Photosynth Res DOI 10.1007/s11120-015-0143-8

REGULAR PAPER

Analysis of spontaneous suppressor mutants from the photomixotrophically grown pmgA-disrupted mutant in the cyanobacterium Synechocystis sp. PCC 6803 Yoshiki Nishijima1 • Yu Kanesaki2 • Hirofumi Yoshikawa2,3,6 • Takako Ogawa4,5 Kintake Sonoike4 • Yoshitaka Nishiyama1 • Yukako Hihara1,6

•

Received: 4 February 2015 / Accepted: 5 April 2015 Ó Springer Science+Business Media Dordrecht 2015

Abstract The pmgA-disrupted (DpmgA) mutant in the cyanobacterium Synechocystis sp. PCC 6803 suffers severe growth inhibition under photomixotrophic conditions. In order to elucidate the key factors enabling the cells to grow under photomixotrophic conditions, we isolated spontaneous suppressor mutants from the DpmgA mutant derived from a single colony. When the DpmgA mutant was spread on a BG11 agar plate supplemented with glucose, colonies of suppressor mutants appeared after the bleaching of the background cells. We identified the mutation site of these suppressor mutants and found that 11 mutants out of 13 had a mutation in genes related to the type 1 NAD(P)H dehydrogenase (NDH-1) complex. Among them, eight mutants had mutations within the ndhF3 (sll1732) gene: R32stop, W62stop, V147I, G266V, G354W, G586C, and deletion of 7 bp within the coding region. One mutant had one base insertion in the putative -10 box of the ndhC (slr1279) gene, leading to the decrease in the transcripts of the ndhCKJ operon. Two mutants had one base insertion and & Kintake Sonoike [email protected]

deletion in the coding region of cupA (sll1734), which is co-transcribed with ndhF3 and ndhD3 and comprises together a form of NDH-1 complex (NDH-1MS complex) involved in inducible high-affinity CO2 uptake. The results indicate that the loss of the activity of this complex effectively rescues the DpmgA mutant under photomixotrophic condition with 1 % CO2. However, little difference among WT and mutants was observed in the activities ascribed to the NDH-1MS complex, i.e., CO2 uptake and cyclic electron transport. This may suggest that the NDH-1MS complex has the third, currently unknown function under photomixotrophic conditions. Keywords Cyanobacteria  Glucose tolerance  NAD(P)H dehydrogenase  ndhF3  Photomixotrophic growth  Suppressor mutation Abbreviations GT HL NDH-1 NDH-1MS complex OD SNP WT

Glucose tolerant High light Type 1 NAD(P)H dehydrogenase A form of NDH-1 complex involved in inducible high-affinity CO2 uptake Optical density Single-nucleotide polymorphism Wild type

1

Graduate School of Science and Engineering, Saitama University, Saitama 338-8570, Japan

2

Nodai Genome Research Center, Tokyo University of Agriculture, Tokyo 156-8502, Japan

3

Department of Bioscience, Tokyo University of Agriculture, Tokyo 156-8502, Japan

4

Faculty of Education and Integrated Arts and Sciences, Waseda University, Tokyo 162-8480, Japan

Introduction

5

Japan Society for the Promotion of Science, Tokyo 102-0083, Japan

6

CREST, Japan Science and Technology Agency (JST), Saitama 332-0012, Japan

Cyanobacteria are primarily photoautotrophic organisms performing oxygenic photosynthesis to convert light energy to chemical energy for their growth and propagation. Some cyanobacterial species additionally have the ability

123

Photosynth Res

to gain chemical energy heterotrophically by catabolizing exogenously added glucose via the glycolytic pathway, the oxidative pentose phosphate (OPP) pathway, and the tricarboxylic acid (TCA) cycle. Regulation of the activities of anabolic and catabolic reactions seems to be essential for the growth of these cyanobacterial species under the photomixotrophic conditions where both light and glucose are available. Although Synechocystis sp. PCC 6803 possesses glucose transporter and has capability of heterotrophic growth (Zhang et al. 1989), the original wild-type (WT) strain (PCC strain) is sensitive to photomixotrophic conditions probably due to the metabolic imbalance (Williams 1988). The glucose-tolerant (GT) non-motile strain was isolated by Williams (1988) from the original PCC strain, and it has been used as the model cyanobacterial strain for researches in molecular biology. The Kazusa strain, whose genomic sequence was determined in 1996 (Kaneko et al. 1996), is a derivative of the GT strain. Kanesaki et al. performed resequencing of the genomes of a GT substrain (GT-I) and two PCC substrains (PCC-P and PCC-N) and compared the sequence with those of the original Kazusa strain. They reported that 14 specific mutations exist between the GT substrains (GT-I and Kazusa) and the PCC substrains (PCC-P and PCC-N) (Kanesaki et al. 2012). The loci responsible for the acquisition of glucose tolerance of the GT strain must exist among these 14 mutations. The glucose-tolerant phenotype of the GT strain is not stable, and it has been often reported that the GT strain became no longer glucose tolerant during laboratory maintenance probably due to spontaneous mutations. One good example is the ‘‘microevolution’’ of the GT strain reported by Hihara and Ikeuchi (1997). During the routine laboratory maintenance of the GT strain under photoautotrophic conditions, a mutant having base substitution of T for C at the position 193 in the pmgA (sll1968) gene appeared spontaneously. The mutation in the gene caused the replacement of Leu with Phe at position 65 of the product, and the mutant showed increased photosynthetic activity as well as enhanced growth rate under photoautotrophic conditions (Hihara et al. 1998). The pmgA mutant grew better than WT and completely took over the culture for a year or so. The pmgA gene encodes a soluble protein of 23 kDa having N, G1, and G2 boxes required for nucleotide binding (Hihara and Ikeuchi 1997). Histidine kinases of the two-component system and anti-sigma factors such as RsbW and SpoIIAB in Bacillus subtilis are known to have these motifs (Min et al. 1993). However, PmgA does not possess the catalytic histidine residue required for histidine kinase, and there is no evidence of interaction of PmgA with a sigma factor to regulate gene expression. Although both the pmgA L65W mutant and the pmgAdisrupted (DpmgA) mutant can grow better than WT under normal laboratory conditions, they suffered severe growth

123

inhibition under continuous high-light (HL) or photomixotrophic conditions. Under HL conditions, the mutants could not maintain photosystem (PS) I content at low level, which caused higher photosynthetic electron transport activity and better growth compared to WT at the beginning of the growth, but led to severe inhibition at a later growth stage (i.e., after 48 h following the inoculation), presumably due to the over-reduction of the acceptor side of PSI (Hihara et al. 1998; Sonoike et al. 2001; Muramatsu and Hihara 2012). Several other mutants defective in the suppression of PSI content under HL were also shown to be glucose sensitive (Fujimori et al. 2005; Ozaki et al. 2007), suggesting the relationship between HL acclimation and glucose tolerance. It should be noted, however, that the large repression of psaAB transcript and PsaAB polypeptide levels observed in the WT cells under HL conditions is not observed under photomixotrophic conditions. Thus, the excess accumulation of PSI complex could not be the cause of growth inhibition of these mutants under photomixotrophic conditions. Elucidation of the cause of growth inhibition of the pmgA mutants seems to be a good approach to clarify the regulatory mechanism required for photomixotrophic growth of the GT strain. Thus, we identified the mutation sites in the genomic DNA of spontaneous suppressor mutants isolated from the DpmgA mutant (Haimovich-Dayan et al. 2011). When the DpmgA mutant was spread on BG11 agar plates supplemented with 5 mM glucose, it stopped growing after 2 days and began to bleach. On the bleached background, however, green colonies of spontaneous suppressor mutants appeared after 5 days. We picked up ten suppressor mutant strains, made libraries from their genomic DNA, and screened genomic clones having ability to complement the light/glucose sensitivity of the DpmgA mutant. As a result, we found that, out of ten suppressor mutant strains, eight strains had the same mutation in the ndhF3 (sll1732) gene encoding a subunit of type 1 NAD(P)H dehydrogenase (NDH-1) complex. There was a base substitution in ndhF3 at position of 1060 that caused the replacement of highly conserved glycine residue at position 354 with tryptophan. The repeated identification of G354W mutation in the ndhF3 gene as mutation sites of spontaneous suppressor mutants suggested that this mutation had been spread throughout the DpmgA mutant strain during cultivation. Based on the results, we planned to identify the mutation sites of the spontaneous suppressor mutants from the DpmgA mutant culture derived from the single colony in the present study. Using the genetically identical culture as a background, isolation of suppressor mutants having various mutations in different genes would be expected. Moreover, instead of the time-consuming screening of genomic DNA library, we employed the whole-genome sequencing by the next-generation sequencer to identify the mutation sites. As expected, we could identify various mutation sites.

Photosynth Res

Surprisingly, however, we found that 11 mutants out of 13 have a suppressor mutation in the genes related to NDH-1 complex, suggesting the importance of this complex in the capability to grow under photomixotrophic conditions.

Materials and methods Strains and culture conditions A glucose-tolerant WT strain of Synechocystis sp. PCC 6803 (GT strain) and mutants were grown at 32 °C in BG-11 medium (Stanier et al. 1971) containing 20 mM HEPES– NaOH, pH 7.0, under continuous illumination at 50 lmol m-2 s-1, bubbled with air containing 1 % (v/v) CO2. Solid medium was buffered with 5 mM TES-KOH, pH 8.2, supplemented with 1.5 % (w/v) agar and 0.3 % (w/v) sodium thiosulfate. To support photomixotrophic growth, glucose was added at a concentration of 5 mM. Cell density in liquid culture was estimated by measuring optical density at 730 nm (OD730) using a spectrophotometer (model UV160A; Shimadzu, Kyoto, Japan). For low-CO2 shift experiment, cells grown under 1 % CO2 conditions were harvested, washed twice with BG-11 medium without Na2CO3, and then cultured in BG-11 medium without Na2CO3, bubbled with air. The single- and double-gene-disrupted mutants of pmgA, ndhF3, ndhD3, and cupA were made by inactivation of genes in the background of the GT strain by insertion of an antibiotic cassette (spectinomycin cassette for pmgA, erythromycin cassette for ndhD3, and kanamycin cassette for ndhF3 and cupA). To maintain gene-disrupted mutants, appropriate antibiotics were added at the final concentration of 20 lg ml-1.

300 bp for a paired-end read format. The library was sequenced on Genome Analyzer IIx or Hiseq 2000 (Illumina Inc., San Diego, CA, USA). Sample preparation, cluster generation, and 100-base paired-end sequencing were performed according to the manufacturer’s protocols with minor modifications (Illumina paired-end cluster generation kit GAII ver. 2, 36-cycle sequencing kit ver. 3). Image analysis and ELAND alignment were performed with Illumina’s Pipeline Analysis software ver. 1.6. Sequences passing standard Illumina GA pipeline filters were retained. For shortread alignment and consensus assembly, we used a recently developed algorithm, BWA ver. 0.5.1 (the outline of the algorithm is available online http://bio-bwa.sourceforge.net/) (Li and Durbin 2010) and MAQ (Li et al. 2008). To call SNPs, we used SAMtools software ver. 0.1.9 (Li et al. 2009) and applied additional filters as follows: minimum read depth for the indel calling = 10, minimum read depth for the SNP calling = 5, and an 80 % cutoff of percent aligned reads calling the SNP per total mapped reads at the SNP sites. We also used BWA to estimate the sequence read depth, which influences the coverage and accuracy of SNP calling. The lists of SNPs/indels were then annotated using the in-house software COVA (comparison of variants and functional annotation) (http://sourceforge.net/projects/cova) (Shiwa et al. 2012). We used the GenBank, RefSeq, and cyanobacterial database Cyanobase (http://genome.kazusa.or.jp/cyanobase) for the annotation. Original read data and data information are deposited in the DRA (DDBJ sequence read archive; http:// trace.ddbj.nig.ac.jp/DRASearch/) and in the BioProject with the accession number of PRJDB3465. Sanger sequencing was performed using ABI3130 Genetic Analyzer (Thermo Fisher Scientific Inc., Waltham, MA, USA). Complementation of the DpmgA mutant under photomixotrophic conditions

Isolation of genomic DNA To isolate genomic DNA, Synechocystis cells were washed with 5 mM EDTA and then incubated with saturated NaI at 37 °C for 30 min. After dilution of the suspension, cells were pelleted and resuspended in 50 mM Tris–HCl, pH 8.0, and 20 mM EDTA. Cells were first incubated with 100 lg ml-1 ribonuclease and 4 mg ml-1 lysozyme at 37 °C for 45 min and then lysed with 0.5 % (w/v) SDS and 200 lg ml-1 proteinase K at 55 °C overnight. DNA was extracted several times with phenol–chloroform and precipitated with ethanol. Resequence analysis using massively parallel sequencer Genomic DNA was uniformly sheared to about 300 bp using a Covaris S-2 sonicator (Covaris, Inc., Woburn, MA, USA). We constructed a DNA library with a median insert size of

One milliliter culture of the DpmgA mutant grown to midlog phase was spread on a BG11 plate supplemented with 5 mM glucose. After drying the surface of the plate, the solution of DNA for complementation was directly applied onto it. The plate was wrapped with Parafilm and placed under the light at 50 lmol m-2 s-1. Transformants capable of growing under photomixotrophic conditions were detected in 5–7 days. RNA gel blot analysis Isolation of total RNA by the hot-phenol method and RNA gel blot analyses, using a DIG RNA Labeling and Detection Kit (Roche, Basel, Switzerland), were performed as described previously (Muramatsu and Hihara 2003). Template DNA fragments were prepared by polymerase chain reaction (PCR) using the following primers to

123

Photosynth Res

generate RNA probes by in vitro transcription: ndhC-F (50 -GTGTTTGTTTTAACCGGT-30 ) and T7-ndhC-R (50 -T AATACGACTCACTATAGGGCGACTAGGACCACTC CAGAGC-30 ). The underlining indicates the T7 polymerase recognition sequence added to the reverse primer in order to use the PCR products directly as templates for in vitro transcription reactions. CO2 exchange measurement Fifty milliliters of cultures of WT and mutant strains at the mid-exponential growth phase were harvested by filtration. Cells on the filter were resuspended with 20 mM Hepes– NaOH, pH 7.0, containing 15 mM NaCl, and placed in a reaction vessel kept at 30 °C. Cell density of the cell suspension was estimated by measuring OD730 and used for the calculation of the rate of CO2 uptake (lmol -1 OD-1 730 h ). The CO2 exchange of the cell suspension in the reaction vessel was measured with an open infrared gas analysis system that records the rate of CO2 exchange as a function of time (Ogawa and Kaplan 1987). Mixed gas consisting of 80 % N2, 20 % O2, and 0.04 % CO2 was generated using a standard gas generator unit (SGGU-712, Standard Technology, Kyoto, Japan) and used to aerate the cell suspension in the reaction vessel through a fritted glass port at a flow rate of one l min-1. The gas leaving the chamber was dried by being passed through a spiral condenser, kept at 3 °C, and then led to an infrared CO2 analyzer (model URA-106, Shimadzu, Kyoto, Japan). The CO2 concentration in the medium surrounding the cells was calculated from the concentration of CO2 in the gas passing through the medium, assuming the equilibrium between them. Cells started to uptake CO2 upon illumination at 200 lmol m-2 s-1 from a slide projector, and the maximum rate was employed as -1 the rate of CO2 uptake (lmol OD-1 730 h ). P700 measurements OD730 of cells determined by a spectrophotometer (UV1800, Shimadzu, Kyoto, Japan) was set to 0.02 at the start of the growth in liquid culture. After the growth for 24 h, cells are concentrated by centrifugation and OD730 was adjusted to 2.0 before the P700 measurements. Oxidation reduction kinetics of P700 was determined by a pulse amplitude modulation photometer (Dual PAM 100, Walz, Effeltrich, Germany). Measuring light (sample wavelength 830 nm, reference wavelength 870 nm) was applied with intensity setting of ten with low gain (setting 1) and high damping (1 ms). Actinic light (5 s) from far-red LED (720 nm) was applied for three times with intervals of 15 s. The three measurements with two independent cultures, i.e., six measurements in total, were averaged for presentation.

123

Results When the DpmgA mutant derived from a single colony was spread on a BG11 agar plate supplemented with 5 mM glucose, green colonies of spontaneous suppressor mutants appeared after the bleaching of the background cells. We picked up ten small colonies (strain S1–S10) using an optical microscope, together with nine middle-sized colonies (strain M11–M19) and 12 large colonies (L20–L31). Thirty one suppressor mutants in total were grown in liquid BG11 medium and harvested for isolation of the genomic DNA as well as for making stock culture without further inoculation to avoid the occurrence of the further mutational events. After the confirmation that the genomic DNA isolated from each suppressor mutant enabled the DpmgA mutant to grow on the BG11 plate with glucose, the nucleotide sequence of the ndhF3 gene was examined by the PCR amplification and Sanger sequencing, since G354W mutation in the ndhF3 gene was repeatedly isolated in the previous study (Haimovich-Dayan et al. 2011). In cases where no mutation was identified within the ndhF3 gene, genomic DNA was sheared by ultrasonic treatment and sequenced by a genome sequencer. The obtained genomic sequences of the suppressor mutants were compared with those of the parental DpmgA strain, and the putative mutation sites of each mutant were checked by PCR amplification and Sanger sequencing. As a result, we identified the mutation sites of ten strains out of L20–L31 and three strains out of M11–M19, but no mutation was detected in S1–S10 strains (Table 1). It should be noted that these may not be the sole mutation sites in each suppressor mutant. Surprisingly, 11 mutants out of 13 had a suppressor mutation in genes related to the NDH-1 complex (Table 1). Among them, eight mutants had mutations within the ndhF3 gene: R32stop (L28), W62stop (L20, L21), V147I (L29), G266V (M18), G354W (L30), G586C (L23), and deletion of 7 bp within the coding region (L24). It is notable that the mutation sites were dispersed throughout the amino acid sequence of the ndhF3 gene unlike the previous result repeatedly identifying single G354W mutation (Haimovich-Dayan et al. 2011). L22 had one base insertion in the putative -10 box of the ndhC (slr1279) gene. The change of the nucleotide sequence from TATAAT to TAATAAT could lead to the decreased activity of transcription. To examine the possibility, we examined the transcript levels of ndhC in WT, the DpmgA mutant, and L22 mutant by RNA gel blot analysis (Fig. 1). Three bands of about 1800, 1400, and 900 nt long were detected in WT and the DpmgA mutant using ndhC probe. These bands that were considered to correspond to the ndhCKJ, ndhCK, and ndhC transcripts, respectively, were hardly observed in L22

Photosynth Res Table 1 Mutation sites in the suppressor mutants

Strain

Position

Gene

Mutation (resulting amino acid substitution)

M16

134,748–135,187

infB

Deletion of 440 bp in the coding region

2,389,715–2,389,632

sll0327

Deletion of 84 bp in the coding region

M18

939,444

ndhF3

Substitution from C to A (G266V)

M19

134,748–135,187

infB

Deletion of 440 bp in the coding region

L20

940,055

ndhF3

Substitution from G to A (W62stop)

L21

940,055

ndhF3

Substitution from G to A (W62stop)

L22

1,874,989–1,874,990

ndhC

Insertion of A to the putative -10 box

L23

938,485

ndhF3

Substitution from G to T (G586C)

L24

938,832–938,826

ndhF3

Deletion of 7 bp in the coding region

L25

936,074

cupA

Deletion of 1 bp (T) in the coding region

L26

936,672

cupA

Insertion of 1 bp (C) in the coding region

L28

940,147

ndhF3

Substitution from C to T (R32stop)

L29

939,802

ndhF3

Substitution from G to A (V147I)

L30

939,181

ndhF3

Substitution from G to T (G354W)

mutant, indicating the low transcription activity of the ndhCKJ operon due to the mutation in -10 box. L25 and L26 had one base deletion and insertion in the coding region of cupA (sll1734), respectively. cupA is known to be co-transcribed with ndhF3 (sll1732) and ndhD3 (sll1733), and the products of these genes together comprise an inducible high-affinity CO2 uptake system, the NDH-1MS complex (Shibata et al. 2001; Herranen et al. 2004). M16 and M19 were the only mutants having mutation sites not

Fig. 1 Transcript levels of the ndhCKJ operon in WT, the DpmgA mutant, and the suppressor mutant L22 (DpmgA/one base insertion to the -10 box of ndhC) examined by RNA gel blot analysis. 5 lg of total RNA was loaded per lane. Total RNA was stained with methylene blue to show the equal loading of RNA

related to the NDH-1 complex. M16 has deletion of 440 bp within infB (slr0744) encoding a translation initiation factor and 84 bp deletion within sll0327 encoding a functionunknown protein. In the M19 genome, the same deletion in infB as M16 was found. The rescue effect of deletion of these genes could not be confirmed, since the double mutants of pmgA and these genes could not be obtained. Previously, we observed that the growth inhibition of the photomixotrophically grown DpmgA mutant was severer under 1 % CO2 conditions than under air-level CO2 conditions (0.04 %) (Haimovich-Dayan et al. 2011). The observation together with the fact that the mutation sites of suppressor mutants concentrated on the genes related to the NDH-1MS complex indicates that the regulation of CO2 uptake activity may be critical for the survival of the DpmgA mutant under photomixotrophic conditions. To examine the rescuing effect of the mutation in genes encoding subunits of the NDH-1MS complex on the photomixotrophic growth of the DpmgA mutant in more detail, growth properties of WT, the DpmgA mutant, the double mutants of DpmgA/DndhF3, DpmgA/DndhD3, and DpmgA/ DcupA, and suppressor mutants of L22 (DpmgA/one base insertion to the -10 box of ndhC), L23 (DpmgA/NdhF3 G586C), and L30(DpmgA/NdhF3 G354W) were examined in liquid BG11 medium supplemented with 5 mM glucose under 1 % CO2 atmosphere (Fig. 2). When the cultures were inoculated every 24 h to minimize the self-shading effect, the delay of growth in the DpmgA mutant became prominent after 24 h, whereas WT did not suffer any growth inhibition. The other mutant strains showed the intermediate phenotype between WT and the DpmgA mutant. After 48 h, the growth of the DpmgA mutant was completely inhibited. The disruption of either ndhF3, ndhD3, or cupA gene could partially suppress the growth inhibition, whereas suppressor mutations that can mitigate

123

Photosynth Res

the growth inhibition on agar plates hardly improved the growth of the DpmgA mutant in liquid. This result is inconsistent with the data shown in Haimovich-Dayan et al. (2011) where the growth of the DpmgA/ndhF3 G354W mutant in liquid culture was far better than that of the DpmgA/DndhF3 mutant after 48 h of growth under photomixotrophic conditions. We suppose that this difference originates from the difference in the genetic background of the mutants. All of the double mutants and suppressor mutants used in this study were obtained from the same DpmgA mutant derived from a single colony, whereas the mutants used in Haimovich-Dayan et al. (2011) were originated from the WT strain comprising heterogonous cell population. Thus, it is expected that the growth property presented in this study reveals more specific effect of mutations in the ndh genes on photomixotrophic growth of the DpmgA mutant. The fact that mutations in genes encoding subunits of the NDH-1MS complex can relieve the growth inhibition of the DpmgA mutant indicates that the active uptake of both CO2 and glucose as a carbon source may be harmful for the mutant. In this context, information on the expression level of the NDH-1MS complex with or without glucose in WT and the DpmgA mutant seems important. Haimovich-Dayan et al. (2011) reported that the ndhF3 transcript accumulated significantly following the addition of glucose in WT but not in the DpmgA mutant under 5 % CO2 conditions. As for protein level, there is a proteomic study showing that the membrane protein pattern of WT did not reveal any significant changes by the addition of glucose under ambient CO2 conditions (Herranen et al.

Fig. 2 The growth properties of WT (open circles), the DpmgA mutant (closed circles), the double mutants of DpmgA/DndhF3 (open triangles), DpmgA/DndhD3 (closed triangles), and DpmgA/DcupA (double triangles), and suppressor mutants of L22 (DpmgA/one base insertion to the -10 box of ndhC) (closed squares), L23 (DpmgA/ NdhF3 G586C) (open squares), and L30 (DpmgA/NdhF3 G354W) (double squares) under photomixotrophic conditions with 5 mM glucose. The cultures were under continuous illumination at 50 lmol m-2 s-1 with bubbling of 1 % (v/v) CO2 and were inoculated every 24 h to minimize the self-shading effect

123

2004). In order to know the changes in the amount of the NDH-1MS complex by the addition of glucose under 1 % CO2 conditions, we performed immunoblot analysis using antibodies raised against NdhF3, NdhD3, or CupA. However, we failed to detect any specific band not only from the cell lysate but also from the membrane fraction of WT and mutant strains (not shown). We assume that low-CO2inducible NDH-1MS complex is active under 1 % CO2 conditions in our experimental system, judging from the large effect of the mutations in NdhF3 and CupA on the photomixotrophic growth of the DpmgA mutant. It would be possible that the NDH-1MS activity may decrease in WT upon the addition of glucose under 1 % CO2 conditions, whereas it may stay at high level in the DpmgA mutant. To examine if such a difference in the activity of CO2 uptake was observed between strains, we measured CO2 uptake activity of the strains examined in the experiments shown in Fig. 3 using an open gas analysis system (Ogawa and Kaplan 1987).

Fig. 3 CO2 uptake activity of WT, the DpmgA mutant, the double mutants of DpmgA/DndhF3, DpmgA/DndhD3, and DpmgA/DcupA, and suppressor mutants of L22 (DpmgA/one base insertion to the -10 box of ndhC), L23 (DpmgA/NdhF3 G586C), and L30 (DpmgA/NdhF3 -1 G354W). a CO2 uptake activity (lmol OD-1 730 h ) before and 24 h after the shift from 1 % CO2 to air. b CO2 uptake activity before and 24 h after the addition of 5 mM glucose under 1 % CO2. The gas exchange of the cells was measured under 400 ppm CO2 concentration with open gas analysis system (an infrared CO2 analyzer)

Photosynth Res Fig. 4 Kinetics of absorption changes due to P700 oxidation/ reduction upon far-red light illumination. a, d the WT cells, b, e the DpmgA mutant cells, c, f the DpmgA/DndhF3 mutant cells. a–c photoautotrophically grown cells, d– f photomixotrophically grown cells in the presence of 5 mM glucose. The kinetics were determined either without the addition (black bold lines), with the addition of 1 mM KCN (black thin lines), or with the addition of 1 mM KCN and 0.1 mM methyl viologen (and gray bold lines). Data are the average of six measurements with two independent cultures

Figure 3a shows the comparison of the CO2 uptake activity of cells before and 24 h after the transfer from 1 % CO2 to ambient CO2 under photoautotrophic conditions. The activity was enhanced after 24 h of incubation under ambient CO2 conditions both in the WT and the DpmgA mutant. This indicates that the induction of CO2 uptake activity upon the downshift of CO2 concentration is operated normally in the photoautotrophically grown DpmgA mutant. In contrast, no induction of the CO2 uptake activity was observed in the DpmgA/DndhF3 mutant presumably due to the absence of low-CO2-inducible NDH-1MS complex in this mutant. The suppressor mutants having a point mutation in ndhF3, L23 (DpmgA/NdhF3 G586C) and L30 (DpmgA/NdhF3 G354W), also did not show the increase of the CO2 uptake activity. In L22 mutant with low transcription activity of ndhCKJ operon, the CO2 uptake activity was lower than that in other strains irrespective of CO2 conditions. The amount of not only NDH-1MS but also other types of NDH complex must be low in this mutant. Next, we examined the change in the CO2 uptake activity after the addition of 5 mM glucose to cultures grown under 1 % CO2 conditions (Fig. 3b). The CO2 uptake activity derived from NDH-1MS complex seems negligible in the absence of glucose (at 0 h) judging from the similar activity between WT and the DpmgA/DndhF3 mutant not having NDH-1MS complex. Nevertheless, decrease in the CO2 uptake activity was observed in every strain after 24 h of incubation under photomixotrophic conditions. The extent of decrease in the CO2 uptake activity was similar between WT and the DpmgA mutant.

Although the growth inhibition of the DpmgA mutant under photomixotrophic condition was rescued by mutations in NDH-1MS complex involved in CO2 uptake, the regulation of CO2 uptake activity upon the change in CO2 concentration or trophic condition seemed normal in the mutant. Thus, we assumed that NDH-1MS may be involved in a certain cellular process other than CO2 uptake, and the process may be critical for the survival under photomixotrophic conditions. It was proposed that NDH-1MS is involved in cyclic electron transport around PSI (Berna´t et al. 2011), besides the well-known role in CO2 uptake. We determined the redox state of P700, the reaction center chlorophyll of PSI, to monitor the possible involvement of cyclic electron transfer in the mechanism of the interaction between the defect in pmgA and that in ndhF3. Far-red illumination, which preferentially excites PSI, caused rapid oxidation of P700 to P700? in photoautotrophically grown WT cells, resulting in the absorbance increase (Fig. 4a, bold black line). This rapid increase is largely suppressed by the addition of KCN, an inhibitor of terminal oxidase in respiratory chain (Fig. 4a, thin black line). Since photosynthesis and respiration shares plastoquinone pool and cytochrome b6/f complex in cyanobacteria (Aoki and Katoh 1982; Peschek and Schmetterer 1982), inactivation of the terminal oxidase should lead to the reduction of redox components between PSI and PSII. This reducing pressure on PSI in the presence of KCN should compete with the oxidation brought about by far-red light, resulting in the slow rising kinetics observed in Fig. 4a. The slow down of the rate of P700 oxidation by KCN was partially restored upon the

123

Photosynth Res

addition of methyl viologen, an electron acceptor from PSI (Fig. 4a, gray line). The observed acceleration of the P700 oxidation rate should reflect the interruption of cyclic electron transfer around PSI by methyl viologen (Berna´t et al. 2011). Thus, we could estimate the changes in cyclic electron transfer by determining the effect of methyl viologen on the redox kinetics of P700. The DpmgA mutant showed somewhat slower oxidation of P700 upon far-red illumination (Fig. 4b, bold black line) compared with the WT strain, but the effect of KCN and methyl viologen on the redox kinetics of P700 is basically the same as that of the WT (Fig. 4b, thin black line and gray line). This is also true for the DpmgA/DndhF3 mutant (Fig. 4c). In this case, the effect of methyl viologen seems to be smaller than that in the single DpmgA mutant, consistent with the earlier report suggesting the involvement of the ndhF3 gene in cyclic electron transfer (Berna´t et al. 2011). When P700 kinetics were determined with photomixotrophically grown cells, extent of the maximum absorption changes due to P700 oxidation decreased to about one-third of that of the photoautotrophically grown cells (Fig. 4d–f, compare with Fig. 4a–c). Glucose, as a substrate of respiration, could be an electron donor to photosynthetic electron transfer in cyanobacteria, acting as a reducing pressure on P700. The decreased extent of absorbance change due to P700 oxidation could be partly explained by the electron flow from respiratory electron transfer to plastoquinone pool (see discussion). The effect of KCN or of KCN/methyl viologen, however, was similar to that observed in photoautotrophic cells, namely slower rise in the presence of KCN and partial recovery by the further addition of methyl viologen (Fig. 4d). These effects were also observed in the DpmgA mutant (Fig. 4e) or in the DpmgA/DndhF3 mutant (Fig. 4f). Since the DpmgA mutant and the DpmgA/DndhF3 mutant did not show much difference in the redox kinetics of P700, the effect of the disruption in ndhF3 gene on the growth of the DpmgA mutant under photomixotrophic condition could not be explained by the change in cyclic electron transfer.

Discussion Previously, we repeatedly identified G354W mutation in the ndhF3 gene as mutation sites of suppressor mutants from the DpmgA mutant cultivated for a long time (Haimovich-Dayan et al. 2011). Since the identified mutation was solely limited to G354W, we could not distinguish whether the mutation in the ndhF3 gene is really important or the mutation had already spread within the culture. In order to distinguish those two possibilities, here we used the single-colony-derived DpmgA mutant in this study and

123

isolated 31 suppressor mutant strains. No mutation site could be identified from the suppressor ‘‘S’’ strains originated from small colonies on the plate with glucose. In this study, an 80 % cutoff value was employed to identify SNP. This means that the heterogeneous mutation cannot be detected if less than 80 % of genomic copies have the mutation. In ‘‘S’’ strains, it is likely that only a small fraction of genomic DNA possesses the suppressor mutation. This may cause the slow growth on the plate with glucose and failure in the identification of the mutation sites. In the case of ‘‘M’’ and ‘‘L’’ strains originated from the middle- and large-size colonies, 11 out of 13 mutants had mutations in genes related to the NDH-1 complex (Table 1). In the genome of Synechocystis sp. PCC 6803, most of the genes encoding subunits of NDH-1 are present as single copies. However, ndhD and ndhF genes encoding large core membrane components comprise small gene families of six and three members, respectively (Kaneko et al. 1996; Ohkawa et al. 2000). Reverse genetics approaches have revealed that NdhF1/D1 is involved in respiratory electron transport, whereas NdhF3/D3 and NdhF4/D4 are constituents of an inducible high-affinity CO2 uptake system and a constitutively expressed low-affinity CO2 uptake system, respectively (Ohkawa et al. 2000; Shibata et al. 2001). Consistent with the information obtained from reverse genetics, proteome analyses of membrane protein complexes revealed that there are two types of NDH-1 complexes in the thylakoid membrane (Herranen et al. 2004; Zhang et al. 2004, 2005). The NDH-1L complex, containing NdhF1/D1 and all known single-copy ndh gene products, is constitutively expressed and may participate in respiratory electron transport. On the other hand, the NDH1MS complex, containing NdhF3/D3, CupA, and all known single-copy ndh gene products, is inducible under low-CO2 conditions and seems to provide energy for CO2 uptake systems. It is notable that eight strains had mutation in the ndhF3 gene and two had in the cupA gene among 13 suppressor mutants. No mutants had mutation in the ndhD3 gene, although NdhD3 is also the constituent of the NDH-1MS complex. L22 strain accumulated quite low level of the ndhCKJ transcripts probably due to the one base insertion into -10 box (Fig. 1). Since NdhC, NdhK, and NdhJ are the common components of the NDH complex, the accumulation level of not only NDH-1MS but also other NDH complexes should be low in this mutant. Since suppressor mutants were isolated from the genetically homogeneous DpmgA mutant, the identified mutations must occur after the exposure of the mutant to lethal photomixotrophic conditions. At first, we expected the identification of suppressor mutations from wide range of genes. However, the actual mutation sites were concentrated

Photosynth Res

on the ndhF3 gene. In bacteria, it has been reported that selective pressure may somehow favor the appearance of beneficial mutations and certain genes exhibit unusually high mutation frequency (Wurtzel et al. 2010; Gunka et al. 2012). In case of long-term evolutional experiment in E. coli with citrate, tandem amplification of citrate/succinate antiporter gene happens reproducibly and frequently (Blount et al. 2008, 2012). This mutation enables the gene to acquire the constitutively expressed promoter, and it enables the cells to utilize citrate aerobically. The ndhF3 gene may be one of such critical genes with high mutation frequency in Synechocystis genome to suppress the phenotype of the DpmgA mutant. The frequency of the mutation in the genes encoding subunits of NDH-1MS suggested that the loss of function of this complex involved in CO2 uptake might be effective to mitigate the growth inhibition of the DpmgA mutant under photomixotrophic conditions. However, the change in CO2 uptake activity upon the addition of glucose under 1 % CO2 conditions was similar between WT and the mutant (Fig. 3b). Before the addition of glucose, the CO2 uptake by NDH-1MS was already suppressed judging from the fact that the similar activity was observed between WT and the DpmgA/DndhF3 mutant not having NDH-1MS complex. Nevertheless, decrease in the CO2 uptake activity to 50–70 % of the initial level was observed in every strain after 24 h of incubation under photomixotrophic conditions. This means that the decrease could not be ascribed to the down-regulation of the NDH-1MS activity. L22 mutant exhibited lower activity than other strains at 0 h, and the decrease of the activity upon the addition of glucose was less conspicuous. Plausible explanation for these observations is that the activity of a constitutively expressed low-affinity CO2 uptake system supported by NdhF4/D4 is down-regulated under photomixotrophic conditions. Decrease of the CO2 uptake activity upon the addition of glucose seems to be a reasonable response to avoid the uptake of excess carbon. It is consistent with the observation that photosynthetic CO2 fixation is down-regulated together with the upregulation of sugar catabolism under the photomixotrophic conditions (Takahashi et al. 2008). Measurement of CO2 uptake activity revealed that the suppression of the DpmgA mutant phenotype by the mutation in genes encoding NDH-1MS subunits could not be explained by the defect in CO2 uptake. Furthermore, it could not be explained by the defect in cyclic electron transfer. The index of the cyclic electron transfer, i.e., the effect of the methyl viologen on the redox kinetics of P700 (Fig. 4), did not show substantial difference between the DpmgA mutant and DpmgADndhF3 mutant. The most striking feature of the redox kinetics of P700 is the effect of growth condition. Under photomixotrophic

growth condition, P700 could not be fully oxidized (Fig. 4d–f), suggesting the strong reducing pressure in photosynthetic electron transport chain, compared with the photoautotrophic condition. The reducing pressure to P700 could be partly explained by the electron donation from glucose to plastoquinone pool. However, the plastoquinone pool was reported to be almost fully reduced in the dark in the photoautotrophically grown WT cells (e.g., Ogawa et al. 2013). Addition of glucose to photoautotrophically grown cells did not induce drastic change in the redox sate of plastoquinone pool (data not shown). Acclimatory process to photomixotrophic growth condition, e.g., changes in the stoichiometry of redox components in the electron transfer chain, would be the cause of the high reducing pressure to P700, and this pressure, in turn, may be the cause of the growth defect observed in the DpmgA mutant. Since the reducing pressure to P700 was not affected by the mutation in the ndhF3 gene, the role of the NDH-1MS complex could not be ascribed to the effect on electron transport chain between plastoquinone and P700. It is evident that the loss of the NDH-1MS activity is effective to rescue the DpmgA mutant under photomixotrophic condition with 1 % CO2. However, there was little difference in the activities of CO2 uptake and cyclic electron transport among WT, the DpmgA mutant, and the DpmgA/DndhF3 mutant. The results suggest that the NDH-1MS complex has the third, currently unknown function. It was reported that the disruptant of ndhF3 gene results in the increased PSI/PSII ratio (Ogawa and Sonoike 2015). However, this also could not be the cause of the rescue effect, since the disruption of the pmgA gene also induces the increase of PSI/PSII ratio (Hihara et al. 1998). As discussed earlier, the addition of glucose could induce the reduction of PSI acceptor-side components. Although the glucose sensitivity cannot be explained by the simple over-reduction of the acceptor side of PSI, the redox balance of the stromal components including ferredoxin and thioredoxin is known to be the key process in many gene/ metabolic regulation in photosynthetic organisms (e.g., Horiuchi et al. 2010). It would be tempting to assume that the defect in the function of the ndhF3 gene restores the redox balance in the stroma that is altered by the disruption of the pmgA gene. The exact role of the NDH-1MS complex in the rescue of the DpmgA mutant should be pursued in future. Acknowledgments We thank Dr. Teruo Ogawa for setting up of an open gas analysis system and helpful discussion. This study was partly supported by MEXT-Supported Program for the Strategic Research Foundation at Private Universities, 2013–2017 (S1311017), by Grant-in-Aid for Young Scientists (B) (24780082) (to YK), by a grant from Research Fellow of Japan Society for the Promotion of Science (No. 26-7221) (to TO), and by The Japan Science and Technology Agency (JST) CREST program (to HY and YH).

123

Photosynth Res

References Aoki M, Katoh S (1982) Oxidation and reduction of plastoquinone by photosynthetic and respiratory electron transport in a cyanobacterium Synechococcus sp. Biochim Biophys Acta 682:307–314 Berna´t G, Appel J, Ogawa T, Ro¨gner M (2011) Distinct roles of multiple NDH-1 complexes in the cyanobacterial electron transport network as revealed by kinetic analysis of P700? reduction in various ndh-deficient mutants of Synechocystis sp. strain PCC6803. J Bacteriol 193:292–295 Blount ZD, Borland CZ, Lenski RE (2008) Historical contingency and the evolution of a key innovation in an experimental population of Escherichia coli. Proc Natl Acad Sci USA 105:7899–7906 Blount ZD, Barrick JE, Davidson CJ, Lenski RE (2012) Genomic analysis of a key innovation in an experimental Escherichia coli population. Nature 489:513–518 Fujimori T, Higuchi M, Sato H, Aiba H, Muramatsu M, Hihara M, Sonoike K (2005) The mutant of sll1961, which encodes a putative transcriptional regulator, has a defect in regulation of photosystem stoichiometry in the cyanobacterium Synechocystis sp. PCC 6803. Plant Physiol 139:408–416 Gunka K, Tholen S, Gerwig J, Herzberg C, Stu¨lke J, Commichau FM (2012) A high-frequency mutation in Bacillus subtilis: requirements for the decryptification of the gudB glutamate dehydrogenase gene. J Bacteriol 194:1036–1044 Haimovich-Dayan M, Kahlon S, Hihara Y, Hagemann M, Ogawa T, Ohad I, Lieman-Hurwitz J, Kaplan A (2011) Cross-talk between photomixotrophic growth and CO2-concentrating mechanism in Synechocystis sp. strain PCC 6803. Environ Microbiol 13:1767–1777 Herranen M, Battchikova N, Zhang P, Graf A, Sirpio S, Paakkarinen V, Aro EM (2004) Towards functional proteomics of membrane protein complexes in Synechocystis sp. PCC 6803. Plant Physiol 134:470–481 Hihara Y, Ikeuchi M (1997) Mutation in a novel gene required for photomixotrophic growth leads to enhanced photoautotrophic growth of Synechocystis sp. PCC 6803. Photosynth Res 53:243–252 Hihara Y, Sonoike K, Ikeuchi M (1998) A novel gene, pmgA, specifically regulates photosystem stoichiometry in the cyanobacterium Synechocystis sp. PCC 6803 in response to high light. Plant Physiol 117:1205–1216 Horiuchi M, Nakamura K, Kojima K, Nishiyama Y, Hatakeyama W, Hisabori T, Hihara Y (2010) The PedR transcriptional regulator interacts with thioredoxin to connect photosynthesis with gene expression in cyanobacteria. Biochem J 431:135–140 Kaneko T, Sato S, Kotani H, Tanaka A, Asamizu E, Nakamura Y, Miyajima N, Hirosawa M, Sugiura M, Sasamoto S, Kimura T, Hosouchi T, Matsuno A, Muraki A, Nakazaki N, Naruo K, Okumura S, Shimpo S, Takeuchi C, Wada T, Watanabe A, Yamada M, Yasuda M, Tabata S (1996) Sequence analysis of the genome of the unicellular cyanobacterium Synechocystis sp. strain PCC6803. II. Sequence determination of the entire genome and assignment of potential protein-coding regions. DNA Res 3:109–136 Kanesaki Y, Shiwa Y, Tajima N, Suzuki M, Watanabe S, Sato N, Ikeuchi M, Yoshikawa H (2012) Identification of substrainspecific mutations by massively parallel whole-genome resequencing of Synechocystis sp. PCC 6803. DNA Res 19:67–79 Li H, Durbin R (2010) Fast and accurate long-read alignment with Burrows-Wheeler transform. Bioinformatics 26:589–595 Li H, Ruan J, Durbin R (2008) Mapping short DNA sequencing reads and calling variants using mapping quality scores. Genome Res 18:1851–1858

123

Li H, Handsaker B, Wysoker A et al (2009) The sequence alignment/ map format and SAMtools. Bioinformatics 25:2078–2079 Min KT, Hilditch CM, Diederich B, Errington J, Yudkin MD (1993) Sigma F, the first compartment-specific transcription factor of B. subtilis, is regulated by an anti-sigma factor that is also a protein kinase. Cell 74:735–742 Muramatsu M, Hihara Y (2003) Transcriptional regulation of genes encoding subunits of photosystem I during acclimation to high-light conditions in Synechocystis sp. PCC 6803. Planta 216:446–453 Muramatsu M, Hihara Y (2012) Acclimation to high-light conditions in cyanobacteria: from gene expression to physiological responses. J Plant Res 125:11–39 Ogawa T, Kaplan A (1987) The stoichiometry between CO2 and H? fluxes involved in the transport of inorganic carbon in cyanobacteria. Plant Physiol 83:888–891 Ogawa T, Sonoike K (2015) Dissection of respiration and photosynthesis in the cyanobacterium Synechocystis sp. PCC6803 by the analysis of chlorophyll fluorescence. J Photochem Photobiol B: Biol 144:61–67 Ogawa T, Harada T, Ozaki H, Sonoike K (2013) Disruption of the ndhF1 gene affects Chl fluorescence through state transition in the cyanobacterium Synechocystis sp. PCC 6803, resulting in apparent high efficiency of photosynthesis. Plant Cell Physiol 54:1164–1171 Ohkawa H, Pakrasi HB, Ogawa T (2000) Two types of functionally distinct NAD(P)H dehydrogenases in Synechocystis sp. strain PCC 6803. J Biol Chem 275:31630–31634 Ozaki H, Ikeuchi M, Ogawa T, Fukuzawa H, Sonoike K (2007) Large-scale analysis of chlorophyll fluorescence kinetics in Synechocystis sp. PCC 6803: identification of the factors involved in the modulation of photosystem stoichiometry. Plant Cell Physiol 48:451–458 Peschek GA, Schmetterer G (1982) Evidence for plastoquinolcytochrome f/b563 reductase as a common electron donor to P700 and cytochrome oxidase in cyanobacteria. Biochem Biophys Res Commun 108:1188–1195 Shibata M, Ohkawa H, Kaneko T, Fukuzawa H, Tabata S, Kaplan A, Ogawa T (2001) Distinct constitutive and low-CO2-induced CO2 uptake systems in cyanobacteria: genes involved and their phylogenetic relationship with homologous genes in other organisms. Proc Natl Acad Sci USA 98:11789–11794 Shiwa Y, Fukushima-Tanaka S, Kasahara K, Horiuchi T, Yoshikawa H (2012) Whole-genome profiling of a novel mutagenesis technique using proofreading-deficient DNA polymerase d. Int J Evol Biol 2012:860797 Sonoike K, Hihara Y, Ikeuchi M (2001) Physiological significance of the regulation of photosystem stoichiometry upon high light acclimation of Synechocystis sp. PCC 6803. Plant Cell Physiol 42:379–384 Stanier RY, Kunisawa R, Mandel M, Cohen-Bazire G (1971) Purification and properties of unicellular blue-green algae (order Chroococcales). Bacteriol Rev 35:171–205 Takahashi H, Uchimiya H, Hihara Y (2008) Difference in metabolite levels between photoautotrophic and photomixotrophic cultures of Synechocystis sp. PCC 6803 examined by capillary electrophoresis electrospray ionization mass spectrometry. J Exp Bot 59:3009–3018 Williams JGK (1988) Construction of specific mutations in photosystem II photosynthetic reaction center by genetic-engineering methods in Synechocystis 6803. Method Enzymol 167:766–778 Wurtzel O, Dori-Bachash M, Pietrokovski S, Jurkevitch E, Sorek R (2010) Mutation detection with next-generation resequencing through a mediator genome. PLoS One 5:e15628 Zhang CC, Durand MC, Jeanjean R, Joset F (1989) Molecular and genetical analysis of the fructose-glucose transport system in the

Photosynth Res cyanobacterium Synechocystis PCC6803. Mol Microbiol 3:1221–1229 Zhang P, Battchikova N, Jansen T, Appel J, Ogawa T, Aro EM (2004) Expression and functional roles of the two distinct NDH-1 complexes and the carbon acquisition complex NdhD3/NdhF3/ CupA/Sll1735 in Synechocystis sp PCC 6803. Plant Cell 16:3326–3340

Zhang P, Battchikova N, Paakkarinen V, Katoh H, Iwai M, Ikeuchi M, Pakrasi HB, Ogawa T, Aro EM (2005) Isolation, subunit composition and interaction of the NDH-1 complexes from Thermosynechococcus elongatus BP-1. Biochem J 390:513–520

123