Journal list menu

Volume 271, Issue 4 p. 685-693
Free Access

Differing involvement of sulfoquinovosyl diacylglycerol in photosystem II in two species of unicellular cyanobacteria

Motohide Aoki

Motohide Aoki

School of Life Science, Tokyo University of Pharmacy and Life Science, Horinouchi, Hachioji, Tokyo, Japan

Search for more papers by this author
Norihiro Sato

Norihiro Sato

School of Life Science, Tokyo University of Pharmacy and Life Science, Horinouchi, Hachioji, Tokyo, Japan

Search for more papers by this author
Ayano Meguro

Ayano Meguro

School of Life Science, Tokyo University of Pharmacy and Life Science, Horinouchi, Hachioji, Tokyo, Japan

Search for more papers by this author
Mikio Tsuzuki

Mikio Tsuzuki

School of Life Science, Tokyo University of Pharmacy and Life Science, Horinouchi, Hachioji, Tokyo, Japan

CREST of Japan

Search for more papers by this author
First published: 26 January 2004
Citations: 73
M. Tsuzuki, School of Life Science, Tokyo University of Pharmacy and Life Science, Horinouchi, Hachioji, Tokyo 192-0392, Japan.
Fax: +81 426 76 6721, Tel.: +81 426 76 6713,
E-mail: [email protected]

Abstract

Sulfoquinovosyl diacylglycerol (SQDG) is involved in the maintenance of photosystem II (PSII) activity in Chlamydomonas reinhardtii[Minoda, A., Sato, N., Nozaki, H., Okada, K., Takahashi, H., Sonoike, K. & Tsuzuki, M. et al. (2002) Eur. J. Biochem.269, 2353–2358]. To understand the spread of the taxa in which PSII interacts with SQDG, especially in cyanobacteria, we produced a mutant defective in the putative sqdB gene responsible for SQDG synthesis from two cyanobacteria, Synechocystis sp. PCC6803 and Synechococcus sp. PCC7942. The mutant of PCC6803, designated SD1, lacked SQDG synthetic ability and required SQDG supplementation for its growth. After transfer from SQDG-supplemented to SQDG-free conditions, SD1 showed decreased net photosynthetic and PSII activities on a chlorophyll (Chl) basis with a decrease in the SQDG content. Moreover, the sensitivity of PSII activity to 3-(3,4-dichlorophenyl)-1,1-dimethylurea and atrazine was increased in SD1. However, SD1 maintained normal amounts of cytochrome b559 and D1 protein (the subunits comprising the PSII complex) on a Chl basis, indicating that the PSII complex content changed little, irrespective of a decrease in the SQDG content. These results suggest that the role of SQDG is the conservation of the PSII properties in PCC6803, consistent with the results obtained with C. reinhardtii. In contrast, the SQDG-null mutant of PCC7942 showed the normal level of PSII activity with little effect on its sensitivity to PSII herbicides. Therefore, the difference in the SQDG requirement for PSII is species-specific in cyanobacteria; this could be of use when investigating the molecular evolution of the PSII complex.

Abbreviations

  • Chl
  • chlorophyll
  • Cyt
  • cytochrome
  • DCMU
  • 3-(3,4-chlorophenyl)-1,1-dimethylurea
  • DGDG
  • digalactosyl diacylglycerol
  • IC50
  • 50% inhibitory concentration
  • MGDG
  • monogalactosyl diacylglycerol
  • PG
  • phosphatidylglycerol
  • PSII
  • photosystem II
  • SQDG
  • sulfoquinovosyl diacylglycerol
  • Biomembranes constructed predominantly from lipids and proteins are sites of energy production, metabolism such as lipid synthesis, and the transportation of substances such as metabolites between the inside and outside of membrane systems. Thylakoid membranes in plant chloroplasts and cyanobacterial cells are unique in possessing photosynthetic electron transport and photophosphorylation systems for the conversion of light to chemical energy. The lipid composition of thylakoid membranes is highly conserved among higher plants, algae, and cyanobacteria, comprised mainly of the following four glycerolipids, monogalactosyl diacylglycerol (MGDG), digalactosyl diacylglycerol (DGDG), phosphatidylglycerol (PG), and sulfoquinovosyl diacylglycerol (SQDG) [1–3]. PG and SQDG possess negatively charged head groups, whereas MGDG and DGDG are noncharged lipids.

    Lipid analyses of subfractions and protein complexes of thylakoid membranes have shown specific lipid compositions, implying roles of some lipid classes in photosynthesis [4]. Significant differences in lipid composition and the fatty acid saturation state were observed, e.g. in extraction of photosystem II (PSII) from photosynthetic membranes with different detergents [5–7]. On the other hand, reconstitution of photosynthetic activity with the addition or removal of thylakoid lipids in vitro has also been performed with thylakoid membranes, subfractions of thylakoid membranes and purified protein complexes, with important observations being summarized in a review by Trémolières and Siegenthaler [8]. However, the conclusion drawn from in vitro experiments is not necessarily valid in vivo.

    The mutants of some lipid metabolism pathways have proved to be useful tools for investigating in vivo the specific roles of lipids in photosynthesis. A mutant of Chlamydomonas reinhardtii, defective in SQDG (hf-2), showed PSII activity that was ≈ 40% lower than that of the wild-type, an increase in sensitivity of the PSII activity to 3-(3,4-dichlorophenyl)-1,1-dimethylurea (DCMU), and a lower growth rate [9–12]. In accordance with these observations, the incubation of isolated thylakoid membranes of hf-2 with SQDG in vitro reversed the lowered PSII activity. Therefore, we concluded that SQDG has the specific function of maintaining PSII properties. We were then faced with two points that required clarification: what is the molecular mechanism of this maintenance function and does PSII interact with SQDG in every taxonomic group in which it is present? Indeed, Benning et al. [13] and Güler et al. [14] reported that the lack of SQDG did not affect growth or photosynthesis detrimentally in Rhodobacter sphaeroides and Synechococcus sp. PCC7942, respectively. In order to understand the second point, especially in cyanobacteria, we isolated SQDG-deficient mutants of Synechocystis sp. PCC6803 and Synechococcus sp. PCC7942 by disruption of a gene for SQDG synthesis. We will discuss the differences in SQDG requirement even between the unicellular cyanobacteria used in this study.

    Materials and methods

    Organisms and culture conditions

    The glucose-tolerant wild-type strain of Synechocystis sp. PCC6803 and the wild-type strain of Synechococcus sp. PCC7942 were provided by M. Ikeuchi (The University of Tokyo) and E. Suzuki (Akita Prefectural University, Japan), respectively. Their mutants were grown photoautotrophically at 30 °C in BG-11 medium [15] supplemented with 30 mm Hepes/NaOH pH 7.5, with air bubbling, or BG-11/1.2% agar plates. Light, 20 µEin·m−2·s−1, was provided constantly by fluorescent lamps. SQDG prepared from the wild-type cells of Synechocystis sp. PCC6803 was sonicated in the growth medium for liposome production (5 mm) and filtered for sterilization. The growth medium was supplemented with the SQDG liposomes at the indicated concentrations. Growth of the cells was monitored by determination of the absorption at 730 nm or by determining the cell number with a new-Neubauer type haemocytometer (Erma, Tokyo, Japan).

    Amplification and disruption of a putative sqdB gene in Synechocystis sp. PCC6803 and Synechococcus sp. PCC7942

    A homologue of the sqdB gene product of Synechococcus sp. PCC7942 responsible for SQDG synthesis was searched for in the genome database of Synechocystis sp. PCC6803 at the Kazusa DNA Research Institute with the use of the blastp system (Cyano Base, http://www.kazusa.or.jp/cyano/). To inactivate the putative sqdB gene of Synechocystis sp. PCC6803, we amplified a DNA fragment, containing its ORF, by PCR with Ex-Taq DNA polymerase (Takara, Ohtsu, Japan) under the following thermo-cycle conditions: 2 min at 95 °C, followed by 25 cycles of 30 s at 94 °C, 1 min at 57 °C, and 3 min at 72 °C. The primers used were (A:5′-CCGGAATTCATGAGAGCTCTGGTTAT TGG-3′) and (B: 5′-TTTCTGCAGCTAGCCGCGCCAG GTAACTT-3′). The sites shown in bold type in primers A and B are added restriction sites EcoRI and PstI, respectively. The amplified DNA fragment was cut with EcoRI and PstI to be cloned into the pUC119 vector. At one restriction site, BalI, of the putative gene of the plasmid, we inserted a spectinomycin-resistant cassette prepared from pRL453 by digestion with SmaI. The plasmid DNA containing the disrupted putative sqdB gene was used for transformation into the wild-type Synechocystis sp. PCC6803 by homologous recombination, as described by Golden and Sherman [16]. Spectinomycin-resistant transformants were screened on selection plates containing 20 µm spectinomycin and 50 µm SQDG, and thereafter subcultured several times to segregate the chromosomes until all copies of the putative sqdB gene had been disrupted. A mutant with complete disruption of the putative sqdB gene was confirmed by PCR using primers A and B.

    The sqdB gene of Synechococcus sp. PCC7942 previously reported by Güler et al. [14] was amplified by PCR with primers 5′-TTGGGTGGCGATGGTTTCTG-3′ and 5′-GAGAGAGTGCGACTTTAGCG-3′, and then cloned into the pGEM-T Easy (Promega) vector for production of the plasmid, pSBC. The plasmid pSBC was cut with BalI for excision of the 0.3-kbp fragment of the sqdB gene, and then ligated with a kanamycin-resistant gene derived from pHSG298 by digestion with HincII/StuI. Transformation of Synechococcus sp. PCC7942 with this vector was performed in the presence of 40 µm kanamycin to select disruptants of the sqdB gene.

    Preparation of thylakoid membranes

    Cells were washed with a buffer comprising 5 mm Hepes/NaOH pH 7.5 and 10 mm NaCl and then agitated with a vortex mixer in this buffer containing zirconia/silica-beads (0.1-mm diameter; Biospec, OK, USA) and recovered as whole cell extracts. Agitation for 30 s was repeated four times with ice-cooling intervals of 60 s. The cell extracts were then centrifuged at 3000 g for 5 min at 4 °C for removal of cell debris, followed by centrifugation of the supernatants at 30 000 g for 20 min at 4 °C. The resultant precipitate was washed several times, resuspended in buffer and recovered as thylakoid membranes.

    Lipid analysis and preparation of SQDG

    Lipids were extracted from cells of the wild-types and their mutants according to the method of Bligh and Dyer [17], and then separated into individual lipid classes by one-dimensional TLC with a solvent system of chloroform/methanol/ammonia (65 : 35 : 5, v/v/v). The lipid content was determined as described previously [9]. SQDG for supplementation to the medium was prepared from lipids of the wild-type cells of Synechocystis sp. PCC6803 by ion exchange column chromatography. DEAE-Toyopearl 650M (TOSO, Tokyo, Japan) was suspended in 1 m sodium acetate buffer pH 7.0, washed with water and then with methanol, and finally packed to a height of 5 cm in a column (3 cm diameter). The column was washed with ≈ 5 column vols chloroform/methanol (1 : 4, v/v), followed by application of the lipids dissolved in the chloroform/methanol to the column. Polar lipids such as MGDG and DGDG, and pigments such as chlorophyll and carotenoids were eluted with ≈ 5 column vols of the chloroform/methanol. PG and SQDG were then eluted in order of decreasing polarity with a linear gradient from the chloroform/methanol to the chloroform/methanol containing 0.2% (w/v) ammonium acetate. The SQDG fraction was desalted by gel chromatography (Tskgel-Toyopearl HW40C, TOSO, Tokyo, Japan) with elution with the chloroform/methanol. SQDG was then dissolved in chloroform/methanol (2 : 1, v/v) and stored at −20 °C until use.

    Determination of pigments content and PSII complex

    After chlorophyll was extracted from cells with 100% methanol, content was determined according to the method of Porra et al. [18]. The total amounts of phycobiliproteins were estimated from the differential absorption spectrum obtained by subtracting the absorbance of heat-treated cells from that of untreated cells at 620 nm, as described by Zhao and Brand [19]. The PSII content was determined by spectrophotometric measurement of oxidized minus reduced difference spectra of Cyt b559, as described previously by Fujita and Murakami [20]. We used a differential extinction coefficient of Cyt b559 that had been reported by Garewal and Wasserman [21]. All spectroscopic determinations were performed with a Beckman DU640 spectrometer (Beckman).

    Measurement of photosynthetic activity

    Photosynthetic O2 evolution (net photosynthesis) was measured for intact cells equivalent to 2.5 µg Chl·ml−1 in culture medium containing 10 mm NaHCO3 using a Clark-type oxygen electrode (Rank Brothers, Cambridge, UK). PSII activity was measured for cells equivalent to 2.5 µg Chl·ml−1 in a solution comprising 50 mm Tricine/KOH pH 7.5, 2 mm NH4Cl, 2 mmp-benzoquinone, as an artificial electron acceptor. For measurement of these photosynthetic activities, the reaction mixture was kept at 30 °C and illuminated with a tungsten projector lamp (1000 µEin· m−2·s−1).

    Protein analysis

    An equal volume of 2× electrophoresis sample buffer (0.125 m Tris/HCl pH 6.8, 20% glycerol, 4% SDS, 1.2% 2-mercaptoethanol, and ≈ 0.002% Bromophenol blue) was used to solubilize the proteins of the whole cell extracts or thylakoid membranes. The solubilized proteins were subjected to SDS/PAGE with a 16–22% (w/v) linear gradient polyacrylamide gel containing 7.5 m urea, as described by Ikeuchi and Inoue [22]. Proteins were stained with Coomassie brilliant blue. For immunoblot analysis, proteins separated by SDS/PAGE were electrophoretically transferred to a poly(vinylidene difluoride) (PVDF) membrane (Immobilon-P PVDF membrane; Millipore). After being blocked with 5% nonfat dried milk in Tris-buffered saline containing 0.1% Tween-20 (TBS/Tween), the PVDF membrane was incubated with 5000-fold diluted polyclonal antibodies against D1 protein of spinach [23]. The membrane was washed with TBS/Tween, and then treated with anti-rabbit IgG, the second antibody, which was coupled with peroxidase. The peroxidase complex was detected with an ECL-plus Western blotting detection kit (Amersham Biosciences) according to the manufacturer's instructions.

    Results

    Inactivation of the putative sqdB gene in Synechocystis sp. PCC6803

    To determine the physiological function of SQDG, we tried to inactivate a putative gene for SQDG synthesis in Synechocystis sp. PCC6803. So far, sqdB genes have been found to participate in SQDG synthesis in some photosynthetic organisms such as R. sphaeroides, Synechococcus sp. PCC7942, and Arabidopsis thaliana. An ORF, slr1020, in the genomic database of Synechocystis sp. PCC6803 was found to encode a polypeptide which is homologous to the sqdB gene product of Synechococcus sp. PCC7942. The deduced amino acid sequence of slr1020, corresponding to a polypeptide of 383 amino acid residues with an approximate molecular mass of 43 kDa, was 44%, 43% and 71% identical to those of the sqdB gene products of Synechococcus sp. PCC7942 and R. sphaeroides, and the SQD1 gene product of A. thaliana, respectively (data not shown). The putative sqdB gene was then disrupted by insertion of a spectinomycin-resistant cassette into the coding region of slr1020, as shown in Fig. 1A. To determine whether or not all copies of the putative gene were disrupted in the genome of the mutant with the spectinomycin-resistant phenotype, we carried out PCR analysis of the wild-type and the mutant, which was designated SD1 (Fig. 1B). A fragment of 1.2 kbp was detected for the wild-type, whereas only one of 3.2 kbp owing to insertion of the 2.0 kbp cassette was observed for the mutant, indicating that all copies of the putative gene were disrupted in SD1.

    Details are in the caption following the image

    Insertional mutagenesis of slr1020 of Synechocystis sp. PCC6803. (A) Disruption of slr1020 with insertion of the Smr/Spr gene in Synechocystis sp. PCC6803. The directions of transcription of the Smr/Spr gene and slr1020 are indicated by open boxed arrows. The PCR primers, A and B, are indicated by open arrows. (B) PCR analysis of slr1020 in the wild-type (WT) and mutant (SD1) of Synechocystis sp. PCC6803. Lane M contains HindIII-digested λDNA size markers. The sizes of amplified fragments are indicated on the right.

    Characteristics of SD1 as to lipid phenotype

    We compared photoautotrophic growth between the wild-type and SD1 to determine the role of the putative sqdB gene in growth. Although the wild-type cells grew in both the presence and the absence of SQDG, SD1 could grow only with supplementation of SQDG to the culture (Fig. 2). We then performed quantitative measurement of the SQDG-requirement for autotrophic growth of SD1 (Fig. 3). SD1 cells could grow only a little without the addition of exogenous SQDG, probably due to SQDG already present in the precultured cells. The supplementation of 20 µm SQDG to the growth medium was enough for exponential growth of SD1, which was comparable to that of the wild-type (data not shown). Thereafter we supplemented the medium with 20 µm SQDG when culturing SD1 cells. These results indicated that the putative sqdB gene is involved in SQDG synthesis, and is essential for the growth of Synechocystis sp. PCC6803.

    Details are in the caption following the image

    Growth of the wild-type of Synechocystis sp. PCC6803 and its mutant, SD1, on agar plates. Cells grown in the medium containing 20 µm SQDG were streaked onto plates with 20 µm SQDG (+SQDG) or without SQDG (–SQDG), and then the plates were incubated at 30 °C for 3 weeks.

    Details are in the caption following the image

    Effect of the SQDG concentration on the growth of SD1. SD1 cells grown to the late logarithmic phase in the medium containing 20 µm SQDG were transferred to the medium supplemented with 0–100 µm SQDG. •, 0 µm; □, 2 µm; ◊, 5 µm; +, 20 µm; ▵, 50 µm; *, 100 µm SQDG.

    SQDG contents of SD1 cells, unlike those of the wild-type, were decreased after removal of SQDG from the medium (Fig. 4). During incubation without SQDG for 120 h, the SQDG content of SD1 decreased remarkably from 4.09 to 0.61 × 10−17 mol·cell−1, whereas the total lipid content of SD1 increased from 16.8 to 22.8 × 10−17 mol·cell−1. An exact amount of SQDG was found in thylakoid membranes of SD1, which was grown with supplementation of SQDG (Table 1). The trace levels of SQDG detected in both cells and thylakoid membranes of SD1 were probably the carried over from precultured cells. We therefore concluded that SD1 lacks the ability to synthesize SQDG, and that SQDG initially present in precultured cells of SD1 was diluted during cell division after transfer to SQDG-free medium. Thus, it was shown that slr1020 is actually the gene related to SQDG synthesis, probably sqdB, and that SQDG is an essential lipid for Synechocystis sp. PCC6803. On the other hand, little effect of SQDG-depletion on the PG content of the wild-type was found in either the cells or the thylakoid membranes. However, SD1 cells supplemented with SQDG, as compared with the wild-type, showed about four to five times increased PG content in both cells and thylakoid membranes. In addition, removal of SQDG from the medium further increased the PG content of SD1 cells (Table 1) from 4.19 × 10−17 mol·cell−1 to 10.7 × 10−17 mol·cell−1 after a 120-h incubation (Fig. 4). As a result, the total amount of anionic lipid in the cell was maintained at ≈ 45% of the total lipid content (Table 1 and Fig. 4B). The mechanism underlying the control of PG synthesis would directly or indirectly sense the SQDG content.

    Details are in the caption following the image

    Lipid contents of SD1 cells after omission of SQDG from the growth medium. (A) SD1 cells grown to the mid-logarithmic phase in the medium with SQDG were transferred to SQDG-free medium for further growth for lipid analysis at the designated times. The values are means ± SD of three independent measurements in a representative of three independent experiments. (B) Individual lipid compositions were estimated from the respective total lipid contents at the designated times. ▴, SQDG; •, PG; ◆, MGDG; ▪, DGDG; ▵, SQDG + PG.

    Table 1. Lipid composition of the wild-type and SQDG-deficient mutant of Synechocystis sp. PCC6803. WT, wild-type.
    Sample Lipid (mol %)
    ±SQDG MGDG DGDG SQDG PG
    Cells
     PCC6803 WT +a 49 16 28 7
    b 54 13 25 8
     SD1 + 31 22 16 30
    29 25 3 43
    Thylakoid membranes
     PCC6803 WT + 47 15 30 8
    48 15 29 8
     SD1 + 31 30 7 32
    27 22 2 48
    • a Cells were grown in the presence of 20 µ m SQDG. b Cells pre-grown in the presence of SQDG were transferred to SQDG-free medium and grown for 3 days.

    Photosynthetic pigment content of SD1 during a decrease in the SQDG content

    Chl a and phycobiliprotein content were monitored for the wild-type and SD1 after transfer to SQDG-free medium. On a per-cell basis, Chl a content of the wild-type decreased gradually by 15% during a 120-h incubation, whereas for SD1 it increased by 30%(Fig. 5A). The Chl a content on a total lipid basis was maintained at almost the same level during incubation for both SD1 and wild-type cells (Fig. 5B), whereas the phycobiliprotein content on a per-cell basis decreased by 20% for the wild-type cells and increased by 30% for SD1; on a total lipid basis the phycobiliprotein content was constant for both the wild-type and SD1 cells (Fig. 5C and D).

    Details are in the caption following the image

    Changes in pigment contents of SD1 after a shift to SQDG-free medium. Wild-type and SD1 cells pre-grown with SQDG supplementation were transferred to SQDG-free medium for analyses of Chl a (A, on a per-cell basis; B, on total lipid basis) and phycobilins (C, on a per-cell basis; D, on total lipid basis). ○, Wild-type; •, SD1.

    Decreases in the rate of photosynthesis and PSII in relation to a reduction in the SQDG content

    The effect of a decrease in the SQDG content on the rate of net photosynthetic oxygen evolution was monitored for SD1 (Fig. 6). The rate of oxygen evolution in SD1 decreased to 45% of the initial level during incubation of the cells for 120 h under SQDG-free conditions. The reduced photosynthetic activity of SD1, together with the repressed growth (data not shown), fully recovered to the initial level during incubation of the cells in the presence of 20 µm SQDG for a further 96 h. These results indicated that SD1 cells were still alive at 120 h after the removal of SQDG, and that the photosynthetic activity depends on the SQDG content of the cells.

    Details are in the caption following the image

    Changes in photosynthetic activity of SD1 after a shift to SQDG-free medium. Wild-type and SD1 cells pre-grown with SQDG supplementation were transferred to SQDG-free medium for measurement of photosynthetic activity with 10 mm NaHCO3. In the case of SD1, the culture was divided into two after incubation for 120 h. One was incubated successively without SQDG, and the other was supplemented with 20 µm SQDG. The arrow indicates the time of addition of SQDG. The values are means ± SD of three independent measurements in a representative of three independent experiments. ○, wild type; •, SD1.

    Table 2 shows the effects of a decrease in the SQDG content on photosynthetic activities. For the wild-type, supplementation of SQDG to the growth medium had little effect on the photosynthetic activities of the cells. For SD1 grown in the presence of SQDG, although the net photosynthetic activity was high enough, the PSII activity recovered to only 76% of the wild-type level. SD1 grown without SQDG supplementation, in comparison to that with SQDG, showed ≈ 40% decreases in both net photosynthesis and PSII activity. Therefore, these results indicated that the decrease in photosynthesis with a reduction in the SQDG content of SD1 was due to lowered PSII activity.

    Table 2. Net photosynthetic and PSII activities in intact cells of the wild-type (WT) and SQDG-deficient mutants of cyanobacteria. Net photosynthetic activity was measured at 30 °C for cells equivalent to 2.5 µg Chl·ml−1 in the medium containing 10 mm NaHCO3, while PSII activity was measured at 30 °C for those in 50 mm tricine/KOH pH 7.5, 2 mm NH4Cl, and 2 mmp-benzoquinone. Each value represents the mean ± SD for three independent experiments.
    Strain ± SQDG Net photosynthetic activity (µmol O2·mg·Chl−1·h−1) PSII activity (µmol O2·mg·Chl−1·h−1)
    PCC6803 WT +a 404.7 ± 9.8 721.0 ± 11.4
    b 416.7 ± 6.5 753.2 ± 11.9
    SD1 + 413.5 ± 6.4 570.1 ± 9.8
    149.9 ± 3.0 204.5 ± 4.5
    PCC7942 WT 111.2 ± 8.5 905.4 ± 19.3
    SDL 142.3 ± 3.8 902.0 ± 14.6
    • a Cells were grown in the presence of 20 µ m SQDG. b Cells pre-grown in the presence of SQDG were transferred to SQDG-free medium and grown for 3 days.

    Increase in the sensitivity of PSII to herbicides along with a decrease in the SQDG content

    The sensitivity of PSII activity to herbicides such as DCMU and atrazine was compared between the wild-type and SD1 (Fig. 7A and B). These herbicides inhibited the photosynthetic electron transfer from QA to QB through association with the QB-binding site of D1 protein [24,25]. The 50% inhibitory concentrations (IC50) of DCMU and atrazine for PSII activity were 115 nm and 2.4 µm, respectively for wild-type cells supplemented with SQDG and 135 nm and 2.0 µm, respectively for SD1 cells. However, a decrease in the SQDG content in SD1 cells reduced the IC50 for DCMU and atrazine to 45 nm and 1.0 µm, respectively. These results suggest that both inhibitors attack the QB-binding site of PSII more easily in SQDG-decreased thylakoid membranes.

    Details are in the caption following the image

    Effect of a decrease in the SQDG content on the sensitivity of PSII to herbicides in cyanobacteria. PSII activities were measured with 2 mmp-benzoquinone in the presence of various concentrations of DCMU or atrazine. The activities were normalized as to the respective maximal values. The values are means ± SD of three independent measurements in a representative of two independent experiments. In the presence of DCMU (A) or atrazine (B), cells of Synechocystis sp. PCC6803 grown without SQDG (○), and ones of SD1 grown with 20 µm SQDG (▴) or without SQDG (•) for 72 h were used for measurement of PSII activities. Cells of Synechococcus sp. PCC7942 (○) and SDL (•) grown without SQDG-supplementation for 72 h were used for measurement of PSII activities in the presence of DCMU (C) or atrazine (D).

    Effect of a decrease in SQDG content on the amount of PSII complex in SD1

    In contrast to PSII activity, PSII content of wild-type cells was affected little by depletion of SQDG from the culture medium (Table 3). SD1, when supplemented with SQDG, showed a 22% higher content of PSII complex than the wild-type, and the removal of SQDG did not affect the PSII content. Immunoblot analysis also showed that the level of D1 protein on a Chl a basis in SQDG-sufficient and -insufficient SD1 cells was similar to that in the wild-type cells (Fig. 8). Essentially the same results were obtained with the samples of thylakoid membranes isolated from both strains. These results indicated that the amount of PSII complex, little affected by the decrease in SQDG content in thylakoid membranes, is not the cause of the changes in the PSII properties in SD1.

    Table 3. Content of PSII in the wild-type and SQDG-deficient mutant of Synechocystis sp. PCC6803. PSII content was estimated as a half content of Cyt b559[20]. Each value represents the mean ± SD. The numbers of independent experiments are shown in parentheses. For +SQDG, cells were grown in the presence of 20 µm SQDG. For – SQDG, cells pre-grown in the presence of SQDG were transferred to SQDG-free medium and grown for 3 days.
    Strain PSII content (mmol·mol Chl a−1)
    +SQDGa – SQDGb
    PCC6803 wild-type 2.3 ± 0.1 (3) 2.1 ± 0.6 (5)‡
    SD1 2.8 ± 0.2 (4) 2.6 ± 0.5 (5)
    Details are in the caption following the image

    Western blotting of D1 protein of the wild-type (WT) and SD1. The cells of Synechocystis sp. PCC6803 was grown in the absence of SQDG, while SD1, pre-grown with SQDG-supplementation, was transferred to the medium containing 20 µm SQDG (+) or no SQDG (–) for incubation for 72 h. Whole cell proteins and thylakoid membrane samples were prepared for Western analysis as described under Materials and methods. Samples equivalent to 0.1 µg Chl a were applied to the lanes, respectively. The signals were not saturated.

    Photosynthetic properties of an SQDG-deficient mutant of Synechococcus sp. PCC7942

    We observed that an SQDG decrease in Synechocystis sp. PCC6803 caused a reduction in PSII activity, as previously reported for an SQDG-deficient mutant of C. reinhardtii[9,12]. In contrast, SQDG deficiency was reported not to impair photosynthesis in R. sphaeroides[13] or Synechococcus sp. PCC7942 [14]. As the effect of SQDG deficiency on PSII activity had not been shown for Synechococcus sp. PCC7942, we produced an SQDG-deficient mutant of Synechococcus sp. PCC7942 through disruption of the sqdB gene to check its PSII property in detail. The mutant, designated SD1, did not contain any detectable SQDG (< 0.1%, if any, in comparison to the wild-type level, data not shown) and had a growth rate similar to that of the wild-type (data not shown). SD1 had almost the same levels of PSII activity and net photosynthesis as the wild-type cells (Table 2). Interestingly, SD1 was also little affected in its sensitivity of PSII to DCMU and atrazine (Figs 7C and D). These results indicated that Synechococcus sp. PCC7942, in contrast with Synechocystis sp. PCC6803, does not require SQDG for its growth, the maintenance of PSII activity or its sensitivity to the herbicides. We therefore conclude that the effect of an SQDG decrease in thylakoid membranes on PSII properties differs between Synechocystis sp. PCC6803 and Synechococcus sp. PCC7942.

    Discussion

    SQDG is known to be synthesized via two steps; the synthesis of UDP-sulfoquinovose from UDP-glucose and sulfur donor, and the transfer of sulfoquinovose from UDP-sulfoquinovose to diacylglycerol. The formation of the UDP-sulfoquinovose is proposed to be catalysed by the bacterial SqdB proteins or the orthologous plant SQD1 proteins [26]. An ORF, slr1020, in the genome of Synechocystis sp. PCC6803 was identified as the sqdB gene, because of the deduced amino acid sequence of slr1020 exhibiting significant homology to those of known SqdB proteins and the inability of the slr1020 disruptant to synthesize SQDG (SD1, Fig. 4). Thus, we believe that the sqdB gene product in Synechocystis sp. PCC6803 catalyses the formation of the UDP-sulfoquinovose.

    We showed that Synechocystis sp. PCC6803 requires SQDG for its growth. This result is in agreement with previous observations of Güler et al. [27] that the gene disruption on SQDG synthesis could not be conducted in Synechocystis sp. PCC6803 without SQDG-supplementation. A direct cause of the suppressed growth with a lack of SQDG in Synechocystis sp. PCC6803 remains to be clarified. In this study, we succeeded in the isolation of sqdB mutant from Synechocystis sp. PCC6803 through exogenously supplementing SQDG. This strategy was previously used for isolating the mutant deficient in PG synthesis from Synechocystis sp. PCC6803 [28]. In Escherichia coli, the barrier posed by the outer membrane prevents efficient uptake of glycerolipids from the growth medium [29]. In contrast, Synechocystis incorporated exogenously supplemented SQDG not only into the cell but also into the thylakoid membranes (up to ≈ 60% and 20% of the wild-type levels, respectively, Table 1), indicating that Synechocystis sp. PCC6803 possesses some mechanisms for lipid incorporation. The amount of SQDG in the mutant cells supplemented with SQDG was relatively higher compared with that of thylakoid membranes. The higher content of SQDG in the cells is presumably due to retention of SQDG in some membranes other than thylakoid membranes, such as plasma membranes, or to nonspecific association of SQDG with the cells.

    After cessation of SQDG supplementation, whereas the contents of both Chl a and phycobilins decreased gradually in wild-type cells − due probably to growth saturation − the contents of both Chl a and phycobilins on a per-cell basis in SD1 increased by 30% (Fig. 5A and C). This effect in SD1 would be caused by an increase in cell size due to the suppression of cell division (data not shown). Besides, the amounts of Cyt b559 and D1 protein (parts of the PSII complex) are little affected by a decrease in the SQDG content (Table 3, Fig. 8). These results suggested that PSII complexes were synthesized continuously even after the suppression of cell division due to a great loss of SQDG.

    SD1, despite possessing a normal amount of PSII complex on a Chl a basis, showed decreased PSII activity with a reduction in the SQDG content (see Table 2). Thus, in Synechocystis sp. PCC6803, the decrease in the SQDG content of the cells reduced the specific activity of PSII. SQDG was found to be bound to the PSII core complex of the wild type of C. reinhardtii[12]. SQDG has also been reported to be present in PSII-related complexes in spinach [6,30]. Therefore, SQDG might be associated with the PSII complex also in Synechocystis sp. PCC6803, maintaining the PSII activity. SD1 showed increased sensitivity of PSII activity to DCMU and atrazine with a decrease in the SQDG content (Fig. 7A and B). Similar observations, i.e. the lowered specific activity of PSII and the enhanced sensitivity to DCMU, were made for SQDG-deficient mutants of C. reinhardtii[9,10,31]. Yu et al. [32] also indicated that the effective quantum yield of PSII was slightly but significantly reduced in SQDG-deficient mutants of Arabidopsis thaliana. Hence, SQDG would maintain, at least, the normal conformation of a region such as the QB-binding site of the D1 protein, and/or the normal state of the lipid environment around the site.

    It had been proposed that SQDG and PG, the other anionic lipid of cyanobacteria and chloroplasts, could be functionally complementary to each other, as SQDG and PG contents have the same inverse correlation depending on available sources of nutrition – keeping a constant anionic lipid content. The sulfolipid-deficient mutants with increased PG content of R. sphaeroides, Synechococcus sp. PCC7942 and A. thaliana showed similar growth rates to those of the wild-type, but were impaired in the growth with decreased PG content under phosphate-limited condition [13,14,32]. In SD1, the content of SQDG was decreased with a complementary increase in the content of PG. However, the increased PG in SD1 could not recover its decreased PSII activity even in normal conditions (Fig. 4). Besides, unlike the effect of SQDG reduction, the decrease in the content of PG in a PG-deficient mutant of Synechocystis sp. PCC6803 caused p-benzoquinone-dependent PSII activity lower than the CO2-dependent photosynthetic activity [28]. Therefore, SQDG and PG would play some distinct role in maintenance of the PSII properties in Synechocystis sp. PCC6803.

    Riekhof et al. [31] suggested that Synechococcus sp. PCC7942 was closely related to alpha-group bacteria rather than to Synechocystis sp. PCC6803, because Synechococcus sp. PCC7942 was positioned in the same cluster in the phylogenetic tree as R. sphaeroides, and apart from other cyanobacteria for known UDP-sulfoquinovose syntheses; also, in contrast with Synechocystis sp. PCC6803, it did not have some fatty acid desaturases or the tocopherol synthesizing ability. We found that there are differences among species in the degree of the contribution of SQDG to photosynthesis. SDL, distinct from SD1 and hf-2, showed no deleterious effect of the SQDG defect on PSII activity or its sensitivity to the herbicides (Figs 7C and D). We consider that the PSII complex of Synechocystis sp. PCC6803 is more evolved than that of Synechococcus sp. PCC7942, because SQDG is not required for the functioning of the photosystem of R. sphaeroides, the prototype of the PSII complex, but it is responsible for the functioning of the PSII complex of a green alga, C. reinhardtii. Thus, the PSII complex may have evolved with use of SQDG as one of the limiting factors. There are some differences among C. reinhardtii, Synechocystis sp. PCC6803, and Synechococcus sp. PCC7942 in the amino acid sequences of the QB-binding site of the D1 protein (comparing amino acid residues 230, 231, 233 and 235). They also possess distinct subunits of the extrinsic oxygen-evolving system of PSII [33]. The SQDG dependency of the PSII properties might change during the evolution of cyanobacteria through nucleotide base substitution of the genes for the PSII complex.

    Both the molecular mechanism underlying the interaction of SQDG with PSII complex and the evolution of PSII biased by the interaction of SQDG remain to be clarified. As we found different types of interaction with SQDG between the cyanobacteria strains, Synechocystis sp. PCC6803 and Synechococcus sp. PCC7942, suitable for analysis by molecular biological techniques, it is now possible to investigate the mechanism more precisely at the molecular level.

    Acknowledgements

    This work was supported by grants from the Ministry of Education, Science, Sports and Culture (13740463, 13874112, and 12206002), the Promotion and Mutual Aid Corporation for Private Schools of Japan, and the Sasagawa Scientific Research Grant from The Japan Science Society to M.A.

    We thank M. Ikeuchi, The University of Tokyo, for the gift of the D1 antibodies.