Evolution of a divinyl chlorophyll-based photosystem in Prochlorococcus

Abstract

Acquisition of new photosynthetic pigments has been a crucial process for the evolution of photosynthesis and photosynthetic organisms. In this process, pigment-binding proteins must evolve to fit new pigments. Prochlorococcus is a unique photosynthetic organism that uses divinyl chlorophyll (DVChl) instead of monovinyl chlorophyll. However, cyanobacterial mutants that accumulate DVChl immediately die even under medium-light conditions, suggesting that chlorophyll (Chl)-binding proteins had to evolve to fit to DVChl concurrently with Prochlorococcus evolution. To elucidate the coevolutionary process of Chl and Chl-binding proteins during the establishment of DVChl-based photosystems, we first compared the amino acid sequences of Chl-binding proteins in Prochlorococcus with those in other photosynthetic organisms. Two amino acid residues of the D1 protein, V205 and G282, are conserved in monovinyl chlorophyll-based photosystems; however, in Prochlorococcus, they are substituted with M205 and C282, respectively. According to the solved photosystem II structure, these amino acids are not involved in Chl binding. To mimic Prochlorococcus, V205 was mutated to M205 in the D1 protein from Synechocystis sp. PCC6803 and Synechocystis dvr mutant was transformed with this construct. Although these transgenic cells could not grow under high-light conditions, they acquired light tolerance and grew under medium-light conditions, whereas untransformed dvr mutants could not survive. Substitution of G282 for C282 contributed additional light tolerance, suggesting that the amino acid substitutions in the D1 protein played an essential role in the development of DVChl-based photosystems. Here, we discuss the coevolution of a photosynthetic pigment and its binding protein.

Photosynthesis is one of the most important biological processes, contributing not only to biological activity but to maintaining the global environment (1). The apparatuses and properties of photosynthesis have been dynamically changed during evolution (2). The first photosynthetic organisms are thought to have been anaerobic photosynthetic bacteria, which did not produce oxygen in the process of photosynthesis (3). Cyanobacteria were the first oxygenic phototrophs whose photosystems evolved from those of preexisting photosynthetic bacteria (4, 5). The conversion of bacteriochlorophyll to chlorophyll (Chl) a was crucial, even in this evolutionary process, because it permitted the oxidation of H₂O (3). In addition to this event, the acquisition of new photosynthetic pigments has greatly contributed to the diversification of photosynthetic organisms (6). All photosynthetic pigments except chlorosomes exist as pigment–protein complexes; therefore, acquisition of the binding protein for the new pigment has been an indispensable process in the establishment of new photosynthetic pigment systems. Because photosynthetic organisms cannot acquire the pigment-binding protein in advance, a new pigment would first need to be incorporated into preexisting proteins, after which the binding proteins would evolve to fit the new pigment better (7). To understand the evolution of pigment systems, phylogenetic analyses of pigment-synthesis enzymes (8) and pigment-binding proteins (9) have been extensively performed. However, the coevolutionary process of a pigment and its binding protein cannot be elucidated by this method because of the lack of information about the intermediate states of the pigment–protein complexes, which appeared and then disappeared during evolution. In vivo experiments that mimic the evolution of the pigment–protein complexes are a potentially powerful tool that could be used to overcome this problem (7).

Prochlorococcus belongs to a marine picophytoplankton clade (10, 11) and is a major participant of the global carbon cycle (12), in part, because of its unique photosynthetic system, which uses divinyl chlorophyll (DVChl) instead of monovinyl chlorophyll (MVChl) (13). DVChl allows Prochlorococcus to photosynthesize and grow under the deep-sea water column, where blue light predominates because DVChl harvests blue light more efficiently than MVChl. Thus, acquisition of DVChl was an important evolutionary event for Prochlorococcus (14). Recently, we identified the 3,8-divinyl chlorophyllide 8-vinyl reductase (DVR) genes, which convert a vinyl group on the C-8 position to an ethyl group, in Arabidopsis (15) and cyanobacteria (16). Genome analysis showed that marine Synechococcus, a closely related species, contains a DVR gene, whereas Prochlorococcus does not, indicating that the progenitor of the Prochlorococcus genus lost the DVR gene and acquired DVChl. However, cyanobacterial and Arabidopsis dvr mutants, which accumulate DVChl instead of MVChl, immediately die under high-light conditions as a result of severe photodamage and can survive only under low-light conditions (16). In contrast, Prochlorococcus adapted to high-light conditions despite the presence of DVChl (17). One possible mechanism for this adaptation is the evolution of Chl-binding proteins to fit to DVChl within the Prochlorococcus lineage.

To elucidate the evolution of DVChl-based photosystems, we first compared the amino acid sequences of Chl-binding proteins of MVChl-based and DVChl-based photosystems. We found two amino acid residues that are conserved only in Prochlorococcus D1 proteins and could not find any Prochlorococcus-specific amino acid residues in other Chl-binding proteins. When these two amino acid residues of Synechocystis D1 proteins were substituted with the residues found in Prochlorococcus and introduced into a Synechocystis dvr mutant, the transgenic Synechocystis acquired light tolerance. In this paper, we discuss the coevolutionary process of a photosynthetic pigment and its binding protein.

Results

Photodamage of DVChl-Containing Organisms.

MVChl and DVChl have an ethyl group and a vinyl group at position C8, respectively (Fig. 1A). When DVR is mutated or lost during evolution, the final product of Chl biosynthesis is DVChl. The Arabidopsis and Synechocystis dvr mutants synthesize DVChl and construct DVChl-based photosystems. These mutants grow photosynthetically under low-light conditions. When Arabidopsis or cyanobacteria that were grown under low-light conditions were transferred to high-light conditions, WT survived but both dvr mutants lost Chl and died within 1 d (Fig. 1B). One possible mechanism for the rapid degradation of Chl is the induction of Chl degradation enzymes in dvr mutants. The other is the severe photodamage under high-light conditions in dvr mutants, although light harvesting systems differ between the two organisms (14).

Fig. 1.

DVChl-accumulating mutants in Arabidopsis and cyanobacterium. (A) Structure of DVChl and MVChl. The positions of the 8-vinyl group and 8-ethyl group are indicated by a circle on each chemical structure. (B) Photodamage of Arabidopsis and Synechocystis dvr mutants accumulating DVChl. Arabidopsis mutants were grown for 3 wk under standard light conditions (40 μmol⋅m⁻²⋅s⁻¹) and exposed to strong light (1,000 μmol⋅m⁻²⋅s⁻¹) for 24 h. Synechocystis grown under low-light conditions (30 μmol⋅m⁻²⋅s⁻¹) was suspended in culture medium (OD₇₅₀ = 1.0) and exposed to strong light (500 μmol⋅m⁻²⋅s⁻¹) for 24 h.

Comparison of Amino Acid Sequences of Core Complexes.

DVChl is structurally different from MVChl, as shown in Fig. 1, and this structural variation may prevent DVChl from fitting into MVChl binding sites, resulting in severe photodamage in dvr mutants. The question then arises as to which Chl–protein complexes caused photodamage in the dvr mutants. Arabidopsis and Synechocystis contain different peripheral antenna systems, but the core antenna and reaction center complex are well conserved in both organisms, suggesting that the core antenna and/or reaction center complexes cause the common photodamage observed in both of the dvr mutants. Because Prochlorococcus uses DVChl instead of MVChl, we sought to identify changes in the Chl-binding proteins specific to Prochlorococcus by aligning the amino acid sequences of the reaction center complexes of photosystem II (D1, D2), the core antenna complexes of photosystem II (CP43, CP47), and the P700 Chl a–protein complex (CP1). We could not find any Prochlorococcus-specific amino acid residues in D2, CP43, CP47, and CP1, suggesting that these core complexes can accept DVChl without modifications. In contrast, D1 proteins of all Prochlorococcus strains (18) contain two conserved amino acid residues (M205 and C282) that are generally substituted with V205 and G282 in other organisms, including marine Synechococcus strains (19) (Fig. S1 and Discussion).

Substitution of D1 Protein.

If the evolution of the D1 protein is a crucial process for acquiring DVChl-based photosystems, the substitution of Synechocystis D1 by Prochlorococcus D1 should increase the light tolerance of Synechocystis dvr mutants. To test this hypothesis, we introduced the Prochlorococcus psbA gene into a Synechocystis dvr mutant. Three psbA genes in the Synechocystis genome were disrupted, and modified psbA was introduced. These transgenic lines were used to investigate the function of the introduced Prochlorococcus psbA gene. To compare the transgenic lines with WT and the dvr mutants, the psbA1 and psbA3 genes in these two lines were disrupted. The MVChl-accumulating Synechocystis (MV mutant), DVChl-accumulating dvr mutant (DV mutant), and DV mutant harboring the Prochlorococcus D1 (DVPro mutant) all grew well under heterotrophic conditions. However, the DVPro mutant could not grow phototrophically (Fig. 2A), indicating that the Prochlorococcus D1 protein does not fit into the Synechocystis photosystem II. Next, instead of switching the whole D1 protein, only V205, G282, or both residues of the Synechocystis D1 protein (PsbA2) were mutated to Met and Cys, respectively, and this protein was introduced into the Synechocystis DV mutant (Fig. S2). Substitution of one or two amino acid residues might not cause drastic structural changes to the D1 protein. Transgenic lines carrying the modified D1 protein grew phototrophically (Fig. 2B). A typical phenomenon of photodamage in the dvr mutant is the rapid degradation of Chl under high-light conditions, as shown in Fig. 1B. To examine whether the amino acid substitutions protect against photodamage, cells grown under low-light conditions were exposed to high light (750 μmol⋅m⁻²⋅s⁻¹) for 6 h and the Chl contents were determined. Chl did not decrease in the MV mutant after high-light treatment, but ∼60% of the Chl degraded in the DV mutant. Interestingly, the V205M/G282C and G282C mutants retained 80% of the initial level of Chl, whereas 40% of the Chl degraded in the V205M mutant during high-light treatment.

Fig. 2.

Effect of the substitution of D1 protein on photodamage. (A) Phototrophic and heterotrophic growth of transformants. MVChl-accumulating cell (MV mutant), DVChl-accumulating cell (DV mutant), and DVChl-accumulating cell whose D1 protein is substituted with the Prochlorococcus D1 protein (DVPro mutant) were cultured on agar. The psbA1 and psbA3 of all Synechocystis cells used in this study were disrupted. Cells were grown phototrophically under continuous illumination (30 μmol⋅m⁻²⋅s⁻¹) or heterotrophically with 5 mM glucose under dim light (5 μmol⋅m⁻²⋅s⁻¹) on agar plates for 1 wk. (B) Chl degradation under high-light conditions. Cells grown photoautotrophically under low-light conditions (30 μmol⋅m⁻²⋅s⁻¹) were suspended in the BG11 medium (OD₇₅₀ = 0.5) and incubated at 30 °C for 6 h under high light (750 μmol⋅m⁻²⋅s⁻¹). Before (white bar) and after (black bar) high-light treatment, Chl was extracted from the cells and measured by the absorbance at 663 nm. The change in Chl content is expressed as the ratio of the absorbance of Chl extracted after high-light treatment to the absorbance of Chl extracted before treatment. Error bars represent the SDs based on the mean values of three samples. *Values differ significantly from the DV mutant after high-light treatments (P < 0.01 by t test). **Values differ significantly from the G282C mutant after high-light treatments (P < 0.05 by t test). MV, MV mutant; DV, DV mutant; V205M/G282C, V205M/G282C mutant; V205M, V205M mutant; G282C, G282C mutant.

Growth Rates and Photosynthetic Capacity of Transformed Cells.

In many cases, substitution of amino acid residues in the D1 protein impairs photoautotrophic growth as a result of the loss of photosynthetic capacity (20), because many amino acid residues participate in water oxidation, charge separation, and other important processes of photosystem II. To examine the impact of the amino acid substitution of the D1 protein, the growth of the transformants was assessed by optical density at 750 nm, cell numbers, and Chl levels under low-light (30 μmol⋅m⁻²⋅s⁻¹) or medium-light (250 μmol⋅m⁻²⋅s⁻¹) conditions (Fig. 3A and Fig. S3). All the cell lines grew well under low-light conditions. Under medium-light conditions, the MV, V205M/G282C, and V205M mutants could grow but the DV and G282C mutants could not proliferate (Fig. 3A). Growth impairment of the G282C mutant is inconsistent with the observation that Chl was more stable in the G282C mutant than in the V205M mutant on strong illumination (Fig. 2B). It is possible that growth impairment under medium light (Fig. 3A) might be caused by a distinct mechanism different from that which influences the stability of Chl under strong illumination (Fig. 2B). Although the optical density of the MV mutant was slightly higher than that of other lines after 5 d of medium light, the doubling times of the MV, V205M/G282C, and V205M mutants calculated from Fig. 3A were 1.18, 1.27, and 1.29 d, respectively. Optical densities and cell numbers exhibited almost the same profiles (Fig. 3A and Fig. S3A). Analysis of the growth of the transformants under low- and medium-light conditions indicates that the substitution of the two amino acid residues played an essential role in the development of a DVChl-based photosystem. To examine the effect of the amino acid substitution on the growth of Synechocystis, which uses MVChl-based photosystem, V205 and G208 were substituted for M205 and C282, respectively, in MV mutant (MV-V205M/G282C mutant). MV and MV-V205M/G282C mutants exhibited the same growth profile (Fig. S4), indicating that substitution of these two amino acid residues has no impact on MVChl-based photosystems.

Fig. 3.

Cell growth under different light conditions. (A) Cell growth under low- or medium-light conditions. Cells growing under low light (30 μmol⋅m⁻²⋅s⁻¹) were diluted with culture medium to OD₇₅₀ = 0.05 and grown under low-light (30 μmol⋅m⁻²⋅s⁻¹) or medium-light (250 μmol⋅m⁻²⋅s⁻¹) conditions, rotating at 30 °C. Cell biomass was monitored by measuring optical density at 750 nm. (B) Effect of light intensity on cell growth. Diluted cells grown under low-light conditions (OD₇₅₀ = 0.05) were incubated under 30, 100, 200, 500, and 1,000 μmol⋅m⁻²⋅s⁻¹. After incubation for 3 d, OD₇₅₀, which is used as a proxy for biomass, was measured. Error bars represent the SDs based on the mean values of three samples. MV, MV mutant; DV, DV mutant; V205M/G282C, V205M/G282C mutant; V205M, V205M mutant; G282C, G282C mutant; MV-V205M/G282C, MV-V205M/G282C mutant.

Next, we measured the increase in biomass of the MV, DV, and V205M/G282C mutants during incubation under various light intensities to examine to what extent these amino acid substitutions contribute to light tolerance (Fig. 3B). The DV mutant was unable to grow at 100 μmol⋅m⁻²⋅s⁻¹. The V205M/G282C mutant grew at 200 μmol⋅m⁻²⋅s⁻¹, but it died at 500 μmol⋅m⁻²⋅s⁻¹. In contrast, the MV mutant survived even at 1,000 μmol⋅m⁻²⋅s⁻¹. Although these amino acid substitutions significantly enhanced the light tolerance of DVChl-based photosystems, these photosystems are still less tolerant to high-light intensities than MVChl-based photosystems. To examine whether amino acid substitutions contribute to further light tolerance of the MVChl-based photosystems, MV-V205M/G282C mutant was grown under high-light conditions. The optical density of MV-V205M/G282C mutant decreased at 1,000 μmol⋅m⁻²⋅s⁻¹ compared with 500 μmol⋅m⁻²⋅s⁻¹ of MV mutant (Fig. 3B), indicating that these amino acid substitutions have no effect on light tolerance of MVChl-based photosystems.

The O₂ evolution of cells grown under low-light conditions was determined under saturating light intensity (Fig. 4A). The O₂ evolution rate was low in the V205M mutant but high in the G282C mutant. After 5 d of medium light, the rates of O₂ evolution of the MV, V205M/G282C, and V205M mutants were 290, 168, and 115 μmol⋅mg Chl⁻¹⋅h⁻¹, respectively, but those of the DV and G282C mutants were under detectable levels (Fig. 4B). These results of O₂ evolution are consistent with the optical density after incubation for 5 d (Fig. 3A).

Fig. 4.

Photosynthetic oxygen-evolving activity of transformants. (A) Oxygen evolution rate of cells grown under low-light conditions. Cells growing under low-light conditions (30 μmol⋅m⁻²⋅s⁻¹) were suspended in BG11 medium to 10 μg Chl⋅mL⁻¹, and oxygen evolution was measured. (B) Oxygen evolution rate of cells grown under medium-light conditions. Cells were grown under medium-light conditions (250 μmol⋅m⁻²⋅s⁻¹) for 5 d as in Fig. 3A, and oxygen evolution rate was measured. N. D., not detected. (C) Effects of light intensity on photoinhibition. Cells growing under low-light conditions (30 μmol⋅m⁻²⋅s⁻¹) were suspended in culture medium at a concentration of OD₇₅₀ = 1.0 and incubated under 30, 100, 250, and 500 μmol⋅m⁻²⋅s⁻¹ for 3 h at 30 °C in the presence of lincomycin (300 μg⋅mL⁻¹). Before and after light treatment, oxygen evolution was measured. Error bars represent the SDs based on the mean values of three samples. Photosynthetic oxygen-evolving activity of intact cells was measured using a Clark-type oxygen electrode at 1,000 μmol⋅m⁻²⋅s⁻¹ in the presence of 5 mM NaHCO₃ at 30 °C. MV, MV mutant; DV, DV mutant; V205M/G282C, V205M/G282C mutant.

Next, we examined whether amino acid substitutions reduce the photoinhibition in the transformants. Cells were exposed to various light intensities in the presence of lincomycin, which inhibits the repair process of photosystem II (Fig. 4C). The rate of O₂ evolution decreased by about 60% and 30% in DV and V205M/G282C mutants, respectively, after 3 h of light treatment (100 μmol⋅m⁻²⋅s⁻¹), indicating that the amino acid substitutions decreased the extent of photoinhibition. The amino acid substitutions did not have much of an impact at higher light intensities of 250 and 500 μmol⋅m⁻²⋅s⁻¹. These results suggest that higher tolerance of the amino acid-substituted lines against medium light can be at least partly explained by decreased rates of photoinhibition.

Taken together, all strains grow at low light and evolve oxygen; however, at medium light, neither the DV mutant nor the G282C mutant grows or evolves oxygen (Figs. 3 and 4). The V205M and G282C substitutions in the D1 protein of the DV mutant both suppressed Chl degradation under high-light conditions (Fig. 2B).

Discussion

Evolution of a DVChl-Based Photosystem.

Although molecular phylogenetic analysis is a powerful tool for understanding the evolution of photosynthesis and photosynthetic organisms, the biochemical processes of the evolution cannot be uncovered by this method (7). However, we can deduce the evolutionary intermediates of photosynthetic machineries based on the comparison of genome sequences. Additionally, we can produce the organism that will use these photosynthetic machineries (21) and examine the physiological responses of this organism. Here, we used this method for understanding the evolution of the DVChl-based photosystem of Prochlorococcus.

Phylogenetic analysis of 16S rRNA has shown that Prochlorococcus is phylogenetically closely related to marine Synechococcus (10). Genomic analyses of Synechococcus and Prochlorococcus revealed that the progenitor of the genus Prochlorococcus lost the DVR gene and acquired DVChl (15). This progenitor had an advantage in that it could efficiently absorb blue light, which is enriched in deep-water layers. However, this progenitor also had a disadvantage because it could survive only under low-light conditions, as suggested by the dvr mutant. We hypothesize that the first step in the acquisition of high-light tolerance was the substitution of the amino acid residues in the D1 protein. Support for this idea is based on the finding that M205 and C282 of D1 are unique amino acid residues that are conserved only in Prochlorococcus D1 proteins. We could not find Prochlorococcus-specific sequences/residues in other Chl-binding proteins. This strongly suggests that these two amino acids were mutated to the Prochlorococcus type (M205 and C282) in a common progenitor of the genus Prochlorococcus. This progenitor then acquired significant tolerance to high-light conditions, despite the accumulation of DVChl, which is normally toxic to cells under high-light conditions. Another hypothesis is that the substitutions of these amino acids were followed by the conversion of MVChl to DVChl in the ancestor of Prochlorococcus. This hypothesis is consistent with the finding that the strain with the altered D1 functioned well in MVChl-based photosystems (Fig. 3B and Fig. S4). Either way, these changes to the D1 protein appear to be a crucial step in the establishment of DVChl-based photosystems.

Between the two amino acid residues, the substitution of V205 to M205 might be more important, because the V205M mutant could grow under medium-light conditions and evolved oxygen. The substitution of G282 for C282 results in less Chl degradation, which indicates that the substitution of G282 for C282 contributed to additional light tolerance (Fig. 2B). Other Chl-binding proteins might tolerate DVChl because photoinhibition occurs mainly in the reaction center of photosystem II and because D1 proteins are targets of photodamage. In addition to these two residues, various other amino acid substitutions in the D1 protein occurred independently in each of the Prochlorococcus lineages. Some of these substitutions might contribute to a better fit with DVChl. The Prochlorococcus genus contains both species adapted to high-light conditions and species adapted to low light (17, 22, 23). Prochlorococcus adapted to low-light conditions lives under low-light intensities that the dvr mutant can tolerate, suggesting that this Prochlorococcus can survive in nature without amino acid substitutions. However, it is known that photodamage is enhanced by other physiological stresses (24). Furthermore, Prochlorococcus adapted to low-light conditions has a large antenna size. Acquisition of stress tolerance might be beneficial even for Prochlorococcus adapted to low light. The hypothesis cannot be excluded that Prochlorococcus adapted to high-light conditions acquired more light tolerance through additional changes in the amino acid sequence of the D1 protein. Prochlorococcus ecotypes might have adapted their D1 protein to specific cellular and environmental conditions to protect against photodamage. In addition to the evolution of D1 protein, a decrease in photosynthetic antenna size might be an important process for the acquisition of light tolerance during the evolution of Prochlorococcus adapted to high-light conditions (25).

Most of the MVChl-based photosynthetic organisms contain the conserved V205 and G282 residues. However, Acaryochloris marina, which uses Chl d as a photosynthetic pigment (26), retains C282 instead of G282, although V205 is conserved (Fig. S1). One possible reason for this amino acid substitution in Acaryochloris is to protect against photodamage by accommodating Chl d, which is structurally different from Chl a (27, 28).

Mechanism of Two Amino Acid Residues for Protection Against Photodamage.

Photodamage in DV mutants is not caused by defects in energy dissipation in the antenna system but by the reaction center of photosystem II (29), because the substitution of amino acid residues in the D1 protein led to a partial recovery from photodamage of the DV mutant. This is consistent with the report that D1 protein is more rapidly degraded by high-light treatment in dvr mutants compared with WT. Antenna systems potentially have the flexibility to accept structurally different pigments; for example, Synechocystis can use Chl b as a photosynthetic pigment (21, 30) and core antenna complexes of both photosystems can bind Chl b (31), although these proteins form Chl a–protein complexes in nature. In contrast, the reaction center complex might have many structural restrictions because the complex is responsible for many functions, such as charge separation, energy and electron transfer, and water oxidation. The atomic structure of photosystem II of cyanobacteria shows that V205 is located near the reaction center (32), which enables V205 to interact with a “special pair” (33) (Fig. S5). It is predicted that V205 is not involved in binding the special-pair Chl and cannot interact with pheophytin (32). V205 also does not directly interact with a vinyl group on pyrrole ring B but with another moiety of special-pair Chl. When V205 is replaced by Met, interactions of the residue with special-pair Chl and pheophytin become stronger because of a larger side chain of Met. If the replacement of this amino acid residue alters the potential of the special pair and/or pheophytin of photosystem II as a result of the strong interaction, the photodamage would be reduced (34). Although the photoinhibition rate was reduced in the V205M/G282C mutant, the possibility cannot be excluded that other processes, such as the repair cycle and other unknown processes, contribute to light tolerance of the V205M/G282C mutant. Although the mechanisms underlying the induction of and protection from photodamage have not yet been elucidated, the study of the transformed cells from this report will contribute to our understanding of these mechanisms of reaction center photoinhibition in cyanobacteria. It should be noted that these two amino acid substitutions contribute to survival under medium-light condition but not under high-light conditions. For Prochlorococcus to become adapted to high-light conditions, further processes would be required, for example, evolution of a high-light-inducible gene family (35, 36).

Distribution of psbA Genes in Cyanophages.

Photosynthesis genes exist not only in photosynthetic organisms but in cyanophages (37–39). Some podoviruses and myoviruses infecting Prochlorococcus and Synechococcus contain psbA genes. It was reported that these cyanophages acquired psbA genes from Prochlorococcus and Synechococcus during evolution (40). With the two exceptions of P-SSM1 and S-ShM1 (40), Synechococcus myoviruses and Synechococcus podoviruses isolated from Synechococcus encode MVChl-type D1 proteins, whereas Prochlorococcus myoviruses contain DVChl-type D1 proteins. This indicates that cyanophages infecting Prochlorococcus and Synechococcus contain DVChl- and MVChl-type D1 proteins, respectively. This finding is consistent with the idea that cyanophage psbA genes contribute to host photosynthesis.

However, this is not a strict rule with respect to cyanophage genes, as observed in Prochlorococcus podovirus, which contains the psbA gene and infects Prochlorococcus (41). Although the psbA gene of Prochlorococcus podovirus is phylogenetically close to the Prochlorococcus psbA genes, podovirus D1 proteins contain the MVChl-type residues V205 and G282. It was also reported that some phages can infect both Prochlorococcus and Synechococcus (41). These observations contradict the idea that the phage psbA gene is expressed during infection and contributes to photosynthesis within the host cells (42, 43), because MVChl-type D1 proteins cause photodamage in the DVChl-containing photosystem. Furthermore, Prochlorococcus podoviruses predominantly infect Prochlorococcus adapted to high-light conditions. One possibility is that the psbA gene does not contribute to photosynthetic capacity but has another unknown function. Another possibility is that D1 protein of these phages evolved to fit to Prochlorococcus and/or Synechococcus by altering amino acid sequence other than the two conserved amino acids (205 and 282). It would also be possible that another protein from phage enables D1 to function in cyanobacteria. The true function of the phage psbA gene still needs to be examined.

Phylogenetic analysis of the psbA genes of cyanobacteria and cyanophages indicated that Prochlorococcus podovirus acquired the psbA gene from a common ancestor of the Prochlorococcus genus (40) (Fig. S6). There are two hypotheses concerning the Chl type of this ancestral Prochlorococcus from which podovirus acquired the psbA gene. The first hypothesis is that the ancestor had already acquired DVChl, and the second is that this ancestor still used MVChl. As shown in this report, MVChl-based photosynthetic organisms are severely photodamaged if they have DVChl-type D1 proteins. If the Prochlorococcus progenitor changed the D1 protein to the DVChl type immediately after acquiring DVChl to protect against photodamage, Prochlorococcus podovirus should have acquired the psbA gene from ancestral Prochlorococcus that still used MVChl. The study of psbA genes in cyanophages will contribute to our understanding of the evolution of Prochlorococcus.

Materials and Methods

Materials and Growth Conditions.

In this report, low, medium, and high light indicate 30, 100–250, and 500–1,000 μmol⋅m⁻²⋅s⁻¹, respectively. Arabidopsis thaliana (Columbia ecotype) was grown at 23 °C under continuous light in a chamber equipped with white fluorescent lamps at a light intensity of 40 μmol⋅m⁻²⋅s⁻¹. For high-light treatment, 3-wk-old Arabidopsis plants were transferred to high light (1,000 μmol⋅m⁻²⋅s⁻¹) for 1 d.

Synechocystis sp. PCC6803 cells were grown at 30 °C in liquid BG11 medium with ambient CO₂ under continuous illumination (30 μmol⋅m⁻²⋅s⁻¹) rotating at 120 rpm, and the logarithmically growing cells were used for further experiments. For the experiments of light-induced Chl degradation, the OD₇₅₀ of the liquid cell cultures was set to 1.0 (2-mL cell culture in a 24-well plate) for Fig. 1B and to 0.5 for Fig. 2B, and the cells were exposed to 500 or 750 μmol⋅m⁻²⋅s⁻¹ at 30 °C. For the growth experiments (Figs. 3 and 4A), the OD₇₅₀ of the liquid cell cultures was set to 0.05 (100-mL cell culture in a 200-mL flask) and the cells were grown under various light intensities at 30 °C.

Constructs and Transformations.

All Synechocystis transformants analyzed in this study were prepared from Synechocystis engineered to express a 6× histidine tag at the C terminus of CP47 selected by coexpressing the kanamycin resistance gene. Furthermore, all transformants were prepared by the disruption of psbA1 and psbA3 by replacing part of the coding region with a cassette encoding an erythromycin resistance gene and a gentamicin resistance gene via homologous recombination. For generating the V205M and G282C site-directed mutants, GAT (codons 613–615) and GGC (codons 844–846) were changed to ATG and TGC, respectively, and the transformants harboring mutated psbA2 were selected by coexpressing the streptomycin resistance gene (44, 45). Segregation of the mutant genomes was confirmed by PCR assay (Fig. S2). Furthermore, the sequence analysis of the cDNA that was reverse-transcribed from psbA mRNA confirmed that only modified psbA gene was expressed in these transformants (Fig. S2). The slr1923 locus encoding DVR was disrupted by the chloramphenicol-resistant cassette as reported previously (16).

Flow Cytometry Analysis.

Cell numbers of transformants were monitored by flow cytometry using a FACS Canto flow cytometer with a 488-nm laser (BD Biosciences). Culturing cells were directly loaded, with the acquisition performance at a low rate (10 μL⋅min⁻¹). Bleached cells generated under medium-light conditions were taken into account to estimate the number of cells.

Chl Measurement.

One milliliter of culture was centrifuged at 22,000 × g for 5 min to precipitate cells. Cells were resuspended in methanol to extract Chl. After centrifugation at 22,000 × g for 5 min, the absorbance at 665 nm of the supernatant was measured and Chl content was calculated following the equations of Porra (46).

Measurement of Oxygen Evolution.

Photosynthetic oxygen-evolving activity of intact cells was measured using a Clark-type oxygen electrode (Hansatech Instruments). Cell suspension (1 mL) was put into the measurement chamber of the oxygen electrode, and oxygen evolution rates were measured under saturating light (1,000 μmol⋅m⁻²⋅s⁻¹) for 2 min at 30 °C in the presence of 5 mM NaHCO₃. For the experiment in Fig. 4A, cells exponentially growing under low-light conditions (30 μmol⋅m⁻²⋅s⁻¹) were suspended in BG11 medium to 10 μg Chl⋅mL⁻¹ and oxygen evolution rates were measured. To evaluate the photosynthetic activity of the cells growing under medium-light conditions, cells were cultured for 5 d at 250 μmol⋅m⁻²⋅s⁻¹ as in Fig. 3A and oxygen-evolving activity was measured (Fig. 4B). For the photoinhibition experiments (Fig. 4C), cells growing under low light were exposed to various light intensities and the oxygen-evolving activity was measured.

Photoinhibition Treatment.

Synechocystis cells grown under low-light conditions were suspended in 2 mL of BG11 medium (OD₇₅₀ = 1.0) and transferred into a 24-well plate. To measure the photoinhibition under various light intensities, cells were exposed to 30, 100, 250, or 500 μmol⋅m⁻²⋅s⁻¹ for 3 h in the presence of 300 μg⋅mL⁻¹ lincomycin, which inhibits repair processes in the photosystem II reaction center.