chlorophyll(ide) - Plant Physiology

slr1923 of Synechocystis sp. PCC6803 Is Essential for Conversion of 3,8-Divinyl(proto)chlorophyll(ide) to 3-Monovinyl(proto)chlorophyll(ide)1 M. Rafiqul Islam, Shimpei Aikawa, Takafumi Midorikawa, Yasuhiro Kashino, Kazuhiko Satoh, and Hiroyuki Koike2* Department of Life Science, Graduate School of Life Science, University of Hyogo, Ako, Hyogo 678–1297, Japan (M.R.I., S.A., Y.K., K.S., H.K.); and Department of Life Sciences (Biology), University of Tokyo, Meguro, Tokyo 153–8902, Japan (T.M.)

The deduced amino acid sequence of an slr1923 gene of Synechocystis sp. PCC6803 is homologous to archaean F420H2 dehydrogenase, which acts as a soluble subcomplex of reduced nicotinamide adenine dinucleotide dehydrogenase complex I. In this study, the gene was inactivated and characteristics of the mutant were analyzed. The mutant grew slower than the wild type under 100 mE m22 s21 but did not grow under high light intensity (300 mE m22 s21). The cellular content of chlorophyll was lower in the mutant, and the absorption spectrum showed a shift in the absorption peak of the Soret band to a longer wavelength by about 10 nm compared with the wild type. It was found, by high-performance liquid chromatography analysis, that the retention time of chlorophyll of the mutant is shorter than that of the wild type and that the peak wavelength of the Soret band was also shifted to a longer wavelength by 11 nm. Proton nuclear magnetic resonance analysis of the chlorophyll of the mutant revealed that the ethyl group of position 8 of ring B is replaced with a vinyl group. The spectrum indicates that the chlorophyll of the mutant is not a normal (3-vinyl)chlorophyll a but a 3,8-divinylchlorophyll a. These results strongly suggest that the Slr1923 protein is essential for the conversion from divinylchlorophyll(ide) to normal chlorophyll(ide). We thus designate this gene cvrA (a gene indispensable for cyanobacterial vinyl reductase).

Land plants, algae, cyanobacteria, and photosynthetic bacteria use various types of chlorophyll molecules (chlorophyll [chl] a, b, c, and d and bacteriochlorophyll a, b, c, etc.) as inevitable photon-capturing pigments in photosynthesis. Extensive studies with various organisms by biochemical and genetic approaches have disclosed many aspects related to chlorophyll metabolism (for review, see Beale, 1993, 1999; Senge and Smith, 1995). However, the functional or regulatory genes related to various chlorophyll biosynthetic pathways have not yet been fully elucidated. In land plants, chlorophyll precursors are sometimes accumulated as both 3-monovinyl (MV-) or 3,8-divinyl (DV-) intermediates, and the ratio between the two forms changes depending on the species, tissues, and growth conditions (Kim and Rebeiz, 1996). Chlorophyll biosynthetic heterogeneity is assumed to originate 1 This work was supported by a grant from the 21st Century Center of Excellence Program to K.S. 2 Present address: Department of Biological Sciences, Faculty of Science and Engineering, Chuo University, Kasuga 1-13-27, Bunkyo, Tokyo 112–8551, Japan. * Corresponding author; e-mail [email protected]. The author responsible for distribution of materials integral to the findings presented in this article in accordance with the policy described in the Instructions for Authors (www.plantphysiol.org) is: Hiroyuki Koike ([email protected]). www.plantphysiol.org/cgi/doi/10.1104/pp.108.123117

1068

mainly in parallel DV- and MV-chl biosynthetic routes interconnected by 8-vinyl reductases (8 VRs) that convert DV-tetrapyrroles to MV-tetrapyrroles by conversion of the vinyl group at position 8 of ring B to the ethyl group. So far, five 8-VR activities have been detected at the levels of DV-Mg-protophorphyrin IX, Mg-protomonomethyl ester, protochlorophyllide (Pchlide) a, chlorophyllide (chlide) a, and chl a (Kolossov et al., 2006). It has not been clear whether the various 8-VR activities are catalyzed by a single enzyme with broad specificities encoded by a single gene or by a family of enzymes with narrow specificity encoded by multiple genes, as is the case of NADPH Pchlide oxidoreductases (Rebeiz et al., 2003). Mutants that accumulate DV-chl(ide) a can be a potential model system. A mutant that accumulates DV-chl was reported by Bazzaz and Govindjee (1974) and analyzed by Bazzaz (1981) in Zea mays. Nagata et al. (2005) and Nakanishi et al. (2005) independently identified gene AT5G18660 of Arabidopsis as a divinyl reductase (DVR) that has sequence similarity to isoflavone reductase. By investigation of the mutant created by insertional inactivation, Chew and Bryant (2007) demonstrated that CT1063 (bciA), which is an ortholog of the Arabidopsis gene, encodes a C-8 DVR of Chlorobium tepidum TLS. They also concluded that BchJ, which had been reported to be a vinyl reductase (Suzuki and Bauer, 1995), is not the enzyme. They assumed that it may play an important role in sub-

Plant Physiology, October 2008, Vol. 148, pp. 1068–1081, www.plantphysiol.org Ó 2008 American Society of Plant Biologists

slr1923 Is Essential for Divinylchlorophyll(ide) Reduction

strate channeling and/or the regulation of bacteriochlorophyll biosynthesis. An ortholog of the DVR gene, however, is not found in the complete genome sequences of cyanobacteria, Synechocystis sp. PCC6803 (Synechocystis 6803; Kaneko et al., 1996), Anabaena sp. PCC7120 (Kaneko et al., 2001), Thermosynechococcus elongatus BP-1 (Nakamura et al., 2002), and Gloeobacter violaceus PCC7421 (Nakamura et al., 2003), or the unicellular red alga Cyanidiochyzan merolae 10D (Matsuzaki et al., 2004). Based on the fact that these organisms synthesize MV-chl(ide), it is reasonable to assume that there exists an 8-vinyl reductase gene(s) whose amino acid sequence is rather different from those found in higher plants or C. tepidum. Thus, the identification of other cyanobacterial DVRs is an important subject for elucidation of the chlorophyll biosynthetic pathway. The model photosynthetic organism Synechocystis 6803 has a unique gene, slr1923, which encodes a hypothetical protein with unknown function. The deduced amino acid sequence of the gene has a similarity with FpoF, which functions as a F420H2 dehydrogenase subunit F. The protein acts as an electron input device to incomplete NADH dehydrogenase complex 1 (NDH1) of the methanogenic archaean species Methenosarcina mazei (Prommeenate et al., 2004). It is an open question whether it still oxidizes coenzyme F420H2 or evolved differently to use NAD(P)H in Synechocystis 6803. Elucidation of a possible function and origin of the protein is of great interest not only with respect to its electron transfer mechanism but also with respect to the evolution and distribution of proteins of archaean origin. In this study, we created a mutant of slr1923 of Synechocystis 6803 by insertional inactivation, and the characteristics of the mutant were analyzed. It was revealed that all of the molecular species of chl a of the

mutant were DV-chl a but not MV-chl a (normal chl a). The molecular identity was confirmed by absorption spectra, HPLC, and 1H-NMR analysis. These results show that slr1923 is inevitable for the conversion of 3,8-DV-(proto)chl(ide) a to 3-MV-(proto)chl(ide) a in Synechocystis 6803. We thus designate the gene cvrA (a gene indispensable for cyanobacterial vinyl reductase). The distribution and evolution of vinyl reductases are discussed. Recently, Ito et al. (2008) considered slr1923 as a candidate of cyanobacterial DVR from bioinformatics analyses. They also knocked out the gene and characterized the mutant. The basic phenotypes were principally the same, but here we present some more detailed properties and discuss the effects of the mutation. RESULTS Construction of the Mutant

In order to elucidate the role of the protein Slr1923, we have constructed an slr1923-inactivated mutant of Synechocystis 6803. Wild-type cells were transformed by the plasmid pGEM-Nf-SpR-Cf, in which a spectinomycin resistance cassette with a reverse orientation with respect to the direction of slr1923 transcription was inserted (Fig. 1A). After transformation by homologous recombination, colonies grown on a spectinomycincontaining BG11 agar medium were picked up and successive transplantations were performed to segregate the mutant lines. Segregation was confirmed by PCR amplification of the corresponding DNA region by genomic DNA as a template. Figure 1B shows that the amplified band of the mutant was totally replaced by the mutated DNA fragment, indicating total segregation in the mutant.

Figure 1. Construction of the slr1923 inactivation mutant, segregation of the slr1923M strain, and northern-blot analysis. A, Schematic representation of the slr1923-containing region and construction of the mutant. B, PCR analysis of the segregation of the slr1923M strain. M1, 100-bp DNA ladder (New England Biolabs); M2, 1-kb DNA ladder (New England Biolabs); W/WT, wild type; M/MT, slr1923M; Specr, spectinomycin resistance cassette (shaded arrow). C, Northern analysis of transcripts of slr1923, slr1924, and slr1925. RNA was isolated as described in “Materials and Methods,” and total RNA (20 mg) was subjected to analysis. The bottom panels show rRNA stained with methylene blue as a loading control. M, Mutant.

Plant Physiol. Vol. 148, 2008

1069

Islam et al.

Based on the genomic sequence of Synechocystis 6803 (Kaneko et al., 1996), two other genes are located downstream of slr1923 (Fig. 1A). slr1924 is only 25 bp apart from slr1923, and slr1924 and slr1925 overlap by 1 bp. It is possible that these three genes are cotranscribed as a single mRNA. Inactivation of slr1923, the most upstream located among the three genes (Fig. 1A), may affect expression of the downstream genes. Figure 1C shows northern analysis of transcripts of the three genes. The signal intensity of the mutant was about half or lower compared with the wild type. However, the signal intensities of the transcripts were not much different between the three genes. This suggests that expression of slr1924 and slr1925 is not inhibited by disruption of slr1923. Although higher molecular mass signals that correspond to the length equivalent of two genes were observed (Fig. 1C, arrow), further analysis is necessary to determine whether the three open reading frames form operons or not. Ito et al. (2008) addressed this problem by inactivating slr1924 and slr1925. They found no difference in phenotypes in the two mutants. This observation also supports our results that the phenotype of the mutant observed is not due to lowered or null expression of slr1924 and slr1925 but to inactivation of the slr1923 gene. Growth Patterns under Different Conditions

When wild-type and mutant cells were grown under 100 mE m22 s21 and aerated by 3% CO2-containing air (normal condition hereafter), they showed similar growth patterns except that the slr1923 inactivation mutant (slr1923M) grew at a slower rate (Fig. 2A). The cells grew exponentially up to about 36 h of cultivation, with doubling times of 10.6 and 12.1 h for the wild type and the mutant, respectively, followed by a gradual decrease in the growth rates, finally reaching a stationary phase after about 80 h. The growth rates of the wild type under 30, 100, and 300 mE m22 s21 were very similar to those reported previously (Ohtsuka et al., 2004). Under moderate light intensities, the mutant cells grew significantly, although the rates were slower compared with those of the wild type (Table I). When slr1923M was grown under strong light (300 mE m22 s21), there was almost no considerable growth from the beginning and the cells started to die at about 34 h of inoculation (Fig. 2B), indicating a high sensitivity to strong illumination. High sensitivity to strong illumination is reported in an Arabidopsis (Arabidopsis thaliana) mutant (Nagata et al., 2005; Nakanishi et al., 2005) and also in a Synechocystis 6803 mutant (Ito et al., 2008). Although the SD is large, slr1923M showed a tendency to grow slower than the wild type under low CO2 (0.03%). The growth rates under high-salt (0.5 M NaCl) plus lowCO2 conditions were not much different between the wild type and the mutant. The doubling times of wildtype and mutant cells grown under various conditions are summarized in Table I. 1070

Figure 2. Effects of light intensity on the growth of wild-type cells (white circles) and slr1923M cells (black circles) under 100 (A) or 300 (B) mE m22 s21.

Spectral Characteristics

An absorption spectrum of wild-type cells showed absorption maxima at 678 and 630 nm in the red region by absorption of chl a and phycobilisomes and at 435 nm in the Soret region (Fig. 3A, solid line). However, the absorption spectrum of slr1923M showed a rather low absorbance in the chlorophyll red region, while absorption by phycobilisome was practically unchanged at any stage of growth on the basis of cell number (Fig. 3A, broken line). The peak wavelengths of the red band region due to chl a and phycobilisomes were the same as those of wild-type cells. It was found that the absorption peak of the Soret band shifted to a longer wavelength by about 10 nm. The absorption spectrum of the same mutant reported by Ito et al. (2008) showed a shift by 6 nm in the Soret band. Judged from the absorption spectrum, the peak of the mutant seems to be affected by the higher amount of carotenoids. This may account for the difference of peak positions between our data and those of Ito et al. (2008). Plant Physiol. Vol. 148, 2008


Table I. Doubling times (hours) of wild-type and slr1923M cells grown under various conditions 3

CO2 Concentration (%): Light Intensity (mE m22 s21):

30

100

300

100

100

100

NaCl (M):

–

–

–

0.5

–

0.5

24.8 6 1.1 34.3 6 2.6

10.0 6 0.4 11.9 6 0.6

8.8 6 0.2 2a

12.9 6 2.5 16.7 6 3.3

13.3 6 3.6 16.4 6 1.8

24.9 6 0.4 26.6 6 1.1

Wild type slr1923M a

0.03

No doubling time, as it died very soon after cultivation.

Figure 3B shows changes of absorption spectra of the mutant cells grown under the high-light condition (300 mE m22 s21). The cells show low chlorophyll absorbance relative to phycobilisome at the initial stage (9.5 h) of cultivation (Fig. 3B, line a). After 17 h, the mutant still retained the spectral characteristics of the initial inoculum (Fig. 3B, line b). However, the color of the culture turned to bluish after 24 h. The spectrum showed a decreased amount of chlorophyll, while phycobilisome content did not change so much based on the number of the cells (Fig. 3B, line c). Finally, the cells lost almost all of the chlorophyll and carotenoids, but a significant amount of phycobilisomes remained. This resulted in total blue color after 33.5 h of growth (Fig. 3B, line d). This is rather different from the responses to high-light illumination of wildtype cells. The wild-type cells show increased carotenoids and decreased phycobilisomes, giving rise to yellowish-green color when grown under strong illumination. Decreased absorption of the chlorophyll red band relative to phycobilisome could result in higher fluorescence from phycobilisomes and reduced energy transfer efficiency from phycobilisomes to PSII core complexes. This possibility was addressed by measuring fluorescence emission spectra at 77 K. When wildtype cells were excited by a broad-band blue light, three emission bands peaking at 650 to 665, 685 to 695, and 725 nm emitted from phycobiliproteins, phycobilisome anchor (LCM)/PSII, and PSI, respectively, were observed (Fig. 4A, solid line). Fluorescence intensity at 725 nm was about two times higher than that at 695 nm. On the other hand, a very high intensity of fluorescence at around 685 nm relative to that at 725 nm was observed in slr1923M. The intensity of fluorescence at 685 to 695 nm was higher than that emitting from PSI (725 nm), reflecting a higher content of phycobilisome in the cells and less efficiency of energy transfer to the PSII core (Fig. 4A, dotted line). The emission peak corresponding to allophycocyanin was shifted to longer wavelengths in the mutant, indicating that the energy trapped by phycobilisome could not be fully transferred to anchor and/or PSII core complexes. These results indicate that the efficiency of energy transfer from phycobilisomes to PSII is decreased in the mutant. When cells were excited by violet light to excite mainly the Soret band of chl a, small but distinct emissions at 685 and 695 nm were observed in addiPlant Physiol. Vol. 148, 2008

tion to a large emission peaking at 725 nm in wild-type cells (Fig. 4B, solid line). The emission spectrum of the mutant was practically the same as that of wild-type cells, although emission at 685 nm was higher (Fig. 4B, dotted line), suggesting that the relative content of PSII is little increased in the mutant (see Table IV below). These results indicate that the microenvironment of chlorophylls responsible for emitting those bands is not much altered in the mutant. Pigment Analysis

The mutant showed a characteristic absorption spectrum. The peak wavelength of chl a of the Soret band is shifted by about 10 nm to a longer wavelength (Fig. 3A, broken line). This suggests that components and/or the composition of the pigments are changed in the mutant. Pigments were extracted from wildtype and mutant cells and their compositions were analyzed by HPLC. On the pigment analysis of the mutant, we recognized that the retention time of chl a (39.4 min) was always a little shorter than that of the wild type (39.9 min; Fig. 5A). The absorption spectra of

Figure 3. Absorption spectra of wild-type and slr1923M cells. A, Absorption spectra of wild-type cells (solid line) and slr1923M cells (broken line) grown under normal conditions. The numbers of cells were adjusted to the same concentration with each other. B, Changes in absorption spectra of slr1923M cells grown under high light intensity (300 mE m22 s21). Absorption spectra were measured at 9.5 h (line a), 17 h (line b), 24 h (line c), and 33.5 h (line d) of cultivation. The base lines were shifted for the sake of clarity. 1071

Islam et al.

species of the mutant is DV-chl a, not MV-chl a (normal chl a). Table II shows the cellular contents of chlorophyll and carotenoids obtained by HPLC analysis. The cellular content of b-carotene was reduced by about 30%, while that of myxozanthophyll was not much different in the mutant compared with the wild type. Zeaxanthin was increased by about 50% in the mutant, suggesting that light intensity under the “normal” culture condition is strong for the mutant (Scha¨fer et al., 2006). On the other hand, chlorophyll content was decreased by about 45%. Accordingly, the ratios of myxoxanthophyll and zeaxanthin to chlorophyll were increased by about 1.5 and 2.7 times in the mutant compared with the wild type. The results obtained by HPLC analysis indicate that inactivation of the slr1923 gene will result in the total replacement of the chlorophyll species from MV type to DV type. The mutant cannot accomplish the reduc-

Figure 4. Fluorescence emission spectra of wild-type and slr1923M cells at 77 K. A, Cells were excited by blue light (Corning 4-96 filter), which excites chl a, carotenoids, and phycobilisomes. B, Cells were excited by violet light (Corning 4-96 + Toshiba V-42 filters), which preferentially excites chl a. Solid lines, wild type; dotted lines, slr1923M. The emission intensities were normalized to those of PSI maxima (725 nm).

separated chlorophyll were different between the wild type and the mutant (Fig. 5B). The spectra showed an absorption maximum at 660 nm in both the wild type and the mutant. The peak wavelength of the Soret band was 431 nm in the wild type; on the other hand, it was shifted to a longer wavelength by 11 nm in the mutant. It is also of note that chl a species were totally replaced by the new chl a species in the mutant. The spectral characteristics and the retention time on HPLC of the modified chlorophyll coincided with those reported by Shedbalkar and Rebeiz (1992), Nagata et al. (2005), and Nakanishi et al. (2005). They identified the modified chlorophyll molecule as a 3,8DV-chl a. It was thus suggested that the chlorophyll 1072

Figure 5. HPLC analysis of chlorophylls from wild-type and slr1923M cells. A, The elution profile of chl a of the wild type (WT; broken line) and slr1923M (solid line). B, Absorption spectra of chl a separated by HPLC. The broken line represents the wild type, and the solid line represents slr1923M. Plant Physiol. Vol. 148, 2008


Table II. Cellular content of pigments of wild-type and slr1923M cells analyzed by HPLC The contents are expressed as 310217 mol cell21. Chl a, MV-chl a (in the wild type) or DV-chl a (in the mutant). Strain

b-Carotene

Myxoxanthophyll

Zeaxanthin

Chl a

Wild type slr1923M

0.96 6 0.05 0.63 6 0.004

0.87 6 0.04 0.75 6 0.26

0.43 6 0.02 0.65 6 0.03

5.0 6 0.2 2.8 6 0.18

tion of the vinyl group of position 8 of ring B to an ethyl group. Analysis of Chlorophyll Molecules by 1H-NMR

In order to confirm the molecular species and determine the structure, 1H-NMR of the chlorophyll purified from the mutant was measured and compared with that of wild-type chlorophyll. Figure 6, A and B, shows 1H-NMR spectra of chlorophyll from the wild type and the mutant, respectively. In the 7.3 to 7.4 ppm region, the spectrum of wild-type chlorophyll shows signals of a doublet of doublets that is normally observed in MV-chl a (Fig. 6A, inset a). Vinyl proton signals are also exhibited at 5.25 and 5.45 ppm (Fig. 6A, inset b). A resonance signal of the ethyl group connected to position 8 of ring B was also observed in the 0.9 to 0.95 ppm region (Fig. 6A, inset c). An NMR spectrum of the chlorophyll purified from the mutant showed differences from the wild type at the above characteristic signal regions. The signal appearing in the 7.3 to 7.5 ppm region showed two sets of double doublets, instead of a single doublet of doublets (Fig. 6B, inset a), while their coupling constants were similar to each other. A difference was also observed in the 5.2 to 5.3 ppm region (Fig. 6B, inset b). It exhibited two well-separated sets of double doublets. In addition, a signal originating from the ethyl group is missing in the 0.90 to 0.95 ppm region (Fig. 6B, inset c). These results confirm the presence of two vinyl groups, one at position 3 of ring A and the other at position 8 of ring B, of the chlorophyll molecule in the mutant (Wu and Rebeiz, 1985). Judging from the results obtained from the absorption and 1H-NMR spectra, it was concluded that the molecular identity of the chlorophyll of the mutant is a 3,8-DV-chl a, not 3-vinyl 8-ethyl (MV-)chl a, which is present in wild-type cells. A mutant that accumulates DV-chl was recently reported independently in Arabidopsis by two groups (Nagata et al., 2005; Nakanishi et al., 2005). Their mutants accumulated DV-chl a and DV-chl b. Genetic analysis revealed that AT5G18660 is inactivated, which acts as a 3,8-divinyl Pchlide 8-vinyl reductase (DVR). Based on our results here and reported findings, we conclude that inactivation of slr1923 results in total inhibition of the conversion from DV-chl(ide) a to MV-chl(ide) a, indicating that the gene product of slr1923 is strongly related to divinyl reduction activity in Synechocystis 6803. Plant Physiol. Vol. 148, 2008

Ito et al. (2008) reached a similar conclusion for the same gene using bioinformatics tools. Their mutant, in which slr1923 was inactivated in Synechocystis 6803, also showed total replacement of the chlorophyll molecule from MV-chl a to DV-chl a. Photosynthetic Activities and Cellular Contents of Photosystems

The photosynthetic activities or reaction center contents were compared between the wild type and the mutant. The CO2 fixation rates on the basis of MV-chl a for the wild type and DV-chl a for the mutant were not much different. However, in the mutant, whole chain electron transport activity and segmented electron transport of PSII measured by a Clark-type O2 electrode were reduced (Table III). The reduced activities could be due to the reduced amount of reaction center or the decreased efficiency of energy transfer from phycobilisome or chlorophyll to the reaction centers. In fact, the light saturation curve of CO2 fixation activity indicated that the efficiency of utilization of light energy is reduced in the mutant, although the activity is the same at the saturating light intensities (data not shown). It should be noted that we observed a strong dark oxygen consumption activity in the presence of ascorbate and 2,6-dichlorophenolindophenol in the mutant for unknown reasons. This resulted in low activity of PSI electron transport in the mutant. Relative contents of PSI and PSII in slr1923M were determined and compared with those in the wild type (Table IV). It was found that PSI and PSII contents on the basis of chlorophyll were not very much different (80%–110% of wild-type values). On a cellular basis, however, both PSI and PSII contents were reduced in the mutant by about 40% and 20%, respectively, compared with the wild type. These results suggest that decreased chlorophyll content on the basis of the cell comes from decreased numbers of active reaction centers but that the number of chlorophylls bound to active reaction center complexes is not changed. These results strongly suggest that the amounts of DV-chl a synthesized are reduced and/or that degradation of the chlorophyll molecule is faster in the mutant. Phylogenetic Analysis of Slr1923 Homologues and Their Distribution

The results obtained in this study strongly suggest that the slr1923 gene product of Synechocystis 6803 is 1073

Islam et al. Figure 6. 1H-NMR spectra of the chl a molecule dissolved in acetone-D6. The chl a molecule was extracted and purified from wild-type or slr1923M cells. The purified chlorophyll was dissolved in deuterated acetone as described in “Materials and Methods,” and 1 H-NMR was measured for the wild type (A) and slr1923M (B).

1074



Table III. CO2 fixation activity and electron transport activities in wild-type and slr1923M cells Asc, Ascorbate; DCIP, 2,6-dichlorophenolindophenol; DCMU, 3-(3,4-dichlorophenyl)-1,1-dimethyl urea; MV, methyl viologen; p-BQ, p-benzoquinone. Activities were measured under the following conditions: H2O / CO2, 5 mM NaHCO3; H2O / MV, 1 mM MV, 10 mM methylamine, 1 mM NaN3, and 10 mM iodoacetamide; H2O / p-BQ, 4 mM p-BQ, 10 mM methylamine, 10 mM 2,5-dibromo-3-methyl-6isopropyl-p-benzoquinone, and 10 mM iodoacetamide; Asc/DCIP / MV, 2 mM Na-ascorbate, 0.1 mM DCIP, 1 mM MV, 10 mM DCMU, and 10 mM iodoacetamide. Activities are expressed as mmol O2 mg21 (MV- or DV-) chl a h21. Strain

Wild type slr1923M

CO2 Fixation or Electron Transport Activities H2O / CO2

H2O / MV

H2O / p-BQ

Asc/DCIP / MV

147.2 6 8.9 153.4 6 22.7

151.0 6 14.0 62.2 6 7.3

137.3 6 5.8 123.3 6 4.1

50.3 6 21.9 16.3 6 6.4

essential for the conversion of 3,8-DV-chl(ide) a to 3-MV-chl(ide) a. The Slr1923 protein has a molecular mass of 45 kD, deduced from its gene sequence, and is a soluble protein, as analyzed by the SOSUI program (http://sosui.proteome.bio.tuat.ac.jp). The homolog of Synechocystis 6803 is also found in Anabaena sp. PCC7120 (all1601; Kaneko et al., 2001), T. elongatus BP1 (tll1845; Nakamura et al., 2002), and Synechococcus sp. PCC6301 (syc0195_d). Interestingly, G. violaceus PCC7421 has two copies of its homolog (gll0878 and glr2543; Nakamura et al., 2003). The sequence alignment among the homologues indicated that Gll0878 has higher sequence similarity with other cyanobacterial counterparts. Comparison of the two reading frames of G. violaceus showed 42% identity and 60% homology. On the other hand, the identity and similarity of Gll0878 to Slr1923 were 70% and 81%, respectively, and those of Glr2543 to Slr1923 were 42% and 63%, respectively. The significance of the presence of two copies in G. violaceus is not clear at the moment. It was found that the homolog is present not only in cyanobacteria but also in eukaryotic algae such as the red alga C. merolae 10D (CMJ076C; Matsuzaki et al., 2004), a diatom, Thalassiosira pseudonana CMP 1335 (gw1.2.195.1; Armbrust et al., 2004), and the green algae Ostreococcus lucimarinus (OSTLU_36251; Palenik et al., 2007) and Chlamydomonas reinhardtii (CHLREDRAFT_106754). Surprisingly, it is also found in land plants such as Arabidopsis (AT1G04620) and rice (Oryza sativa; 4335472). The genes found in eukaryotic photosynthetic organisms are all encoded in nuclear genomes, and chloroplast-targeting signal peptides were found in their N-terminal flanking side ana-

lyzed by TargetP (http://www.cbs.dtu.dk/services/ TargetP). In addition to the oxygenic photosynthetic organisms, the slr1923 homologues are also found in the genomes of the purple bacteria Rhodopseudomonas palstris CGA009 (RPA1501) and Rhodobacter rubrum (Rru_A0937). Multiple sequence alignment from photosynthetic bacteria to higher plants constructed by ClustalW is shown in Figure 7. The aligned sequences indicate that the homologues have a common sequence cluster showing CXXCXXCX[12]C in the 30th to 50th amino acids, except R. palstris. This sequence motif is found in the N4 iron-sulfur cluster of Nqo3 protein, which is one of the subunits of NADH dehydrogenase of Thermus thermophilus (Sazanov and Hinchliffe, 2006). It is possible that Slr1923 bears an iron-sulfur cluster and transfers electrons from NAD (P)H to the substrate. It should be emphasized that R. palstris also has a four-Cys-containing cluster in the corresponding region, although the motif is different from that mentioned above. Thus, the homolog of R. palstris could also possess the same electron transfer activity. In addition, there are two highly conserved regions within the sequence. They are indicated by boxes in Figure 7. However, no motifs were picked up from the consensus sequences from motif databases. A phylogenetic tree of Slr1923 homologues was constructed by the neighbor-joining method (Saitou and Nei, 1987) using F420H2 dehydrogenase subunits as an outgroup. It indicates that the Slr1923 homologues are separated into two large groups: a purple bacterial group and other photosynthetic organisms

Table IV. Contents of PSI and PSII in wild-type and slr1923M cells on the basis of (MV- or DV-) chl a or cells Numbers in parentheses are expressed as 310217 mol cell21. P700 content of the mutant was calculated on the assumption that P700 chlorophyll is composed of DV-chl a in which the extinction coefficient of P700 is proportional to the ratio of that of extracted chl a (73.30) to DV-chl a (69.29; Shedbalkar and Rebeiz, 1992). Thus, the extinction coefficient, «slr1923, was estimated as «slr1923 = 643 (69.29/73.30) mM21 cm21. Strain

PSI Chl a/P700

PSII Chl a/PSII

Wild type slr1923M

99 6 4 (0.052 6 0.002) 121 6 11 (0.031 6 0.003)

461 6 40 (0.011 6 0.001) 419 6 13 (0.009 6 0.0003)


1075

Islam et al.

Figure 7. Multiple sequence alignment of Slr1923 and its homologues. The sequences were aligned by the ClustalW algorithm. The sequences whose genes are encoded in nuclei are trimmed to those of mature proteins by estimating the cleavage sites by Target P. Asterisks indicate strictly conserved residues, colons represent residues with high similarity, and periods represent residues with similarity.

(Fig. 8). Oxygenic photosynthetic organisms form a subcluster closely related to each other. It is of note that among them, a cyanobacterial group and green lineages were clearly separated. A red algal homolog, CMJ076C of C. merolae, was classified into the cyanobacterial group as reported by Ito et al. (2008). Interestingly, a homolog of the diatom T. pseudonana was categorized into a group of green lineages, although diatoms originated by second symbiosis of red algae. This is in sharp contrast with other genes, such as 16S rDNA, PsbA, or PsaA proteins, in which eukaryotic groups branched out from the cyanobacterial group. This suggests that slr1923 homologues of eukaryotic algae did not originate from some cyanobacterial strain. It is also noted that one of the homologues present in G. violaceus, gll0878, is classified into the cyanobacterial group, while glr2543 has highest affinity with the green bacterium Chlorobium phaerobacteroides homolog. Recently, Nagata et al. (2005) and Nakanishi et al. (2005) independently identified DVR in the land plant Arabidopsis. Chew and Bryant (2007) confirmed the presence of the corresponding enzyme in a green sulfur bacterium, C. tepidum TL3 (CT1063). On the other hand, Nagata et al. (2005) could not identify DVR homologues in most cyanobacteria. Only a small number 1076

of strains, such as Synechococcus WH8102 and Synechococcus spp. CC9605, CC9902, and CC9311, have DVR homologues. Slr1923 and its homologues found in other cyanobacteria could be important factors of the second type of vinyl reductase they predicted. Recently, Ito et al. (2008) identified the same gene by means of bioinformatics analysis, which is totally different from our approach. Interestingly, an slr1923 homolog was found in the green sulfur bacterium C. phaeobacteroides DSM266 (Cpha266_0188). This indicated that the slr1923 homolog is widely distributed among photosynthetic organisms irrespective of oxygen evolution. Neither the Slr1923 nor DVR homologous gene is found in divinylchlorophyll-harboring cyanobacteria, Prochlorococcaceae. It is also of note that photosynthetic organisms possess either DVR or slr1923 orthologues, except green lineages. The distribution of the slr1923 homologues has no relationship with the acquisition of the gene and taxonomy (Table V). DISCUSSION

We have created the slr1923 inactivation mutant and analyzed its phenotypes. slr1923 is located within a cluster comprising slr1924 and slr1925, separated by Plant Physiol. Vol. 148, 2008


Figure 8. Phylogenetic tree of Slr1923 homologues. A neighbor-joining tree was constructed based on the multiple sequence alignment constructed by the ClustalW algorithm. Archaean homologues were used as an outgroup.

only 25 bp from slr1923 (Fig. 1A). Northern analysis indicated that inactivation of slr1923, which is located upstream of the cluster, does not affect the transcription of other genes located downstream. The created mutant cells, slr1923M, exhibited a characteristic absorption spectrum. It has an absorption maximum at 678 nm in the red band, which is the same as that of the wild type. However, the Soret band was shifted to a longer wavelength by about 10 nm (Fig. 3A). Absorption and 1H-NMR spectral analyses of the extracted and purified chl a (Figs. 5B and 6) revealed that the chlorophyll species of the mutant is DV-chl a, not MVchl a (normal chl a). It is also found that all of the chlorophyll species present in the mutant were replaced with DV-chl a (Fig. 5A). According to the pathway of chlorophyll biosynthesis (Kim and Rebeiz, 1996), DV-(P)chl(ide) a is a precursor of MV-(P)chl(ide) a. Judged from the accumulation of DV-chl a in the mutant, it is strongly suggested that Slr1923 is essential to convert 3,8-divinyl(proto)chlorophyll(ide) a to 3-vinyl(proto)chlorophyll(ide) a of Synechocystis 6803. It is highly possible that Slr1923 of Synechocystis 6803 is a divinylchlorophyll(ide) reductase or a part of the Plant Physiol. Vol. 148, 2008

complex. Thus, we designate this gene cvrA (a gene indispensable for cyanobacterial vinyl reductase). The cellular content of DV-chl a of the slr1923M cells was reduced by about half compared with the content of MV-chl a of the wild type. Electron microscopic analysis has shown that the number of thylakoid membranes was reduced in the mutant (Ito et al., 2008). Reduction of chlorophyll content to about half accounts for the reduced number of thylakoid membranes. It is of note that DV-chl a is incorporated into PSI and PSII complexes and that they seem to be somehow functioning. However, the complexes do not seem to be fully functional and/or they are less stable. The inactivated mutant was sensitive to high-light illumination and started to die under such conditions (Fig. 2B). A similar phenotype was also observed with the dvr (AT5G18660) mutant of Arabidopsis (Nagata et al., 2005) and the slr1923 inactivation mutant of Synechocystis 6803 (Ito et al., 2008). They correlated the sensitivity with the accumulation of DV-chl and assumed that DV-chl completely substitutes for MV-chl in the preexisting protein pigment system and that this substitution leads to photodamage under high-light conditions. The same situation could take place in slr1923M. Ito et al. (2008) considered that free DV-chl a existed in the slr1923 mutant of Synechocystis 6803. Judged from fluorescence spectra at room temperature (data not shown) and 77 K (Fig. 4), however, we did not observe signals consistent with emissions from free chlorophyll. Thus, we conclude that all of the chlorophyll present in the mutant is bound to proteins. Incomplete matching of DV-chl a to the binding niche of PSI and PSII chlorophyll-protein complexes would lead to reduced efficiency of energy transfer within and between the complexes. This was found to be the case. The mutant shows a high fluorescence peak at 663 nm emitting from phycobiliproteins at room temperature (data not shown) and at 660 and 685 nm at 77 K emitting from phycobilisomes and the core linker polypeptide, LCM (Fig. 4A). This reflects the reduced efficiency of energy transfer from phycobilisomes to chlorophyll-protein complexes and the requirement of higher light intensity for the saturation of CO2 fixation (data not shown). It is also possible that the redox potential of DV-chl a is different from that of MV-chl a and, thus, that the efficiency of the photochemical reaction might be reduced. Determination of the redox potential of the DV-chl a molecule and reaction center is now in progress. The reduced contents of PSI and PSII on a cell basis (Table IV) can also be accounted for by the replacement of MV-chl a (normal chl a) with DV-chl a. Insufficient spatial fitting of DV-chl a to the binding site will bring about instability of the chlorophyll-protein complexes. Due to their instability, the degradation rate of the complex would be higher than in the wild type. When the mutant cells are grown under strong illumination, the degradation rate becomes further increased due to photodamage of the complex and finally exceeds the 1077

Islam et al.

Table V. Distribution of DVR (bciA) and cvrA homologues *, P. marinus = Prochlorococcus marinus. Gene Name (Gene Identifier)

Species

R. palstris CGA009 R. sphaeroides 2.4.1 C. tepidum TLS C. phaeobacteroides DSM266 Synechocystis PCC6803 T. elongatus BP1 Anabaena PCC7120 G. violaceus PCC7421 Synechococcus WH8102 P. marinus MED4* C. merolae 10D T. pseudonana CCMP 1335 C. reinhardtii A. thaliana O. sativa

dvr (bciA)

cvrA

2 rsp-3070 CT1603 2 2 2 2 2 SYNW 0963 2 2 2 CHLREDRAFT_195952 At5G18660 4332843

RPA1501 2 2 Cpha266_0188 slr1923 tll1845 all1601 gll0878, glr2543 2 2 CMJ076C Gw1.2.195.1 CHLREDRAFT_106754 At1G04620 4335472

synthesis rate of the complex, leading to a decrease in chlorophyll protein complexes under strong illumination. This was indeed the case depicted by the absorption spectral changes shown in Figure 3B. Similar phenomena were also observed by Nagata et al. (2005) and Ito et al. (2008). The deduced amino acid sequence of slr1923 has homology to that of FpoF protein, which functions as an electron input device of archaean NDH1 as an F420H2 oxidoreductase (Prommeenate et al., 2004). In archaea, F420 acts as an electron and proton carrier such as NAD (P)+ in bacteria or eukarya. F420 has a deazaflavin moiety within the molecule, and the structure of deazaflavin is quite similar to that of flavin. Although the structure of the redox center of F420 is similar to that of flavin, the manner of redox reaction is similar to that of NAD(P)+; F420 accepts two electrons and one proton, like NAD(P)+ when it is reduced. Taking into account the reports that vinyl reductase is specific to NADPH (Kolossov and Rebeiz, 2001; Nagata et al., 2005; Chew and Bryant, 2007; Ito et al., 2008) and the structural similarity of F420 to flavin, a flavin-binding site is expected in the Slr1923 protein. The deduced amino acid sequence shows the motif of a 4Fe4S-type iron-sulfur cluster on its N-terminal side (Fig. 7). This sequence motif is quite similar to that of Nqo3 protein of the N4 ironsulfur cluster of T. thermophilus (Sazanov and Hinchliffe, 2006). Thus, the following electron transfer pathway within the Slr1923 protein is assumed: electrons are transferred from NADPH to flavin, which is assumed to be bound to the C-terminal side, and then to the 4Fe4S-type iron-sulfur cluster situated on the N-terminal side. Another type of vinyl reductase (DVR) has been found in the land plants Arabidopsis and rice and in the marine cyanobacterium Synechococcus WH8102 (Nagata et al., 2005) and some other marine cyanobacteria. It is also found in the green sulfur bacterium C. tepidum TLS (Chew and Bryant, 2007). However, it was not found in other cyanobacteria. The existence of a 1078

second vinyl reductase that has low or no homology to DVR was predicted by Nagata et al. (2005), and Ito et al. (2008) reached a similar conclusion from slr1923. Indeed, the slr1923 (cvrA) homolog was found in other cyanobacteria, in the red alga C. merolae (Matsuzaki et al., 2004), in the diatom T. pseudonana (Armbrust et al., 2004), and in the green algae O. lucimarinus (Palenik et al., 2007) and C. reinhardtii. The deduced amino acid sequences of the homologues of Slr1923 have no homology to that of DVR. It is notable that a strain of green sulfur bacterium, C. phaeobacteroides DSM266, possesses a cvrA ortholog but not DVR, and the opposite is the case for R. sphaeroides 2.4.1 (Table V). The distribution of the two genes indicated that the cyanobacteria possess either a dvr or cvrA homolog, except Prochlorococcaceae. This clade does not have either gene. On the other hand, green lineages harbor both genes in their nuclei. There seems to be no relationship between the distribution of both proteins and taxonomy (Table V). Taking into account these facts, it is very likely that dvr and cvrA homologous genes were acquired independently, as pointed out by Ito et al. (2008). However, the real origin of the cvrA gene is still to be clarified, because the gene is assumed to have originated from archaea. In green lineages, both dvr and cvrA homologues are present. If both enzymes are functioning in the chloroplast, inactivation of dvr alone will not suffer from incapability of the conversion from DV-chlide to MVchlide. However, the DVR-inactivated mutant accumulated only DV-chl and no MV-chl (Nagata et al., 2005; Nakanishi et al., 2005). The following explanations could be possible. (1) The activity of the cvrA homolog is very low or negligible in higher plant chloroplasts. (2) Expression of the two genes is organ or tissue specific. For example, the dvr gene is expressed mainly in mesophyll cells of leaves, while the cvrA homolog could be expressed mainly in the epiderm of branches or trunks. (3) Expression depends on developmental stage. In rapidly growing plants, both Plant Physiol. Vol. 148, 2008


enzymes seem to be functioning, because DVR activity was observed in both membrane and soluble fractions (Rebeiz et al., 2003). However, in mature leaves, DVR is the only reductase that catalyzes the conversion from DV-(P)chlide to MV-(P)chlide. This hypothesis should be tested in young seedlings or by double mutation. (4) The function of the Slr1923 homolog in green lineages has been altered and the protein does not participate in the reduction of the vinyl group of 3,8-DV-chl(ide). Ito et al. (2008) raised the interesting possibility that the cvrA homolog is expressed during senescence in Arabidopsis. Thus, the function of Slr1923 of green lineages could be different from that in cyanobacteria. In any case, analysis of the enzymatic activity is necessary by expressed protein. MATERIALS AND METHODS Experimental Organism and Growth Conditions A Glc-tolerant Synechocystis 6803 strain was used as the wild-type organism. The cells were grown photoautotrophically in a liquid BG11 medium at 30°C. Cultures were grown under white light at a light intensity of 100 mE m22 s21 under 3% CO2-containing air, except where indicated otherwise. For measurements of growth rates, cells were grown in liquid BG11 medium and aliquots of the culture solution were sampled at various time points. The cell density was determined by monitoring the A750.

Mutant Construction and Segregation The nucleotide sequence of slr1923 was obtained from Cyanobase (http:// www.kazusa.or.jp/cyano/cyano.html). Two DNA fragments consisting of 549 bp of the N-terminal side (N-fragment, Nf) and 557 bp of the C-terminal side (C-fragment, Cf) of the slr1923 gene were amplified by PCR separately. For N-fragment amplification, forward (N-5) and reverse (N-3) primers with the sequences 5#-CATGACCGTTCCTGCCC-3# and 5#-GGGGAATTCCCACACAGGGCGTTCCC-3#, respectively, and for the C-fragment, forward (C-5) and reverse (C-3) primers with the sequences 5#-GGGGAATTCATGTTTCCCGGGCTGGG-3# and 5#-GAGGGACGTGGTCAGCC-3#, respectively, were used. The EcoRI recognition site was introduced into the 5# end of reverse (N-3) and forward (C-5) primers for the N- and C-fragments, respectively. The amplified N- and C-fragments were cloned separately into the pGEM-T vector (Promega) and introduced into competent Escherichia coli cells (XL-1 blue). N- and C-fragment-containing recombinant plasmids were cut out by double digestion by EcoRI and SpeI. The C-fragment, which was purified by gel electrophoresis, was ligated into gel-purified N-fragment containing pGEM-T vector (pGEM-Nf), giving rise to pGEM-Nf-Cf plasmid. A spectinomycin resistance cassette (SpR) cut out from pHSG398-SpR by EcoRI was inserted between N- and C-fragment-containing recombinant plasmid (pGEM-Nf-Cf), which was linearized by EcoRI digestion. The constructed recombinant vector (pGEM-Nf-SpR-Cf) was amplified in E. coli and purified for transformation. The wild-type Synechocystis 6803 cells were transformed by the resulting plasmid pGEM-Nf-SpR-Cf vector. The transformed cells, which have a disrupted slr1923 gene, were segregated by successive streaks (about five to six times) on BG11 agar plates that contained 20 mg mL21 spectinomycin. The segregation was confirmed by PCR using N-fragment forward (N-5) and C-fragment reverse (C-3) primers, with genomic DNA prepared from mutant cells as a template.

Isolation of Total RNA and Northern Hybridization Total RNA from wild-type and mutant cells at mid-log phase was extracted according to Hihara et al. (1998) with some modifications. The RNA was separated by agarose gel electrophoresis, blotted onto a nylon membrane, and fixed by UV illumination. Specific probes for slr1923, slr1924, and slr1925 were prepared by PCR using specific primers. The primers used for the slr1923 probe were the same as those used for the segregation check. The probes for slr1924 and slr1925 were obtained by amplification of the coding region by


PCR using primers 5#-ATCCTTATTTATGGTCGG-3# and 5#-GCCAAGCTTGGCTGGTCG-3# for slr1924 and 5#-ATGATGGCGGTAGATCCC-3# and 5#-TATTGAAGCTTATCTCCG-3# for slr1925. The amplified probes were labeled by digoxigenin following the manufacturer’s instructions (Roche). Luminescence from the hybridized probe was detected with a Bio-Image analyzer (Fuji Photo Film).

Measurements of Absorption and Fluorescence Spectra Absorption spectra of the intact cells were measured with a Shimadzu MPS-2000 spectrophotometer (Shimadzu). Fluorescence emission spectra were recorded by a laboratory-constructed setup at 77 K (Yamane et al., 1997). Excitation light from a 100-W halogen lamp was passed through a Corning 4-96 filter to excite chlorophyll, carotenoids, and phycobilisomes or a combination of Corning 4-96 and Toshiba V-42 filters to excite chlorophyll. The fluorescence intensities were normalized to those of PSI maxima (725 nm).

Pigment Analysis A cell suspension (1.3 mL) in which the A750 was adjusted to 0.2 was precipitated, and then 10 mL of pure water and 190 mL of dimethyl formamide were added. After a short vortexing, the solution was transferred to a brown plastic bottle and kept at 220°C overnight. After a short vortexing, an aliquot of 50 mL was withdrawn and centrifuged. Ten microliters of the supernatant was injected onto an HPLC column (Phenomenex). Other chemicals and instrumental settings were as described by Kashino and Kudoh (2003). Data were transferred to a personal computer, and the amounts of the pigments were calculated from their extinction coefficients at the monitoring wavelength.

Preparation of Thylakoid Membranes Cells were harvested by centrifugation (3,200g, 5 min) and washed in buffer A (25% [w/v] glycerol, 10 mM CaCl2, 10 mM MgCl2, and 50 mM MESNaOH [pH 6.5]). The precipitated cells were resuspended in buffer A. The cells were ruptured by zirconia beads (f = 0.1 mm, 1.6 g mL21 cell suspension) with the Mini Bead Beater (Cole-Parmer). Agitation was performed for 20 s followed by cooling on ice for 1 min to minimize heat denaturation. This cycle was repeated three times. The supernatant after precipitation of zirconia beads was recovered and centrifuged at 6,000g for 10 min at 4°C to remove unbroken cells. Thylakoid membranes and soluble fractions were separated by centrifugation at 300,000g for 30 min. The thylakoid membranes recovered as a precipitate were resuspended in buffer A and stored at 280°C until use.

Chlorophyll Purification and NMR Analysis Chlorophylls were extracted by sonicating the mixture of 1.5 mL of thylakoid membrane preparation and 6.0 mL of acetone-Na2HPO4 for 5 min. The resulting solution was filtered by a filter paper and then a glass filter. Filtrate volume was adjusted to 10 mL with acetone, and 1.33 mL of dioxane was added with stirring on ice. Then, water was slowly added until green aggregates were formed. Samples were kept at 220°C for 1 h and centrifuged at 2,300g for 15 min at 4°C. The green pellet was dissolved in 3 mL of ethanol and again filtered through a glass filter. Chlorophyll purification was repeated to remove contaminating carotenoids. Finally, purified chlorophyll was lyophilized overnight. Dried samples were dissolved in 700 mL of NMR-grade deuterated acetone. Two hundred microliters of the sample was loaded in the NMR machine. 1H-NMR was measured on a JEOL ECP-600 at 600 MHz using a solvent peak as a reference (J = 2.00).

Measurement of Oxygen Evolution or Consumption Activities Cells were suspended in liquid BG11 medium, and chlorophyll concentration was adjusted to 20 mg mL21. Oxygen evolution or consumption activities were measured by a Clark-type oxygen electrode (Rank Brothers) as described previously (Inoue et al., 2001).

Determination of PSI and PSII For the determination of PSI content, light-induced absorption changes of P700 were measured at 703 nm by a Shimadzu MPS-2000 spectrophotometer by thylakoid membranes at 15 mg chl mL21 as described by Nakayama et al. (1979). P700 content was determined using an extinction coefficient of 64 mM21

1079

Islam et al. cm21 (Hiyama and Ke, 1972) for the wild type. The extinction coefficient of the DV-chl a-containing mutant was estimated according to the ratio of extinction coefficients of pure MV- and DV-chlorophylls reported by Shedbalkar and Rebeiz (1992). Intact cell suspension was used for the measurement of PSII content. Oxygen evolution under repetitive flash excitation (10-ms duration) was recorded by a Clark-type O2 electrode (Rank Brothers) at 30°C in the presence of 5 mM NaHCO3 and 1 mM K3Fe(CN)6. The PSII content was determined from an average yield of oxygen per flash on the basis of chlorophyll and then on the basis of cells.

Phylogenetic Analysis of slr1923 Homologues The deduced amino acid sequences of slr1923 homologues were trimmed to the predicted mature forms by TargetP (http://www.cbs.dtu.dk/services/ TargetP/) for nucleus-encoded genes. The amino acid sequences were aligned by the ClustalW algorithm. A neighbor-joining tree was constructed based on multiple sequence alignment (Saitou and Nei, 1987). Sequence data from this article can be found in the GenBank/EMBL data libraries. Accession numbers of cvrA homologues are as follows: Rhodopseudomonas palstris CAE26943, Q6N9N9; Roseobacter denitrificans, Q161A4; Jannaschia sp. CCS1, Q28VP5; Halorhodospira halophila, A1WXJ2; Chlorobium phaeobacteroides DSM266, A1BCX8; Synechocystis sp. PCC6803, P74473; Thermosynechococcus elongatus BP1, Q8DHV0; Anabaena sp. PCC7120, P46015; Gloeobacter violaceus PCC7421 gll0878, Q7NM89; Gloeobacter violaceus PCC7421 glr2543, Q7NHJ2; Trichodesmium erythraeum, Q118A1; Synechococcus sp. PCC7942, Q31NI0; Cyanobacteria Yellowstone B-Prime (Cyanobacteria CYB), 3318347HEL; Synechococcus sp. RCC307, A5GS18; Synechococcus sp. WH7803, A5GM11; Ostreococcus lucimarinus, A4RSL2; Chlamydomonas reinhardtii, A8JC46; Arabidopsis thaliana, Q8GS60; Oryza sativa, O23023; Methenosarcina berkeri FpoF, 3223320LC; Methenosarcina mazei FpoF, Q8PZ67; Methenosarcina berkeri beta, 3223320VV; Methenosarcina mazei beta, Q46A78. Accession numbers of DVR are as follows: Rhodobacter sphaeroides 2.4.1, Q3IXP7; Chlorobium tepidum TLS, Q8KDI7; Synechococcus WH8102, Q7U7L8; Chlamydomonas reinhardtii, A8HMQ3; Arabidopsis thaliana, Q1H537; Oryza sativa, Q10LH0.

ACKNOWLEDGMENTS We are especially thankful to Prof. Takashi Sugimura and Ms. Aya Inoue (Department of Material Science, Graduate School of Material Science, University of Hyogo) for their great help with 1H-NMR measurements and analysis. We are also grateful to Yohei Ikeda and Takeshi Takahashi (Graduate School of Life Science) for helping with HPLC manipulation and data analysis and to M. Ikeuchi (Department of Life Sciences [Biology], University of Tokyo, Komaba) for helpful discussions. Received May 18, 2008; accepted August 14, 2008; published August 20, 2008.

LITERATURE CITED Armbrust EV, Berges JA, Bowler C, Green BR, Martinez D, Putnam NH, Zhou S, Allen AE, Apt KE, Bechner M, et al (2004) The genome of the diatom Thalassiosira pseudonana: ecology, evolution, and metabolism. Science 306: 79–86 Bazzaz MB (1981) New chlorophyll chromophores isolated from a chlorophyll deficient mutant of maize. Photobiochem Photobiophys 2: 199–207 Bazzaz MB, Govindjee (1974) Effects of lead chloride on chloroplast reactions. Environ Lett 6: 175–191 Beale SI (1993) Biosynthesis of cyanobacterial tetrapyrrole pigments: hemes, chlorophylls, and phycobillins. In DA Bryant, ed, The Molecular Biology of Cyanobacteria. Kluwer Academic Publishers, Dordrecht, The Netherlands, pp 519–558 Beale SI (1999) Enzymes of chlorophyll biosynthesis. Photosynth Res 60: 43–73 Chew AGM, Bryant DA (2007) Characterization of a plant-like protochlorophyllide a divinyl reductase in green sulfur bacteria. J Biol Chem 282: 2967–2975

1080

Hihara Y, Sonoike K, Ikeuchi M (1998) A novel gene, pmgA, specifically regulates photosystem stoichiometry in the cyanobacterium, Synechocystis species PCC6803 in response to high light. Plant Physiol 117: 1205–1216 Hiyama T, Ke B (1972) Difference spectra and extinction coefficients of P700. Biochim Biophys Acta 267: 160–171 Inoue N, Taira Y, Emi T, Yamane Y, Kashino Y, Koike H, Satoh K (2001) Acclimation to the growth temperature and the high-temperature effects on photosystem II and plasma membranes in a mesophilic cyanobacterium, Synechocystis sp. PCC6803. Plant Cell Physiol 42: 1140–1148 Ito H, Yokono M, Tanaka R, Tanaka A (2008) Identification of a novel vinyl reductase gene essential for the biosynthesis of monovinyl chlorophyll in Synechocystis sp. PCC6803. J Biol Chem 283: 9002–9011 Kaneko T, Nakamura Y, Wolk CP, Kuritz T, Sasamoto S, Watanabe A, Iriguchi M, Ishikawa A, Kawashima K, Kimura T, et al (2001) Complete genomic sequence of the filamentous nitrogen-fixing cyanobacterium Anabaena sp. strain PCC 7120. DNA Res 8: 205–213 Kaneko T, Sato S, Kotani H, Tanaka A, Asamizu E, Nakamura Y, Miyajima N, Hirosawa M, Sugiura M, Sasamoto S, et al (1996) Sequence analysis of the genome of the unicellular cyanobacterium Synechocystis sp. strain PCC 6803. II. Sequence determination of the entire genome and assignment of potential protein-coding regions. DNA Res 3: 109–136 Kashino Y, Kudoh S (2003) Concerted response of xanthophyll-cycle pigments in a marine diatom, Chaetoceros gracilis, to shifts in light condition. Phycol Res 51: 168–172 Kim JS, Rebeiz CA (1996) Origin of the chlorophyll a biosynthetic heterogeneity in higher plants. J Biochem Mol Biol 29: 327–334 Kolossov VL, Bohnert HJ, Rebeiz CA (2006) Chloroplast biogenesis 92: in situ screening for divinyl chlorophyll(ide) a reductase mutants by spectrofluorometry. Anal Biochem 348: 192–197 Kolossov VL, Rebeiz CA (2001) Chloroplast biogenesis 84: solubilization and partial purification of membrane-bound [4-vinyl]chlorophyllide a reductase from etiolated barley leaves. Anal Biochem 295: 214–219 Matsuzaki M, Misumi O, Shin-i T, Maruyama S, Takahara M, Miyagishima S, Mori T, Nishida K, Yagasawa F, Nishida K, et al (2004) Genome sequence of the ultrasmall unicellular red alga Cyanidioschyzon merolae 10D. Nature 428: 653–657 Nagata N, Tanaka R, Satoh S, Tanaka A (2005) Identification of a vinyl reductase gene for chlorophyll synthesis in Arabidopsis thaliana and implications for the evolution of Prochlorococcus species. Plant Cell 17: 233–240 Nakamura Y, Kaneko T, Sato S, Ikeuchi M, Katoh H, Sasamoto S, Watanabe A, Iriguchi M, Kawashima K, Kimura T, et al (2002) Complete genome structure of the thermophilic cyanobacterium Thermosynechococcus elongatus BP-1. DNA Res 9: 123–130 Nakamura Y, Kaneko T, Sato S, Mimuro M, Miyashita H, Tsuchiya T, Sasamoto S, Watanabe A, Kawashima K, Kishida Y, et al (2003) Complete genome structure of Gloeobacter violaceus PCC 7421, a cyanobacterium that lacks thylakoids. DNA Res 10: 137–145 Nakanishi H, Nozue H, Suzuki K, Kaneko Y, Taguchi G, Hayashida N (2005) Characterization of the Arabidopsis thaliana mutant pcb2 which accumulates divinyl chlorophylls. Plant Cell Physiol 46: 467–473 Nakayama T, Yamaoka T, Katoh S (1979) Chromatographic separation of photosystems I and II from the thylakoid membrane isolated from a thermophilic blue-green alga. Plant Cell Physiol 20: 1565–1576 Ohtsuka M, Oyabu J, Kashino Y, Satoh K, Koike H (2004) Inactivation of ycf33 results in an altered cyclic electron transport pathway around photosystem I in Synechocystis sp. PCC 6803. Plant Cell Physiol 45: 1243–1251 Palenik B, Grimwood J, Aerts A, Rouze P, Salamov A, Putnam N, Dupont C, Jorgensen R, Derelle E, Rombauts S, et al (2007) The tiny eukaryote Ostreococcus provides genomic insights into the paradox of plankton speciation. Proc Natl Acad Sci USA 104: 7705–7710 Prommeenate P, Lennon AM, Markert C, Hippler M, Nixon PJ (2004) Subunit composition of NDH-1 complexes of Synechocystis sp. PCC 6803. J Biol Chem 279: 28165–28173 Rebeiz CA, Kolossov VL, Briskin D, Gawienowski M (2003) Chloroplast biogenesis: chlorophyll biosynthetic heterogeneity, multiple biosynthetic routes, and biological spin-offs. In HS Nalwa, ed, Handbook of Photochemistry and Photobiology. American Scientific, Los Angeles, pp 183–248



Saitou N, Nei M (1987) The neighbor-joining method: a new method for reconstructing phylogenetic trees. Mol Biol Evol 4: 406–425 Sazanov LA, Hinchliffe P (2006) Structure of the hydrophilic domain of respiratory complex I from Thermus thermophilus. Science 311: 1430–1436 Scha¨fer L, Sandmann M, Woitsch S, Sandmann G (2006) Coordinate upregulation of carotenoid biosynthesis as a response to light stress in Synechococcus PCC7942. Plant Cell Environ 29: 1349–1356 Senge MO, Smith KM (1995) Biosynthesis and structure of the bacteriochlorophylls. In RE Blankenship, MT Madigan, CE Bauer, eds, Anoxygenic Photosynthetic Bacteria. Kluwer Academic Publishers, Dordrecht, The Netherlands, pp 137–151


Shedbalkar VP, Rebeiz CA (1992) Chloroplast biogenesis: determination of the molar extinction coefficients of divinyl chlorophyll a and b and their pheophytins. Anal Biochem 207: 261–266 Suzuki JY, Bauer CE (1995) Altered monovinyl and divinyl protochlorophyllide pools in bchJ mutants of Rhodobacter capsulatus. J Biol Chem 270: 3732–3740 Wu SM, Rebeiz CA (1985) Chloroplast biogenesis: molecular structure of chlorophyll b (E489 F666). J Biol Chem 260: 3632–3634 Yamane Y, Kashino Y, Koike H, Satoh K (1997) Increases in the fluorescence Fo level and reversible inhibition of photosystem II reaction center by hightemperature treatments in higher plants. Photosynth Res 52: 57–64

1081