CP29, a Monomeric Light-harvesting Complex II Protein, Is Essential for State Transitions in Chlamydomonas reinhardtii*

In oxygen-evolving photosynthesis, the two photosystems, photosystem I (PSI) and photosystem II (PSII), function in parallel, and their excitation levels must be balanced to maintain an optimal photosynthetic rate under various light conditions. State transitions balance excitation energy between the two photosystems by redistributing light-harvesting complex II (LHCII) proteins. Here we describe two RNA interference (RNAi) mutants of the green alga Chlamydomonas reinhardtii with one of the minor monomeric LHCII proteins, CP29 or CP26, knocked down. These two proteins have been identified in PSI-LHCI supercomplexes that harbor mobile LHCII proteins from PSII under a state where PSII is preferentially excited (State 2). We show that both the CP29 and CP26 RNAi mutants undergo reductions in the PSII antenna size during a transition from State 1 (a state where PSI is preferentially excited) to State 2, as reflected by nonphotochemical quenching of fluorescence, low temperature fluorescence spectra, and functional absorption cross-section. However, the undocked LHCIIs from PSII do not re-associate with PSI in the CP29-RNAi (b4i) mutant because the antenna size of PSI was not complementary increased. The mobile LHCIIs in the CP26-RNAi (b5i) mutant, however, re-associate with PSI, whose PSI-LHCI/II supercomplex is visualized on a sucrose density gradient. This study clarifies that CP29, not CP26, is an essential component in state transitions and demonstrates that CP29 is crucial when mobile LHCIIs re-associate with PSI under State 2 conditions.

In photosynthetic organisms, the conversion of light energy into electrochemical energy is driven by two photosystems, photosystems I (PSI) 2 and II (PSII). These photosystems possess peripheral antenna systems that are termed light-harvesting complexes I (LHCI) and II (LHCII), respectively. To opti-mize photosynthetic electron transport efficiency and avoid damage during light stress conditions, plants and green algae control the distribution of excitation energy between the two photosystems (state transition) (1,2). When the plastoquinone pool is reduced by preferential excitation of PSII (3), a protein kinase that phosphorylates LHCII bound to PSII is activated through the cytochrome bf complex (4,5). The phosphorylation of LHCII leads to the lateral migration of LHCII to the PSI in the unappressed region of thylakoids, where it acts as the peripheral antenna for PSI (State 2). Oxidation of the plastoquinone pool induces the opposite effect, recovering State 1 (6). The peripheral antenna system for PSII consists of major and minor LHCII proteins as outer trimeric and inner monomeric antennas, respectively (7). Single particle image analysis of electron micrographs revealed that these peripheral antenna proteins are bound to both sides of the PSII core, and each consists of one LHCII trimer and two LHCII monomers in higher plants (8) and green algae (9). CP29 and CP26 are the two conserved minor monomeric LHCII proteins among higher plants and green algae, whereas another minor monomeric LHCII, CP24, is only present in higher plants (10). Recent reports have suggested pivotal roles for minor monomeric LHCII(s) for state transitions in the unicellular green alga Chlamydomonas reinhardtii. A transfer of the cells from State 1 to State 2 increases the number of phosphorylated residues in CP29 from 2 to 4, and these residues were mapped at the interface of the PSII core and peripheral antenna proteins (11). CP29 and CP26 are dissociated from the PSII core upon phosphorylation (12). Furthermore, the two minor LHCIIs, in addition to one major LHCII type II (also known as LhcbM5), are associated with PSI under State 2 conditions (13). Because these minor LHCIIs border major LHCII trimers as well as the PSII core (14), any LHCII proteins associated with PSII are forced to dissociate when the minor LHCIIs are undocked. Thus, it has been hypothesized that both major trimeric and minor monomeric LHCIIs migrate from PSII to PSI, and the minor monomeric LHCIIs act as linker proteins between the former and the PSI core (13,15). A recent report of an Arabidopsis thaliana mutant lacking CP26 shows plants undergo normal state transitions without CP26 (16). It is not clear whether CP26 is essential for driving state transitions in green algae, because it might be a passenger in the mobile LHCIIs and passively associated with the PSI-LHCI/II supercomplex in State 2. Currently, it is not known whether CP29 is dispensable for state transitions.
In this study, we generated individual RNAi mutants in C. reinhardtii that knock down either CP29 or CP26, and we characterized their traits. Both mutants were studied to examine whether they undergo state transitions by several potential methods, including phosphorylation of LHCII proteins, fluorescence quenching analysis, antennae measurements, fluorescence emission and excitation spectra, and formation of a PSI-LHCI/II supercomplex. The results clearly demonstrate that the peripheral LHCII proteins in the CP29-RNAi (b4i) mutant dissociate from PSII but do not re-associate with PSI under State 2-promoting conditions. This observation indicates that CP29 is an essential component for state transitions in C. reinhardtii. The CP26-RNAi (b5i) mutant behaves like the wild type (WT), differentiating the role of CP26 in state transitions. We also discuss the low fluorescence nature of the free LHCIIs that appeared in the b4i mutant in State 2.

EXPERIMENTAL PROCEDURES
Strains and Media-The C. reinhardtii WT strain TW3 (thi10 cw15 mt ϩ ) and a state transition-less mutant DLS⌬ (17) were acquired from Drs. K. Shimogawara and F.-A. Wollman, respectively. These control strains and the two RNAi mutants generated in this study (b4i and b5i) were grown photoautotrophically in a high salt minimal medium (18) supplemented with 10 ng/ml thiamine and 1.5 mM L-tryptophan, with bubbling 5% (v/v) CO 2 -enriched air at 23°C under moderate light conditions (120 microeinsteins m Ϫ2 s Ϫ1 ).
Plasmid Construction-To generate the inverted repeat Lhcb4 and Lhcb5 constructs, short fragments corresponding to the N-terminal part of CP29 and CP26 were PCR-amplified using the WT cDNA as a template and the primers b4-75F (gaggatccgaattccctcccggcacccagaagaa; the BamHI and EcoRI sites are underlined), b4-408R (tcgtctgcagtctccgagctaacctcatcg; the PstI site is underlined), b5-76F (tgggatccgaattccccgccaagaacggcactgc; the BamHI and EcoRI restriction sites are underlined), and b5-248R (ctgtctgcagggatctccgagcggtcgta, the PstI site is underlined). Longer fragments were amplified using the primer pairs b4-75F and b4-318R (ctgtctgcagcgccaatcaccacgaactcc; the PstI site is underlined) and b5-76F and b5-355R (gtctctgcagcgtgcagcagctcgttctcg; the PstI site is underlined). These longer regions contained the adjacent coding sequence that functions as a loop structure in the inverted repeats. Both fragments were cloned in the antisense direction into the multiple cloning sites of the vector pBluescript II SK(ϩ) (Stratagene). The resulting inverted repeat cassettes were excised with EcoRI and inserted into the vector NE537 (19), kindly provided by Dr. H. Cerutti, to make NE537/IRLhcb4 and NE537/IRLhcb5.
Isolation of Photosynthetic Protein Complexes-Thylakoid membranes were isolated as described previously (13). A slightly modified buffer containing 25 mM MES, 0.33 M sucrose, 5 mM EDTA, and 1.5 mM NaCl (pH 6.5) was used to suspend the membranes. Cells were disrupted two times by BioNeb (Glas-col, Terre Haute, IN) at 2.5 kg/cm 2 . Sucrose density gradient ultracentrifugation was performed as described previously (13).
SDS-PAGE, Two-dimensional Electrophoresis, and Immunoblotting-SDS-PAGE, two-dimensional electrophoresis, and immunoblotting were conducted as described previously (12). The phosphorylated polypeptides were detected with an antibody against phosphothreonine (Zymed Laboratories Inc.).
MS/MS Analysis-MS/MS analysis was performed with a Finnigan LTQ linear ion trap mass spectrometer (Thermo Fisher Scientific) and SEQUEST software (revision 3.3, Thermo Fisher Scientific) as described previously (12).
Fluorescence Quenching Analysis-Traces of fluorescence yield were recorded under light-dark cycles as reported previously (22).
Fluorescence Induction and Relaxation Fluorimetry-The functional absorption cross-section ( PSII ) that reflects the functional antennae size of PSII was measured using a FIRe fluorometer system (Model FIRe, Satlantic, Nova Scotia, Canada) as described previously (12).
Low Temperature Fluorescence Emission and Excitation Spectra-Both fluorescence emission and excitation spectra were measured at 77 K using a Hitachi F-2500 fluorescence spectrophotometer with a red-sensitive photomultiplier, R-928F (Hamamatsu Photonics, Hamamatsu, Japan). For fluorescence emission spectra, cells were excited at 440 nm (slit ϭ 10 nm), and fluorescence was monitored between 650 and 760 nm (slit ϭ 5 nm). Fluorescence emission spectra were normalized to the maximum fluorescence yields simultaneously measured at room temperature. Excitation spectra for the PSI fluorescence at 718 nm were measured with an excitation scan between 620 and 690 nm and normalized to 697 nm, the position of the PSI absorption band.

Efficient and Specific Inhibition of CP29 and CP26 Expression-
To determine the specific roles of CP29 and CP26, we generated individual mutants devoid of each of these minor monomeric LHCII proteins. RNAi has been recognized as an effective method to knock down genes in organisms whose genomes are not readily amenable to gene targeting (23,24), and it has also recently become a powerful tool in studies on C. reinhardtii (24 -26). To knock down CP29 and CP26, we constructed plasmids carrying hairpin RNA sequences corresponding to a sequence from the Lhcb4 and Lhcb5 genes, respectively. The hairpin RNA sequences (consisting of Ϸ200 bp from the 5Ј-translated region stem structure and Ϸ100 bp of loop structure) were inserted downstream of the RBCS2 promoter into the plasmid NE537; this plasmid also harbors a hairpin RNA sequence for the tryptophan synthase ␤-subunit, which helps in screening for targeted gene interference (19). C. reinhardtii transformants with the resultant plasmids NE537/b4i and NE537/b5i were selected using 5-fluoroindole resistance. The amount of CP29 and CP26 proteins in these transformants was quantified by Western blot analysis using an antibody raised against each protein (Fig. 1A). Mutants showing a near complete inhibition of the CP29 and CP26 proteins were named b4i and b5i, respectively, and were used throughout this study. The other LHCII proteins appear not to be affected (supplemental Fig. S1 and supplemental Table S1). These RNAi mutants show more extensive suppression of the target proteins than our previous RNAi lines (27), whose state transition ability was briefly mentioned in Kargul et al. (15).
Phosphorylation of LHCIIs during a Transition to State 2-LHCII phosphorylation is crucial for the migration of LHCIIs from PSII to PSI during state transitions (28). We first tested phosphorylation of the LHCII proteins in vivo under State 2-promoting conditions to examine the effects of the RNAi (Fig. 1B). A state transition-less mutant DLS⌬ in which the cytochrome bf complex subunits PetD and PetL are fused (17) and WT were used as a negative and positive control, respectively. Normal LHCII phosphorylation, except for each of the targeted minor LHCIIs, was observed in the two RNAi mutants as well as in WT under State 2 conditions. This normal phosphorylation was not the case for the negative control DLS⌬, where all LHCII proteins were only minimally phosphorylated. Both RNAi mutants thus meet a prerequisite for LHCII migration.
Dissociation of the Mobile LHCIIs from PSII-We performed three sets of fluorescence experiments, fluorescence quenching, functional absorption cross-section of PSII ( PSII ), and low temperature fluorescence spectra, to investigate whether the mobile LHCII proteins dissociate from PSII during state transitions. In WT, the maximal yield of fluorescence (F m and F m Ј ) decreases upon shifting the cells from light to dark (Fig. 2). The observed decrease in fluorescence corresponds to an induction of q T quenching (22,29,30). Both RNAi mutants exhibited q T quenching as WT (Fig. 2). On the other hand, the DLS⌬ mutant showed no decrease in maximal fluorescence yield (Fig. 2). Likewise, decreases in PSII were observed in both the b4i and the b5i mutants, as in WT (Table 1). Intriguingly, the extent of q T quenching and decrease in PSII in the b4i mutant was large compared with WT and the b5i mutant (Table 1); this observation suggests that a larger number of LHCIIs are undocked from PSII during state transitions in the absence of CP29.
Re-association of the Mobile LHCIIs to PSI-During state transitions, undocked LHCIIs from PSII re-associate with PSI to form a PSI-LHCI/II supercomplex (31,32). Functional coupling of the mobile LHCIIs to PSI is reflected in an increase in the fluorescence emission peak at 718 nm at 77 K (33). The b5i mutant showed a substantial increase in the fluorescence emission at 718 nm under State 2-promoting conditions (Fig. 3A), indicating that the mobile LHCIIs could be functionally coupled to PSI in the absence of CP26. However, the b4i mutant showed no increase in the PSI fluorescence, although a decrease in PSII fluorescence was observed (Fig. 3A). The DLS⌬ mutant showed no change in either PSII or PSI fluorescence during state transitions.
Another line of evidence for the noncoupled mobile LHCIIs in the b4i mutant was provided by the fluorescence excitation spectra at 77 K. Samples were scanned by excitation from 620 to 690 nm, and the fluorescence was monitored at the PSI-specific emission at 718 nm to record spectral components transferring excitations to PSI. To visualize additional spectral components that transfer excitations to PSI only in State 2, the excitation spectra in State 1 was subtracted from that of State 2 (Fig. 3B). The difference spectra for WT shows two peaks, one at 650 nm (chlorophyll b) and the other at 678 nm (chlorophyll a), which is superim-  4). B, phosphorylated polypeptides in WT, b4i, b5i, and DLS⌬ cells in State 1 (S1) and State 2 (S2) were detected with anti-phosphothreonine antibody.

TABLE 1 Relative antenna sizes during state transitions
The relative maximal fluorescence yield (F m Ј ) and an absorption cross-section of PSII ( PSII ) of the WT, b4i, b5i, and DLS⌬ cells in State 1 (S1) and State 2 (S2) shown in Fig. 2 MARCH 20, 2009 • VOLUME 284 • NUMBER 12 posable on the absorption spectra of purified LHCII proteins (Fig. 3B). Thus, in State 2, certain LHCIIs feed energy to component(s) that has a fluorescence peak at 718 nm in WT. Whereas the b5i mutant showed the same difference spectra as WT, the b4i mutant displayed no spectral differences. Like the DLS⌬ mutant, this CP29-less mutant has no LHCII that could feed energy to PSI in State 2 (Fig. 3B). These data strongly indicate that the undocked LHCIIs are not functionally re-coupled with PSI in the absence of CP29, whereas a normal state transition occurs even in the absence of CP26.

CP29 Is Essential for Photosynthetic State Transitions
Formation of PSI-LHCI/II Supercomplex in State 2-To further examine physical associations of the mobile LHCIIs to PSI, we performed sucrose density gradient ultracentrifugation of the thylakoid membranes from the control and mutant strains under State 2-promoting conditions (Fig. 4A). In WT, the chlorophyll protein complexes were separated into four green bands, as reported previously (Fig. 4A); A1, A2, A3, and A3Ј bands represent fractions containing free LHCII, the PSII core complex, the PSI-LHCI supercomplex, and the PSI-LHCI/II supercomplex, respectively (13). Whereas the A3Ј band was present in the thylakoids from both WT and the b5i mutant, it was absent in the b4i mutant under State 2 conditions (Fig. 4B). The PSI-LHCI/II supercomplex in the b5i mutant contains all the LHCIIs associated with the supercomplex in WT except for the knocked out CP26 (Fig. 4B). A pool of phosphorylated CP26, CP29, and LHCII type I, which are dissociated from PSII during a transition to State 2 (12), was detected in the PSI-LHCI/II supercomplex fraction in WT and the b5i mutant; however, all of the phosphorylated LHCII proteins were found in the free LHCII fraction in the b4i mutant (Fig. 5). These results supported that the mobile LHCIIs do not re-associate with PSI in the absence of CP29.

DISCUSSION
In this study, we demonstrated that one of the minor monomeric LHCII proteins CP29 is crucial for the formation of the PSI-LHCI/II supercomplex during state transitions. In our previous report using a WT strain, CP29 and CP26 were attached to PSI in State 2 (13) (see Fig. 4). Their tight association with the PSI core suggested that they have roles as linkers between the PSI core and the major LHCII trimers within the PSI-LHCI/II supercomplex. The present result is important because it shows CP29 is essential for state transitions, whereas another minor monomeric LHCII CP26 is not. The results obtained with the b5i mutant agree with the recent report on the koCP26 mutant in A. thaliana (16). Furthermore, the specific role of CP29 dur-  Fig. 2) and State 2-promoting conditions (thick line, measured at a point indicated by the closed triangle in Fig. 2). Samples were excited at 440 nm. Spectra were normalized to the maximum fluorescence yields simultaneously measured at room temperature. B, State 2-specific excitation spectra for the PSI fluorescence. Excitation spectra for the PSI fluorescence at 718 nm were normalized to 697 nm, the position of the PSI absorption band, and differences between the excitation spectra under State 2-and State 1-promoting conditions were calculated. For comparison, absorption spectra of purified LHCII are shown (green). . Chlorophyll protein complexes separated on a sucrose density gradient. A, thylakoids from WT, b5i, and b4i cells prepared under State 2-promoting conditions were solubilized and subjected to a sucrose density gradient ultracentrifugation. The green bands were designated A1, A2, A3, and A3Ј as described previously (13). B, immunoblot analysis of the A3Ј fractions were performed with specific antibody. Proteins with 2.5 g of chlorophyll were loaded in each lane. Polypeptides of the A1, A2, A3, and A3Ј bands from the State 2 thylakoids were subjected to immunoblotting with an anti-phosphothreonine antibody. The sample representing the A3Ј band in the b4i mutant, which was not observed, corresponds to the fraction equivalent to the A3Ј band in WT and b5i mutant.
ing state transitions was clarified in this study. It is required for re-association of the mobile LHCIIs with PSI (the latter step of transition from State 1 to 2) and not for dissociation of the mobile LHCIIs from PSII (the former step of transition from State 1 to 2). Previously, an electron density near a PSI subunit PsaH was observed on a single particle image of the PSI-LHCI/II supercomplex; this density was ascribed to CP29 because it was the only LHCII protein that was present in the sample (15). CP29 is a hyperphosphorylated protein under State 2 conditions (11,13,15), and phosphorylated CP29 is indeed undocked from PSII upon phosphorylation (12). The current results corroborate previous reports and further provide evidence for the specific role of CP29 in state transitions.
Interestingly, the mobile LHCIIs were dissociated from the PSII core even in the absence of CP29 during a transition from State 1 to State 2; this dissociation was shown via an induction of q T quenching (Fig. 2), a decreased fluorescence emission at 680 nm (Fig. 3A), and a reduced functional absorption crosssection of PSII (Table 1). It is thus inferred that the other phosphorylated protein(s), including another minor LHCII CP26 and the PSII core subunits CP43 and D2, are sufficient to undock the mobile antennas. In fact, the mobile LHCIIs are even more extensively dissociated from PSII in the absence of CP29 ( Fig. 2 and Fig. 3A). Because CP29 locates at the border between PSII core and LHCII trimers (34), its absence could make the binding of LHCII trimers rather weak. The unstable PSII-LHCII supercomplex was also reported in an antisense line of CP29 in A. thaliana, where it was impossible to isolate the PSII-LHCII supercomplex (14). The extended dissociation of mobile LHCIIs from PSII observed in the b4i mutant is compatible with these previous findings.
The b4i mutant is the second state transition mutant that has a defect in the photosynthetic components. The first mutant carries a mutation in PsaH in A. thaliana (35). This mutant did not increase the PSI antenna size under State 2-promoting conditions, which led the authors to propose that PsaH is a docking site of the mobile LHCIIs for PSI. This is an attractive model because this subunit is located on one side of the PSI core, whereas the PsaG/F/J/K subunits (where the "LHCI-belt" is bound) are located on the opposite side (36). Indeed, Kouřil et al. (37) and Kargul et al. (15) observed an electron density near PsaH in their single particle analysis of the State 2 samples from A. thaliana and C. reinhardtii, respectively. Although the PsaH mutant and the b4i mutant have a similar phenotype, such as impaired state transitions because of blocked mobile LHCII binding to PSI, the fate of the mobile LHCIIs is clearly distinct. Whereas the mobile antennas stay on PSII in the former mutant, they are undocked from PSII and do not re-associate with PSI (Fig. 4) and PSII ( Fig. 2 and Fig. 3A) in the latter mutant. The difference could be the result of the structural instability of the PSII supercomplex caused by a lack of CP29 in the latter mutant.
In the b4i mutant, "free LHCIIs" likely exist under State 2-promoting conditions. Such unbound LHCII proteins have been thought of as highly fluorescent because they have nothing to transfer the energy to (38). Surprisingly, they exhibited low fluorescence under State 2-promoting conditions (q T quenching; Fig. 2), suggesting that the free LHCIIs are in a quenching state. q T quenching observed here could be accounted for by a simple undocking of LHCIIs from PSII and their conversion to a quenching state, although q T quenching has been conventionally interpreted as fluorescence lowering resulting from migration of peripheral antenna from PSII (a high fluorescence center) to PSI (a low fluorescence center).