very limited, either at the level of protein translation or posttranslational steps leading to a functional PSII. The de novo synthesis, membrane insertion and assembly of D1 processes are most likely to require the participation of nuclear-encoded auxiliary proteins.
In earlier studies from this lab (Zhang et al., 1997;Park et al., 2007), DNA insertional mutagenesis in the model organism Chlamydomonas reinhardtii was applied for the isolation of mutants defective in photoautotrophic growth. Isolated from this screening protocol, the rep27 strain was found to grow normally in the presence of acetate, but displayed low photosynthetic activity even under low light conditions. Under weak growth irradiance, the rep27 showed a limited water oxidation and O 2 evolution capacity, whereas under moderate to high irradiance PSII activity ceased to exist, indicating inability of the chloroplast to perform the D1 protein turnover. Gene cloning and biochemical analysis of the rep27 strain resulted in the identification of the REP27 nuclear gene, which was deleted from the rep27 mutant. The REP27 cDNA sequence was isolated and used for complementation of the rep27 mutant, permitting the recovery of the wild-type phenotype (Park et al., 2007). From these preliminary results, it was suggested that REP27 is a nuclear gene involved in the rapid light-dependent turnover of the D1 protein during the PSII repair process.
The present work investigated and elucidated the mechanism of the REP27 protein action in the D1 protein turnover and PSII repair from photodamage. It was concluded that REP27 plays a dual role in the regulation of the D1 protein turnover by facilitating co-translational biosynthesis-insertion (C-terminal domain) and activation (TPR motifs) of the nascent D1 during the PSII repair process.

BN-and SDS-PAGE analysis of wild type and rep27 mutant
DNA insertional mutagenesis with the model organism Chlamydomonas reinhardtii was applied for the isolation and characterization of putative PSII repair mutants (Zhang et al., 1997;Park et al., 2007). The rep27 strain, defective in photoautotrophic growth, was isolated from this screening library. It was found that rep27 grew in the presence of acetate, but it displayed a lower level of photosynthetic activity under low light conditions. Under moderate to high light intensity, PSII photochemical activity was absent in this mutant. Further physiological and biochemical analyses showed lower steady-state levels of the D1/32-kD reaction center protein in the rep27 than in the wild type, while others subunits of the PSII holocomplex, such as D2, CP47, were only somewhat reduced. Peripheral subunits of PSII, such as the PsbO proteins, were in equivalent amounts in wild type and rep27. Subunits of non-PSII complexes (Cyt b 6 -f, PSI, and ATP synthase) also occurred in equivalent amounts in wild type and rep27 mutant (Park et al., 2007). Results from the earlier work supported a working hypothesis, whereby de novo biogenesis/assembly of the PSII holocomplex occurred more-or-less normally in the rep27 mutant; however, replacement (turnover) of the photodamaged D1/32 kD protein was severely impaired.
A comparative thylakoid membrane protein profile analysis of wild type and rep27 mutant was undertaken, aiming to identify biochemical and functional differences between the two strains. For better resolution of the integral and hydrophobic components of the thylakoid membrane, holocomplexes were separated in a first dimension via non-denaturing polyacrylamide gel electrophoresis. Fig. 1A (upper lane) shows a Blue-Native gel of wild type thylakoid membrane proteins, compared with the green native gel of the same sample (Fig. 1A, lower lane). The native form of the resolved holocomplexes was then separated into their constituent subunits upon running in a second dimension by denaturing SDS-PAGE and visualized by silver staining (Fig. 1B). Combination of BN-PAGE and SDS-PAGE was previously used to identify specific thylakoid membrane proteins of C. reinhardtii by peptide mass fingerprinting and MALDITOF-MS (Rexroth et al., 2003). This analysis permitted separation and identification of the main PSII core subunits (D1, D2, CP43 and CP47 proteins), PSI proteins (PsaA, PsaB and PsaF and light-harvesting complex I protein), the light-harvesting complexes of PSII in trimeric and monomeric forms, cytochrome b 6 -f complex proteins, and chloroplast ATP synthase complex (CF 0 F 1 ) proteins (Rexroth et al., 2003). A selective comparison of the 2-DE proteome of photo-mixotrophically grown Chlamydomonas wild type and rep27 mutant is given in Fig. 2. Demarcated by dashed line are subunits of the PSII supercomplex (Fig. 2, CC425). In the rep27 mutant, these proteins appeared to occur at substantially lower levels (Fig. 2, rep27) relative to the wild type. The 2-DE gel also showed a higher relative amount of LHCII proteins in the rep27 mutant relative to the wild type, demarcated by a dashed line in Fig. 2 (rep27). Since the overall Chl/cell ratio in the rep27 mutant was similar to that in the wild type (Zhang et al., 1997;Park et al., 2007) and the 2-d gels were loaded on the basis of equal Chl, the results suggest a higher LHCII/RCII ratio in the rep27 compared to that in the wild type. This is a consequence of the rep27 mutation, adversely affecting the RCII and PSII-core complexes but not the LHCII.
Quantitative Western blot analysis was employed to more directly compare levels of the PSII core and RC proteins in wild type and rep27 mutant. Fig. 3 shows that D1 and CP43 were depleted from the thylakoid membrane of the rep27 mutant, whereas D2 and CP47 occurred in The above proteome-based results are consistent with previous spectrophotometric and Western-blot analyses probing the steady-state level of PSII and D1 reaction center protein in Chlamydomonas wild-type and rep27 mutant (Zhang et al., 1997;Park et al., 2007).

Occurrence of REP27 in wild type and photochemical apparatus mutants
The cDNA sequence of REP27 (GenBank Accession Number EF127650) was used for complementation of the rep27 mutant, which permitted recovery of the wild-type phenotype.
Based on these findings, specific polyclonal antibodies were generated in rabbit against a portion of the REP27 protein, defined by amino acids F273-to-L367. SDS-PAGE and Western-Blot 7 analysis were then applied to probe the occurrence and steady-state level of REP27 proteins in Chlamydomonas reinhardtii wild type and selected photochemical apparatus mutants. Results from such comparative analysis are shown in Fig. 4, where a specific cross reaction was detected between the REP27 antibodies and a protein band migrating to 47 kD (Fig. 4, CC425). The REP27 molecular weight, obtained by this Western-blot analysis was similar to that deduced from the amino acid composition of the protein (45.6 kD, Park et al., 2007). The REP27 specific polyclonal antibodies failed to recognize any protein band in extracts from the rep27 mutant ( Fig. 4, rep27), consistent with the notion that the latter is a REP27 knockout. Conversely, rep27-complemented strains showed a distinct protein band in the ~48 kD region (Fig. 4,rep27comp) [the recombinant REP27 is larger by about 2 kD due to the presence of a 2xMYC tag]. It is of interest that both the D1-less (∆psbA) and D2-less (∆psbD) mutant strains of C. reinhardtii showed an equal to wild type abundance for the REP27 protein ( Fig. 4), in spite of the fact that these strains have no PSII reaction center complex, display no variable Chl fluorescence or oxygen evolution (Bennoun et al., 1986;Erickson, 1986;Minai et al., 2006) and, therefore, do not undergo a photodamage and repair cycle. It is suggested that the REP27 gene is expressed constitutively in Chlamydomonas reinhardtii under all growth conditions. Moreover, it appears that the REP27 gene transcription and translation are not down regulated by the absence of the D1/D2 reaction center proteins of PSII. It may be concluded that a lack of de novo biogenesis/assembly of the PSII reaction center does not adversely affect the synthesis of the REP27 protein.

Thylakoid membrane fractionation and localization of REP27
It was previously suggested that REP27 is a putative chloroplast-targeted protein (Park et al., 2007), as the REP27 amino acid sequence analysis (Emanuelsson et al., 1999)  mutant, and rep27-complemented strains, probed with specific polyclonal antibodies generated against the REP27, RbcL, or D1 proteins. Total cell extract (TE), soluble (S) and total membrane (TM) fractions were assayed. RbcL and D1 specific polyclonal antibodies served in testing for the enrichment of the soluble and total membrane fractions. Results showed absence of the REP27 protein from the total protein extract of the rep27 mutant (Fig. 5, TE), but presence in the CC425 and rep27-complemented strains (Fig. 5, REP27). Importantly, the REP27 protein was found in exclusive association with the total membrane extract of the CC425 and rep27complemented strains (Fig. 5, TM) and not with the soluble protein fraction (Fig. 5, S). Resolved thylakoid membranes and isolation of a light membrane fraction, including the chloroplast envelopes, was achieved by differential ultracentrifugation, as recently done in this lab (Lindberg and Melis, 2008). Western blot analysis of these membrane fractions with REP27-specific antibodies showed presence of the REP27 protein in the thylakoid membrane fraction and absence from the lighter envelope-containing fraction (results not shown). These results clearly suggest integral thylakoid membrane localization for the REP27 protein, probably spanning the thylakoid lipid bilayer through the TMH1 and TMH2 regions.
In an effort to more precisely localize the REP27 in the thylakoid membrane of photosynthesis, appressed and non-appressed thylakoid membrane domains were isolated upon mechanical fractionation of C. reinhardtii wild type (Neale and Melis, 1991). It was shown previously that the 10k fraction is enriched in thylakoid grana membranes and the 140k fraction in stroma exposed membranes, as evidenced from direct measurements of PSI and PSII content using P700, Q A and pheophytin quantitations (Neale and Melis, 1991). The differential distribution of the photosystems among such thylakoid membrane domains was also reflected in the Chl a/b ratio of the fractions, thus providing an easy to measure assay and convenient marker. Fig. 6 shows that the Chl a/b ratio was 2.60 in the total thylakoid extract, it was lower to 2.39 in the heavy grana-enriched 10k fraction, and higher at 3.24 in the light stroma-exposed thylakoid containing PSI-enriched 140 k fraction. Fig. 6 also shows Western-blot analysis of the various thylakoid membrane fractions with D1-specific ( Fig. 6A) and REP27-specific polyclonal antibodies (Fig. 6B). Measurement of the REP27/D1 ratio showed that the REP27 protein is relatively more prevalent in the 40k than in the 10k or 140 k fractions, suggesting that the membrane domain of the REP27 protein is intermediate to the fully appressed and stromaexposed thylakoids. These could be the "fret" domains of the thylakoid membrane (Morrissey et  , 1986), where the PSII repair takes place, localized adjacent to the appressed grana membranes, yet exposed to the stroma medium so as to permit the D1 protein turnover.

Functional role of the REP27 tetratricopeptide motifs and C-terminal domain
By using Inter-ProScan 13.1 software analysis (Quevillon et al., 2005), two distinct tetratricopeptide (TRP) motifs were also identified in REP27 (Park et al. 2007). These were termed TPR1A/TPR1B and TPR2A/TPR2B, and were shown to occur near the N-terminal end of the protein. TPR motifs form a compact unit of two helices interacting with each other in the antiparallel direction (Blatch and Lässle, 1999). Accordingly, the structure of the REP27 protein entails the N-terminus, followed by two tetratricopeptide repeat motifs (TPR1 and TPR2), the two transmembrane helices and the long C-terminal hydrophilic portion of the protein. On the basis of topology analysis, using the TMAP transmembrane program and the positive inside rule (Persson and Argos, 1994), we propose that the N-and C-termini, including the two TPR motifs, are exposed to the chloroplast stroma phase. A folding model, depicting the structural association of REP27 with the thylakoid membrane was thus constructed (Fig. 7).
To gain a better understanding on the function of REP27 during the D1/32-kD PSII reaction center protein turnover, the role of the TPR motifs and of the C-terminal region were investigated. It has been suggested that TPR motifs are involved in a variety of critical proteinprotein interactions in the living cell (Blatch and Lässle, 1999), and this is consistent with a putative functional role of the REP27 in the unique D1 replacement process. Therefore, REP27 cDNA constructs missing the TPR1 (pSLREP27-∆T1), TPR2 (pSLREP27-∆T2), both TPR1 and TPR2 (pSLREP27-∆T1+2), or the C-terminal region (pSLREP27-∆Ct) were made (Fig. 8, see also Materials and Methods). These were used for transformation of the rep27 mutant strain.
Complementation of the rep27 mutant (rep27-comp) with a full-length REP27 cDNA construct (pSLREP27-comp) was used as a control. The paramomycin resistance cassette was used as the first selectable marker for the isolation of transformants. Putative rep27-∆T1, rep27-∆T2, rep27-∆ T1+2, and rep27-∆Ct transformant strains were screened further via Western-blot analysis with specific polyclonal antibodies generated against the REP27 protein. Only strains with a positive expression of the modified REP27 protein were selected for further analysis. Positive transformant strains rep27-∆T1, rep27-∆T2, rep27-∆T1+2, and rep27-∆Ct were tested for photoautotrophic growth on both HSM and TBP minimal media (Fig. 9). It is known that a functional REP27 protein is needed for autotrophic growth of C. reinhardtii in minimal media (Park et al. 2007). Accordingly, wild type (Fig. 9A) and rep27-comp strains (Fig. 9C) grew viable colonies on both HSM and TBP plates. Conversely, the rep27 mutant (Fig. 9B) and transformant strains rep27-∆T1 (Fig. 9D), rep27-∆T2 (Fig. 9E), rep27-∆T1+2 (Fig. 9F), and rep27-∆Ct (Fig. 9G) all failed to rescue the acetate-requiring phenotype of the rep27 mutant. It is concluded that both TPR motifs and the C-terminal portion of the REP27 protein are needed for the completion of the PSII repair cycle and the survival of C. reinhardtii under photoautotrophic growth conditions.
To probe the status of the truncated REP27 protein in transformant strains rep27-∆T1, rep27-∆T2, rep27-∆T1+2, and rep27-∆Ct, these were grown on TAP media, followed by biophysical and biochemical analysis of the function of PSII. Growth in the presence of acetate (TAP) under weak illumination enables biosynthesis and assembly of functional PSII (Zhang et al. 1997;Park et al. 2007), i.e., under conditions when the rate of photodamage is rather slow (Melis 1999), thereby permitting accumulation of a smaller-than-wild-type fraction of functional PSII reaction centers. Fig. 10 (left panel) shows that wild type, rep27 mutant, and transformant strains employed in this work, all grew viable colonies and greened normally on TAP plates. To probe the functional properties of PSII in these strains, the fluorescence yield F V /F M ratio was measured. The F V /F M ratio provides a measure of the photochemical charge separation efficiency of the PSII reaction centers (Kitajima and Butler, 1975); therefore it offers an indication of the proportion of functional PSII reaction centers under in vivo conditions. The F V /F M ratio of the wild type was equal to 0.72, whereas F V /F M for the rep27 mutant (0.30) was only about 40% of that in the wild type, suggesting that only a fraction of the PSII reaction centers were functional in the rep27 mutant (growth on TAP). As a positive control for the recovery of the REP27 function, the rep27-comp strain showed elevated F V /F M ratio (=0.68) and similar to that of the C. reinhardtii wild type. The rep27-∆T1, rep27-∆T2, rep27-∆T1+2, and rep27-∆Ct transformants all showed low F V /F M ratios and similar to that of the rep27 mutant (see Fig. 10, right panel). It is concluded that the TPR motifs and C-terminal region of the REP27 protein are all necessary to restore the proportion of functional PSII reaction centers and electron transport capacity in the chloroplast of this green microalga.

11
Western-blot analysis was employed with the wild type, rep27, rep27-comp, rep27-∆T1, rep27-∆T2, rep27-∆T1+2 and rep27-∆Ct transformant strains to test for the presence and relative steady-state amounts of the REP27 and D1 proteins (Fig. 11). Consistent with earlier results (Park et al. 2007), the rep27 mutant lacked the REP27 protein and showed low levels of the D1 protein in its thylakoids (Fig. 11, rep27). The rep27-comp strain offered positive evidence of the REP27 protein and substantially enhanced levels of the D1 protein, which were similar to those in the wild type (Fig. 11, rep27-comp). This is consistent with the elevated F V /F M ratio (=0.68) in the latter. Interestingly, all truncated REP27 transformant strains (Fig. 11, rep27-∆T1, rep27-∆ T2, rep27-∆T1+2, and rep27-∆Ct) clearly showed presence of the modified REP27 protein in the cell. Moreover, they also showed elevated amounts of the D1 protein in the chloroplast thylakoid membranes. Specifically, strains rep27-∆T1 and rep27-∆T2 displayed levels of the D1 protein that were nearly equivalent to those in the wild type and rep27-comp. Strains rep27-∆ T1+2 and rep27-∆Ct contained substantial amounts of the truncated REP27 protein, however, they did not accumulate D1 protein to levels equivalent to that in the other transformants.
The results clearly showed that removal of TPR1, TPR2, TPR1+TPR2, or 61 amino acids from the C-terminal domain did not interfere with the synthesis and assembly of the REP27 protein in the thylakoid membrane. Moreover, these transformants had elevated steady-state amounts of the D1 protein in the respective thylakoid membranes. However, evidenced from the low F v /F M ratio and inability to grow on minimal media, it may be concluded that synthesis and assembly of the truncated REP27 to the thylakoid membrane is not sufficient by itself to restore function in these transformants. Presence of both TPR motifs, and of the C-terminal portion of the protein, is required for the completion of the D1 protein turnover during the PSII repair process. It is hypothesized (a) that the C-terminal portion of the REP27 protein is needed for the insertion of nascent D1 proteins in the D1-less PSII template, and (b) that TPR motifs are required for the "activation" of newly inserted and bound D1 proteins, which, in the case of the truncated REP27 transformants remained inactive. Such "activation" step was inferred earlier from biophysical studies (Guenther et al., 1990;Neale and Melis, 1990) and could involve a posttranslational modification and/or proper membrane deployment / folding of the nascent D1 protein during its assembly within the PSII template and subsequent maturation. Since the TPRless strains had a low PSII photochemical charge separation efficiency and lack of photoautotrophic growth capacity, it appears that TPR domains are needed to assure a full integration of nascent D1 proteins into the PSII reaction center template, leading to a functional PSII reaction center.
The rep27-∆Ct mutant showed the lowest steady-state level of D1 present, compared to the others rep27 transformant strains, but the level was still higher than that of the original rep27 mutant. It is most likely that the REP27 C-terminal domain plays an essential role in the de novo D1 biosynthesis at the level of ribosomal psbA mRNA translation and it may also be required for insertion of nascent D1 proteins in the D1-less PSII template. Taken together, these results illuminate, for the first time and in great detail, the role that is played by the REP27 protein in the D1 protein turnover and PSII repair from photodamage.

DISCUSSION
The D1 protein is subject to a frequent turnover, which far surpasses that of all other thylakoid membrane and PSII subunits. Turnover can take place at all light intensities, but is accelerated under increasing irradiance (Melis, 1999). The D1 protein turnover was demonstrated in pulse-chase experiments, in which the D1 was preferentially labeled over that of other thylakoid membrane proteins (Mattoo and Edelman, 1987;Schuster et al., 1988;Adir et al., 1990). In a previous study, presenting the isolation and characterization of the rep27 mutant strain (Park et al., 2007), comparative [35S]-sulfate labeling experiments showed that the rep27 mutant accumulated radio-labeled D1 in tandem with the accumulation of the other PSII subunits. However, and opposite to that in the wild type, it did not show the prolific and preferential accumulation of D1 over the other PSII subunits. Since the latter is evidence for active D1 turnover (Park et al., 2007), it was suggested that rep27 has a specific defect in the active turnover of this PSII reaction center protein. Furthermore, steady-state levels of psbA and psbD mRNA, measured by northern-blot analysis, were similar in the wild-type and mutant.
From these results, it was proposed that the rep27 mutant is defective in the D1 protein turnover and, therefore, unable to complete the PSII repair process. The comparative thylakoid membrane functional and protein analysis of wild type and rep27 mutant in this work confirms and strengthens this hypothesis. Since the REP27 knockout mutant contains functional PSII units, when grown in the presence of acetate (Fig. 10) and under weak irradiance, it is likely that the REP27 protein does not play a direct role in the de novo biogenesis and assembly of the PSII holocomplex. This conclusion is strengthened by the finding that REP27 is localized on fully assembled thylakoid membrane domains, in which there is a fully assembled PSII and electron-transport apparatus , and where turnover of the D1 protein is the only function requiring "assembly". On the contrary, de novo biosynthesis and assembly of PSII, and of thylakoid membranes, is expected to take places at the chloroplast "polar regions", where thylakoid membranes begin to emanate, and probably far away from the fully assembled grana complexes. The de novo biogenesis/assembly of PSII was in fact proposed to occur in areas of the chloroplast where there is a low density of membranes in a process of assembly biochemically related to the chloroplast inner envelope (Zerges and Rochaix, 1998). The spatially separated de novo biogenesis/assembly of PSII from the D1 repair machinery explains why substantial residual O 2 -evolving photosynthetic activity is encountered even after the knockout inactivation of REP27. Consistent with this reasoning is also the finding that regulation of psbA translation by auto-feedback repression for PSII biogenesis/assembly is clearly distinct from the D1 biosynthesis for PSII repair (Minai et al., 2006). Therefore, we conclude that the rep27 mutant performs biogenesis/assembly of a functional PSII holocomplex but fails to undertake the D1 protein turnover, as would be required by the PSII repair process.
Regulation of D1 biosynthesis takes place primarily during the psbA mRNA translation, with translation initiation, elongation and co-translational assembly of the D1 protein into PSII all being regulated (Kettunen et al. 1997;Zhang et al., 1999;Zhang et al., 2000;Zhang and Aro, 2002). Association of cofactors to the D1 protein during its de novo synthesis has been suggested to stabilize and properly fold the nascent D1 chain (Kim et al. 1994). Adding to this earlier information, results from the present work showed how the REP27 protein plays a role in facilitating, and probably regulating, different stages of the de novo D1 biosynthesis, assembly, and activation during the PSII repair process. For example, the REP27 C-terminal is essential for the psbA mRNA translation initiation and assembly of the nascent D1. The TPR domains of REP27 are required for the "activation" of the bound D1 in the PSII reaction center during the PSII repair process. 14 The REP27 is not the only TPR motif-containing chloroplast protein in Chlamydomonas reinhardtii, which is implicated in posttranscriptional steps of the chloroplast gene expression.
The nuclear-encoded TPR protein Mbb1 is involved in psbB mRNA processing, stability and translation (Vaistij et al., 2000). Moreover, the Nac2 factor is required for the stabilization, processing and translation of the psbD mRNA, permitting the proper folding of the D2 protein during the de novo biosynthesis-assembly of PSII (Boudreau et al., 2000). It is interesting to observe that the REP27 has a similar functionality with the Mbb1 and Nac2 proteins (i.e., mediating protein-protein interactions) during chloroplast mRNA metabolism, even though the three proteins probably act independently in different signaling pathways.
Other possibly significant TPR motif-containing proteins that play a role in photosynthesis include the periplasmic Prat1 in Synechocystis sp. PCC 6803 (Klinkert et al., 2004;Schottkowski et al., 2009) and the thylakoid membrane-bound LPA1 in Arabidopsis thaliana (Peng et al., 2006). Klinkert et al. (2004) showed that targeted inactivation of PratA There is greater similarity between the deduced amino acid sequence alignment of REP27 and LPA1 (CLUSTAL 2.0.8 analysis, supplemental Fig. 1S, B). LPA1 has been localized to the thylakoid membrane of Arabidopsis chloroplasts (Peng et al. 2006). However, and somewhat at variance with the conclusions drawn from this work, the latter authors assigned a primary function to LPA1 in the de novo biosynthesis and assembly of the PSII holocomplex, rather than to a specific function in the turnover of the D1 protein, occurring during the PSII damage and repair cycle. More work is needed to delineate between these alternative possibilities of the three proteins in the different photosynthetic systems. In conclusion, REP27, a nuclear encoded protein, is essential for the D1 reaction center protein turnover, permitting a completion of the translation process, maturation and activation of D1 into a functional PSII reaction center holocomplex. Figure 12 is a schematic illustration of the role of the REP27 protein during the PSII repair from photodamage. It portrays the critical role played by the REP27 protein in the translation of the psbA mRNA, insertion of the nascent D1 protein in the D1-less PSII template, and in the activation of the newly assembled reaction center complex. According to this proposed mechanism, REP27 C-terminal is permitting the initiation of ribosomal psbA mRNA translation and protein insertion, whereas the TPR motifs enable functional activation of the newly assembled D1 within the existing PSII template.

Growth media and culture conditions
Wild-type, rep27 mutant and related transformants of the green alga C. reinhardtii were grown mixotrophically in acetate-containing TAP media (Gorman and Levine, 1965) at 25°C under illumination of 50 µmol photons m -2 s -1 provided by cool white fluorescent lamps. Algal cultures in early exponential growth phase were used for experiments, with cells either in liquid culture or on 1.5% agar plates. To test photoautotrophic growth of strains, cells were grown on TBP minimal media, in which sodium bicarbonate (25 mM, pH 7.4) replaced the acetate as the growth carbon source (Polle et al., 2003). Cells were collected by centrifugation at 5,000 x g for 5 min at 20ºC. Cells pellet was stored at -80ºC, or used immediately for extraction of total proteins. Chlorophyll concentration was determined in 80% acetone extracts according to Arnon (1949), with equations corrected as in Melis et al. (1987). Each experiment was repeated three times with independently grown cell cultures.

Mutant strain generation
Generation of truncated REP27 proteins and the corresponding cDNA constructs were implemented using the QuikChange II XL Site-directed Mutagenesis Kit (Stratagene), according to the manufacturer's instructions. Oligonucleotides (Bioneer, CA) carrying the desired mutations are listed in Table 1. Plasmids carrying the targeted mutations were identified by sequencing, isolated and reintroduced into pSL18 for complementation of C. reinhardtii rep27 mutant. Deletion of the specific amino acids from the mature REP27 protein in the various constructs is as follows: ΔTPR1, A58-E87; ΔTPR2, A97-Y126; ΔTPR1+2, A58-Y126; ΔCt, L389-E449.

Blue-Native gel electrophoresis (BN-PAGE)
Thylakoid membranes were diluted to 0.5 mg Chl per ml in BN-PAGE solubilization buffer (50 mM Bis-Tris-HCl, pH 7.0, 750 mM ε -amino-n-caproic acid and 20% glycerol according to Schägger et al. (1994). Dodecyl-β-d-maltoside was added to a final concentration of 1% (w/v), solubilization was carried out on ice for 40 min, followed by centrifugation at 1,000xg for 10 min. The supernatant was supplemented with Serva blue G-250 from a 5% (w/v) stock in 500 mM ε -amino-n-caproic acid to a detergent/Serva blue G-250 ratio of 4:1 (w:w) and directly loaded onto the gel. BN-PAGE was carried out in a 5-12.5% gradient by using the Hoefer SE 600 and Hoefer SE 250 electrophoresis apparatuses according to Schägger et al. (1994) with modifications according to Thidholm et al. (2002).

Denaturing SDS-poly-acrylamide gel electrophoresis (SDS-PAGE)
For the isolation of total protein, cell biomass equivalent to 100 μ g Chl were resuspended in 400 μ L of 0.1 M dithiothreitol and 0.1 M Na 2 CO 3 . Following incubation for 5 min, 400 μ L of 2× sample solubilization buffer containing 250 mM Tris-HCl (pH adjusted to 6.8), 7% SDS, 20% glycerol, 2 M urea and 10% β -mercaptoethanol was added and incubated for 1 h at room temperature. Unsolubilized material was removed by centrifugation at 15,000xg for 5 min prior to loading samples onto the SDS-PAGE.
Aliquots corresponding to an equal amount of Chl were loaded in the wells of the stacking gel and electrophoresed through the 12.5% SDS-polyacrylamide running gels containing 1 M urea, as described by Tetali et al. (2007).

Generation of REP27 specific polyclonal antibodies and Western-blot analysis
Specific polyclonal antibodies were generated in rabbit against the REP27 protein using a Cterminal portion of the REP27 protein as recombinant antigen. Nucleotide fragments corresponding to F273-L367 of the C. reinhardtii REP27 were amplified using the primers listed in Table 1  horseradish peroxidase conjugated secondary antibodies (Bio-Rad). Specific polyclonal antibodies against the D1/32 kD PSII reaction center protein were also employed, as described by Park and Rodermel (2004). Cross-reactions between protein bands and antibodies were visualized by the Supersignal ECL (Pierce) detection kit following the manufacturer's specifications.

Measurement of photosynthetic parameters
The maximum quantum yield of primary PSII photochemistry was determined from  et al., 1995).

Cell and thylakoid membrane fractionations
Algal cells were harvested upon centrifugation (Beckman JA-10 Rotor) at 5,000 × g for 5 min.
Pelletted cells were resuspended in sonication-buffer containing protease inhibitors (50 mM Tricine/NaOH, pH 7.8, 10 mM NaCl, 5 mM MgC1 2 , 1 mM aminocaproic acid, 1 mM aminobenzamidine, 0.1 mM phenylmethylsufonylfluoride and 2 mM Na-ascorbate) and sonicated on ice 3 times for 60 s in a 50 % duty cycle pulse mode, with 60 s cooling intervals in-between (Branson sonifier). The crude homogenate was then centrifuged at 3,000 x g for 5 min in order to remove unbroken cells and large cell fragments. To separate membrane-bound from soluble proteins, the crude homogenate was subjected to centrifugation at 20,000 g for 60 min. The pellet was used as the total thylakoid membrane fraction. The supernatant was used as the soluble fraction. Appressed and non-appressed thylakoid membrane regions were isolated upon mechanical fractionation and differential centrifugation, as described previously (Neale and Melis, 1991). Thylakoid membrane vesicles were precipitated by differential centrifugation at 10,000 × g (10K fraction), 40,000 × g (40K fraction) and 140,000 × g (140K fraction). The Chl a/b ratio was determined in all fractions to provide a measure of the differential enrichment of grana and stroma-       Western-blot analysis of total proteins from Chlamydomonas reinhardtii wild type (CC425), rep27 and rep27-comp strains. Cells were fractionated into soluble (S) and total membrane (TM) portions. Total extract (TE) was obtained upon breaking the cells with acidwashed glass beads. The total membrane fraction (TM) was obtained by centrifugation at 20,000 g for 60 min. Lanes were loaded with 2 μ g Chl. The supernatant was used as the soluble fraction (S). Relative protein content was estimated from the intensity of the antibody cross-reaction in the Western-blots. Specific polyclonal antibodies were used, generated against REP27, RbcL or D1, respectively.  (TMH1 and TMH2) are shown spanning the thylakoid membrane such that N-and C-termini are exposed in the chloroplast stroma phase. The model further shows two tetratricopeptide repeat motifs (TPR1 and TPR2) occurring near the N-terminus, and the transit peptide that is cleaved upon chloroplast import. REP27 displays a rather extensive C-terminal portion, which is exposed in the chloroplast stroma.  transformant strains grown on HSM and TBP medium, respectively. A, CC125 wild type; B, rep27 mutant; C, rep27-comp; D, rep27-∆T1; E, rep27-∆T2; F, rep27-∆T1+2; G, rep27-∆Ct.   C-terminal is essential for the de novo D1 biosynthesis at the level of ribosomal psbA mRNA translation and initial assembly in the PSII template. The TPR motifs participate in posttranslational modification (D1 activation). Both TPR motifs are needed to activate the nascent D1 and to confer functional status to the PSII holocomplex.