Genetics of the Blue Light-Dependent Signal Cascade That Controls Phototaxis in the Cyanobacterium Synechocystis sp. PCC6803

The Synechocystis sp. PCC6803 can move on a solid surface in response to light, a phenomenon called phototaxis. Although many of the photoreceptors involved in phototaxis have been identified, the mechanisms that regulate directional motility of Synechocystis are not well understood. Previous studies showed that a mutant lacking the blue light-using flavin (BLUF) photoreceptor PixD exhibits negative phototaxis under conditions where the wild type responds positively. PixD interacts with the pseudo-response regulator-like protein PixE in a light-dependent manner, suggesting that this intermolecular interaction is important for phototaxis regulation, although genetic evidence has been lacking. To gain further insight into phototaxis regulation by PixD–PixE signaling, we constructed the deletion mutants ΔPixE and ΔPixD–ΔPixE, and characterized their phenotypes, which matched those of the wild type (positive phototaxis). Because ΔPixD exhibited negative phototaxis, PixE must function downstream of PixD. Under intense blue light (>100 μmol m⁻² s⁻¹; 470 nm) the wild type exhibited negative phototaxis, but ΔPixD–PixE exhibited positive phototaxis toward low-intensity blue light (∼0.8 μmol m⁻² s⁻¹; 470 nm). These results suggest that an unknown light-sensing system(s), that is necessary for directional cell movement, can be activated by low-intensity blue light; on the other hand, PixD needs high-intensity blue light to be activated. We also isolated spontaneous mutants that compensated for the pixE deletion. Genome-wide sequencing of the mutants revealed that the uncharacterized gene sll2003 regulates positive and negative phototaxis in response to light intensity.

Introduction

Cyanobacteria are the oldest known oxygenic phototrophs. To perform photosynthesis efficiently, certain cyanobacterium strains move directionally on solid surfaces toward or away from light, a process known as phototaxis (Bhaya 2004, Yoshihara and Ikeuchi 2004). The model unicellular cyanobacterium Synechocystis sp. PCC6803 (hereafter referred to as Synechocystis) uses type IV pili for taxis. Sequential extraction, adhesion and retraction of pili located on one side of the cell are responsible for its directional movement (Bhaya et al. 2001, Yoshihara et al. 2001, Schuergers et al. 2015, Wilde and Mullineaux 2015). Recent high-resolution imaging of Synechocystis indicated that the bacterium, acting as a spherical microlens, focuses directional light at the cell edge distal to the light source (Schuergers et al. 2016). Although we now know how Synechocystis senses the light direction, the intracellular signaling mechanisms that control type IV pili movement as a function of the direction and intensity of a light source are still unclear.

Synechocystis responds to various wavelengths of light. It moves toward light between approximately 560 nm and approximately 720 nm (green and red light, respectively), and away from UV-A, blue and strong red light (∼360, ∼470 and ∼720 nm, respectively) (Choi et al. 1999, Ng et al. 2003), indicating that it uses different photoreceptors for phototaxis. These photoreceptors include the cyanobacteriochromes PixJ1/TaxD1 (Yoshihara et al. 2000, Bhaya et al. 2001), Cph2 (Wilde et al. 2002) and PixA/UirS (Narikawa et al. 2011, Song et al. 2011), the cryptochrome-like cyanopterin sll1629 (Moon et al. 2010) and the blue light-using flavin (BLUF) photoreceptor PixD/slr1694 (Masuda and Ono 2004, Okajima et al. 2005). Although the phototaxis responses of all photoreceptor mutants currently available are different from that of the wild type (WT), their responses are still directionally related, indicating that none of them is responsible for sensing the direction of the light source.

The BLUF photoreceptor PixD is relatively well characterized (Masuda 2013). The pixD deletion mutant (ΔPixD) moves away from white light whereas WT cells move toward the source (Masuda and Ono 2004, Okajima et al. 2005). PixD physically interacts with the response regulator-like protein PixE in the dark to form a PixD₁₀–PixE₄ (or PixD₁₀–PixE₅) complex in vitro (Okajima et al. 2005, Yuan et al. 2006, Yuan and Bauer 2008, Tanaka et al. 2011, Tanaka et al. 2012, Ren et al. 2013). However, PixE does not have the two conserved aspartates and the lysine found in response regulators (Ren et al. 2013), suggesting that the light signal is relayed as the protein–protein interaction instead of by a phosphorylation event. Light excitation causes the PixD–PixE complex to dissociate, and it has been suggested that monomeric PixE might act to switch phototaxis from the positive to negative mode (Masuda et al. 2008, Yuan and Bauer 2008, Tanaka et al. 2012, Ren et al. 2013, Ren et al. 2015). However, no in vivo evidence has been provided to support this hypothesis. PixE-related mechanisms that influence phototaxis behavior are also unknown. To characterize the physiological significance of the PixD–PixE interaction in vivo, for this report, we constructed a pixE deletion mutant (ΔPixE) and a pixD–pixE double deletion mutant (ΔPixDE) and compared their phototaxis characteristics with those of the WT and ΔPixD. PixE was found to function downstream of PixD to control phototaxis. We also found that an uncharacterized gene, sll2003, is involved in the phototaxis response of Synechocystis.

Results

To construct ΔPixD, pixD was replaced with a spectinomycin/streptomycin resistance gene cassette (Fig. 1A), and to construct ΔPixDE, pixE, pixD and the region separating the two were replaced with a kanamycin resistance gene cassette (Fig. 1B). Given that pixE is located upstream of pixD, inactivation of pixE by a gene cassette insertion may influence pixD expression. To avoid such a potential artifact, we first constructed a plasmid with pixE partially deleted and a spectinomycin/streptomycin resistance gene cassette placed downstream of pixD (Fig. 1C). This plasmid was transferred into ΔPixDE, and a spectinomycin resistant/kanamycin-sensitive colony was selected as ΔPixE. PCR indicated that the genetic manipulations were successful (Fig. 1D), which were further verified by genome re-sequencing analysis as described in the following section.

Fig. 1

Construction of ΔPixD, ΔPixE and ΔPixDE. Schematic representation of the construction of ΔPixD (A), ΔPixDE (B) and ΔPixE (C). Ω-Sp/Sm^r, spectinomycin/streptomycin resistance gene with the transcription/translation termination Ω-fragment (Fellay et al. 1987). Km^r, kanamycin resistance gene. (D) PCR to validate the segregation of the mutants from the WT. Positions of the primers used for PCR are shown in (B).

We assessed the phototaxis behaviors of mutants ΔPixD, ΔPixE and ΔPixDE. As reported (Masuda and Ono 2004, Okajima et al. 2005), ΔPixD moved away from white light (Fig. 2A). Conversely, ΔPixE, ΔPixDE and the WT moved toward white light (Fig. 2A). To assess the sensitivity of the strains to light, a suspension of each strain was spotted onto solidified agar at various distances from a white-light source, and their migration speeds (mm d⁻¹) were measured. (Note that light intensity is inversely related to the distance from the source since we used a normal fluorescent lamp for the experiment.) The speeds of the WT, ΔPixE and ΔPixDE were exponentially and directly dependent on the distance of each strain from the light source (Fig. 2B). Even under very low light, the cells did not move away from the light. The negative phototaxis of ΔPixD also depended on the light intensity, i.e. the initial distance from the light source (Fig. 2B; red). Interestingly, ΔPixD showed positive phototaxis under low-intensity light, although its migration speed was very slow. To analyze the light sensitivity of ΔPixD phototaxis more carefully, we spotted WT and ΔPixD cells on a plate at two distinct positions (left and right), which are separated by 25 mm, and then white light was illuminated from the left side to induce phototaxis (Fig. 2C). Under this condition, the WT, spotted on both sides, showed positive phototaxis. Interestingly, ΔPixD, spotted on the left-hand side, showed negative phototaxis at first, but, after passing the right-hand side position, the cells turned back and showed positive phototaxis. ΔPixD, spotted on the right-hand side, showed no taxis response. The light intensity at the right-hand side is approximately 15 μmol photons m⁻² s⁻¹.

Fig. 2

Phototaxis behavior of the WT, ΔPixD, ΔPixE and ΔPixDE. (A) Samples of the WT and its mutants were spotted onto a plate containing solidified 0.8% (w/v) agar infused with BG11 medium. The strains were illuminated laterally with white light provided by a fluorescent lamp (∼80 μmol m⁻¹ s⁻¹), as illustrated, for 4 d. (B) Samples of the WT and its mutants were spotted as described in (A), but at various distances from the white light source (fluorescent lamp). After incubation for 4 d, the distances that the cells had migrated toward (positive value) or away from (negative value) the light was measured. Black, WT; red, ΔPixD; blue, ΔPixE; green, ΔPixDE. (C) Cell cultures of the WT and ΔPixD were spotted onto a 0.8% (w/v) agar-solidified BG11 plate at two positions separated by 25 mm. White light provided by a fluorescent lamp was illuminated from approximately 3 cm to the left of the closest samples for 9 d. The light intensity for the samples on the right was 15 μmol m⁻¹ s⁻¹.

We next characterized the phototaxis behavior of the WT and ΔPixDE at the single-cell level. We individually spotted suspensions of these strains onto solidified agar in a glass-bottomed plate and observed the edges of each suspension with an inverted microscope. Red or blue light was irradiated from one side of a suspension at different intensities to induce phototaxis, and time-lapse images of the cells were recorded. The movements of single cells were traced and quantified by a computer program. Phototaxis motility was expressed as μm s⁻¹, with positive and negative signs to indicate movement toward and away from the light, respectively.

In the dark, individual WT and ΔPixDE cells move randomly (Fig. 3A and B, respectively) as reported previously (Choi et al. 1999). Under red light, the WT showed positive phototaxis; consequently, cells stacked up at the edge of the plate closest to the light source (Fig. 4A; right-hand side). By increasing the intensity of the red light, the phototaxis motility of the WT increased and the distribution of the motility became more Gaussian like (Fig. 3A). ΔPixDE also moved toward the red light (Fig. 3B). Notably, the average speed of ΔPixDE under low-intensity red light (≤600 μmol m⁻¹ s⁻¹) was greater than that of the WT (Fig. 4B). Under blue light illumination, the WT moved away from the light, with their average speed directly proportional to the intensity at > 100 μmol m⁻¹ s⁻¹ (Figs. 3A, 4B). However, when illuminated with blue light, ΔPixDE showed positive phototaxis (Figs. 3B, 4C). By increasing the blue light intensity, statistical deviation from a Gaussian curve of the negative and positive migration distances for WT and ΔPixDE cells, respectively, was reduced (Figs. 3A, B, 4C).

Fig. 3

Phototaxis motility distribution of single WT cells. The distribution of the migration velocity as dependent on the light intensity for singe WT (A) and ΔPixDE (B) cells. Data were fit to a Gaussian curve. Phototaxis was induced by lateral illumination of red or blue light (λ_max = 660 and 470 nm, respectively). The taxis response of single cells was measured using time-lapse images taken by an inverted microscope. Light intensity, reported as Einstein (E), is shown in each panel.

Fig. 4

Phototaxis behavior of single WT cells. (A) WT cells spotted onto 0.4% (w/v) solidified agar infused with BG11 medium were kept in the dark or illuminated laterally with red light (λ_max = 660 nm). Images were acquired at 0 and 30 min. Scale bars = 50 μm. (B) Dependence of phototactic mobility of WT and ΔPixDE cells on red light intensity (λ_max = 660 nm). (C) Dependence of phototactic mobility of WT and ΔPixDE cells on blue light intensity (λ_max = 470 nm). Data are the same as in Fig. 3.

We isolated mutants that arose spontaneously and exhibited negative phototaxis from ΔPixDE (Fig. 5A; Mutant 1) and ΔPixE cultures (Fig. 5A; Mutants 2 and 3). Then we performed genome re-sequencing analysis of the isolated mutants and WT to identify mutation sites responsible for the suppressor phenotype. Given that we used the substrain PCC-P as the WT, resequencing data were compared with the genome sequence of the PCC-P strain deposited previously (Kanesaki et al. 2012). Several single nucleotide polymorphisms were identified in our WT strain compared with the reference PCC-P strain; most of them are also found in the isolated mutants (Supplementary Table S1). The re-sequencing data indicated that a deletion in a cheA gene (sll0043) between positions 968 and 981 had occurred in Mutant 1, and a single deletion at position 809 in sll2003 had occurred in both Mutants 2 and 3 (Fig. 5B; Table S1), which are most probably candidates for the suppressor mutations, although other mutations in a hypothetical protein (sll0040), pixJ1 (sll0041) and/or a transcription–repair coupling factor (sll0377) are found in a specific mutant or all of the mutants (Supplementary Table S1). The PixJ1 mutation in Mutant 3 is an in-frame deletion mutation at the 3' end, suggesting that it is not a null mutation (Supplementary Table S1).

Fig. 5

Phototaxis mutants isolated from ΔPixE and ΔPixDE. (A) Schematic representation of mutant isolation. Spontaneous mutants (Mutants 1, 2 and 3) showing negative phototaxis were isolated from ΔPixDE and ΔPixE that showed positive phototaxis. (B) Mutation sites in Mutants 1, 2 and 3 as revealed by whole-genome sequencing. Mutant 1 had a deletion in cheA. Mutants 2 and 3 had a mutation in sll2003. Nucleotide sequence numbers are indicated.

To assess the roles of cheA and sll2003 during phototaxis, a cheA mutant (ΔCheA), cheA–pixE double mutant (ΔCheAPixE), sll2003 mutant (Δsll2003) and sll2003–pixE double mutant (Δsll2003PixE) were constructed. To determine their phenotypes, each strain was illuminated with red light in the lateral direction to induce phototaxis in the presence or absence of blue light illuminated from above. The WT, ΔPixD, ΔPixE and ΔPixDE moved toward the red light in the absence of blue light; in the presence of blue light, however, the WT and ΔPixD, but not ΔPixD and ΔPixDE, moved away from the red light (Fig. 6). ΔCheA and Δsll2003 showed negative phototaxis toward red light in the absence of blue light, whereas under this condition ΔCheAPixE and Δsll2003PixE moved toward the red light. In the presence of blue light, only Δsll2003PixE showed positive phototaxis, whereas the other mutants showed negative phototaxis or did not respond to red light (Fig. 6). The phototaxis behaviors of ΔCheAPixE and Δsll2003PixE were identical to those of Mutants 1 and 2, respectively (Supplementary Fig. S1).

Fig. 6

Phototaxis behavior of the WT and its mutants in response to red and blue light. Cell cultures of the WT and its mutants were spotted onto solidified 0.8% (w/v) agar infused with BG11 medium and laterally illuminated with red light (λ_max = 660 nm; intensity = 24 μmol m⁻¹ s⁻¹), as illustrated, in the presence or absence of vertically illuminated blue light (λ_max = 470 nm; intensity = 154 μmol m⁻¹ s⁻¹). Images were acquired after 3 d.

Discussion

We provide herein the first genetic evidence that PixD functionally interacts with PixE in vivo. Under conditions where ΔPixD showed negative phototaxis toward white light, ΔPixE and ΔPixDE showed positive phototaxis, the same response as the WT (Fig. 2A), indicating that the signal for negative phototaxis in ΔPixD is cancelled by the loss of PixE. PixD is known to interact with PixE in the dark in vitro to form a PixD₁₀–PixE₄ (or PixD₁₀–PixE₅) complex, and blue light irradiation destroys the interaction (Yuan and Bauer 2008, Ren et al. 2013). Because PixE is released from the complex in the presence of light, PixD probably also modulates the level of monomeric PixE in a light-dependent manner in vivo. The amount of monomeric PixE determines the direction of cell movement but is not required for the directional tactic response itself (Fig. 7). In ΔPixD, PixE constitutively inhibited positive phototaxis, allowing for negative phototaxis. White and red light caused the WT to move toward the light, and vertical illumination of intense blue light (154 μmol m⁻² s⁻¹) resulted in red light-dependent negative phototaxis (Fig. 6), suggesting that under lateral white light illumination with an intensity of 70 μmol m⁻² s⁻¹ (Fig. 2A), the lack of the PixD–PixE interaction in the WT is not strong enough to switch the tactic response from positive to negative.

Fig. 7

Schematic of the light-dependent signal cascade controlling phototaxis in Synechocystis. An unidentified light-sensing system(s) is responsible for recognition of the illumination direction and controls the retraction of type IV pili to induce biased movement. The unknown light-sensing system(s) is activated by red light and low-intensity blue light to induce positive phototaxis. PixE inhibits positive phototaxis. Conversely, CheA and PixJ1 up-regulate the positive phototaxis. sll2003 and sll1737 are involved in up- and down-regulation of positive phototaxis in response to low- and high-intensity light, respectively. PixD interacts with PixE in the dark or under low-intensity blue light, thereby inhibiting PixE activity. In the dark, some functional PixE monomer may exist and partially inhibit positive phototaxis. PixD functions as a sensor of blue light intensity; intense blue light is required to dissociate the PixD–PixE complex to release functional PixE. See text for details.

In our single-cell tracking experiments, red and blue light caused the WT to move in the positive and negative direction, respectively (Figs. 3A, 4). Notably, ΔPixDE moved toward not only red light but also blue light (Figs. 3B, 4B), suggesting that unknown light-sensing system(s), required for the directional tactic response, sense red and blue light (Fig. 7). Blue light-dependent positive phototaxis of ΔPixDE was observed under blue light of low intensity (∼0.8 μmol m⁻² s⁻¹) (Figs. 3B, 4C), suggesting that the unidentified light-sensing machinery can be activated by low blue light (Fig. 7). Conversely, blue light-dependent negative phototaxis of the WT was observed only under intense blue light (>100 μmol m⁻² s⁻¹; Fig. 4C), suggesting that the PixD-dependent inhibitory signal for positive phototaxis operates only under blue light of high intensity (Fig. 7). Disassembly of the PixD–PixE complex in vitro occurs when at least two PixD monomers in the complex are simultaneity photo-activated (this could be done only with strong blue light irradiation) (Tanaka et al. 2012), supporting the idea that PixD senses the intensity of blue light rather than simply the presence of blue light.

What is the ‘unknown light-sensing system(s)’ (Fig. 7) responsible for recognizing light direction? The photosynthetic apparatus is not the one responsible, since photosynthesis inhibitors such as DCMU did not affect phototactic behavior (Choi et al. 1999). As described in the Introduction, Synechocystis mutants of all photoreceptors characterized to date still exhibit directional movement. From these observations, one could assume that no specific photoreceptor is responsible for recognizing the light direction, but certain light-dependent metabolism is involved in the processes that are controlled by multiple light inputs. Switching of the tactic movement of ΔPixD from negative to positive by lowering the light intensity (Fig. 2C) takes a rather long time, since, if it were rapid, the cells would be stacked at the point of the light intensity of 15 μmol m⁻² s⁻¹, and would not pass through and turn back. Given that light signal transduction of photoreceptors is generally rapid, subsequent changes in certain metabolic processes may be involved in sensing the light direction.

We isolated the phototaxis mutants Mutant 1, 2 and 3 from ΔPixDE and/or ΔPixE mutational backgrounds (Fig. 5). The mutation sites map to cheA in Mutant 1 and to sll2003 in Mutants 2 and 3 (Fig. 5). CheA functions downstream of PixJ1 to control phototaxis, and ΔCheA and ΔPixJ1 showed negative phototaxis under the same conditions in which the WT showed positive phototaxis (Yoshihara et al. 2000). ΔCheA showed negative phototaxis under lateral red light illumination (Fig. 6). ΔPixE and ΔPixDE move toward red light, but their movement was inhibited by vertical blue light illumination (Fig. 6), suggesting that red light and PixJ1 signaling (producing positive phototaxis) are suppressed (Fig. 7). Given that ΔCheAPixE showed positive phototaxis toward red light (Fig. 6) and that ΔCheA moved away from red light, some monomeric PixE may exist in ΔCheA (and by extrapolation in the WT) to oppose positive phototaxis (Fig. 7). Therefore, even when PixD associated completely or almost completely with PixE (i.e. in the dark or in the absence of blue light), an excess of active monomeric PixE may have been present in the WT.

ΔPixD did not show negative phototaxis when illuminated with white light at an intensity of approximately 15 μmol m⁻² s⁻¹ (Fig. 2C), suggesting that signaling for positive and negative phototaxis at this intensity is in equilibrium. Given that PixE-dependent signaling—inhibiting positive phototaxis—would have its greatest effect in ΔPixD, the activation signal for positive phototaxis may be up-regulated as the light intensity decreases. Interestingly, Δsll2003 showed negative phototaxis under lateral red light illumination (Fig. 6). Under this condition, ΔPixD showed positive phototaxis, suggesting that sll2003 is involved in the up-regulation of positive phototaxis in response to low-intensity light (Fig. 7). Positive phototaxis of ΔPixE towards red light was cancelled out by vertical illumination of intense blue light; on the other hand, Δsll2003PixE showed positive phototaxis under these conditions (Fig. 6), suggesting that sll2003 is also involved in the down-regulation of positive phototaxis in response to high-intensity light (Fig. 7). The primary structure of sll2003 is similar to that of glycosyltransferases. A yeast-two-hybrid study indicated that sll2003 interacts with sll1737, a cyanobacterial homolog of Tic20, which is part of a protein import apparatus associated with the inner envelope membrane of the chloroplast (Sato et al. 2007). These observations suggest a relationship between a cytoplasmic glycosyltransferase(s) and protein import/export activities for phototaxis control, although details of this interaction(s) have not been elucidated. One possibility is that such cellular activities of sll2003 and sll1737 are important for controlling metabolic changes responsible for directional movement. Elucidation of the catalytic activity and substrates of sll2003 should help us understand its functional role(s) during phototaxis; such studies are currently underway in our laboratory.

Materials and Methods

Culture conditions

Synechocystis sp. PCC6803 substrain PCC-P (Yoshihara et al. 2000), kindly provided by Dr. Naoki Mizusawa at Hosei University, served as the WT. Cells were cultured in BG11 medium at 30 °C under white, red or blue light, provided by a Hitachi FHF32EX fluorescent lamp, a Sanyo MIL-RI8 red light-emitting diode (LED; λ_max = 660 nm) or a Sanyo MIL-BI8 blue LED (λ_max = 470 nm), respectively. The phototaxis behavior of the strains was examined with the bacteria spotted onto solidified 0.4% or 0.8% (w/v) agar containing BG11 medium and with lateral white or red light illumination. Escherichia coli were cultured in Luria–Bertani medium. Antibiotics were added to Synechocystis cultures [spectinomycin (8 μg ml⁻¹) or kanamycin (10 μg ml⁻¹)] and to E. coli cultures [ampicillin (100 μg ml⁻¹), spectinomycin (40 μg ml⁻¹) or kanamycin (50 μg ml⁻¹)] when appropriate.

Mutant constructions

Primers used for construction of the mutants are listed in Supplementary Table S2. The construction of ΔPixD has been reported (Masuda and Ono 2004). ΔPixDE was constructed as follows. First, 1,000 bp regions upstream and downstream of pixE and pixD were individually amplified with primer pairs PixE-UP-EcoRI-Fsion/PixE-UP-R-PstI and PixD-Down-F-PstI/PixE-Down-R-Fsion, respectively. The two fragments are mixed and inserted into an EcoRI–SphI-cut pUC18 plasmid (TAKARA) using In Fusion Cloning kit reagents (Clontech). This plasmid was digested with PstI, and a PstI-cut kanamycin resistance gene cassette was ligated into it (Alexeyev 1995). The resulting plasmid was transferred into the WT, and ΔPixDE was isolated from a plate containing agar infused with BG11 medium and kanamycin. Complete gene segregation was validated by PCR with primer pairs Primer 1/Primer 2 and Primer 2/Primer 3 (Fig. 1B).

ΔPixE was constructed as follows. First, the 1,000 bp region upstream of pixE to the pixD stop codon was amplified with primers PixE-UP-EcoRI-Fsion2 and PixD-R-SmaI. The 1,000 bp downstream region of pixD was amplified with primers PixD-Down-F-SmaI and PixD-Down-R-SphI-Fsion2. The two PCR-amplified fragments were mixed and then inserted into EcoRI–SphI-cut pUC18 using In Fusion Cloning kit reagents. The resultant plasmid contained full-length pixE and pixD, the regions 1,000 bp upstream of pixE and 1,000 bp downstream of pixD, and an SmaI restriction site just after the pixD stop codon. Using this plasmid as the template, the 1 kb region upstream of pixE that also contained the pixE start codon was amplified with primers PixE-UP-EcoRI-Fsion2 and PixE-UP-R-FLAG-dell. The fragment covering the final five codons of pixE, all of pixD and the 1,000 bp region downstream of pixD was amplified using primers PixED-F-FLAG and PixD-Down-R-SphI-Fsion2. The two fragments were mixed and then cloned into EcoRI–SphI-cut pUC18 using In Fusion Cloning kit reagents. The resultant plasmid was digested with SmaI and ligated with an SmaI-cut spectinomycin/streptomycin resistance gene cassette (Fellay et al. 1987). The resulting plasmid (Fig. 1C) was transferred into ΔPixDE, and ΔPixE was isolated from a plate containing agar infused with BG11 medium and spectinomycin. Complete gene segregation was first validated by the sensitivity of the clone to kanamycin and then by PCR with Primer 3 and Primer 4 (Fig. 1B).

ΔCheA and ΔCheAPixE were constructed as follows. First, the 500 bp regions upstream and downstream of cheA were amplified with primer pairs cheAupF/cheAupR and cheAdownF/cheAdownR, respectively. The two fragments were mixed and inserted into BamHI-cut pUC18 using In Fusion Cloning kit reagents. The BamHI site was not conserved at the vector ends but was situated between the two inserted fragments. The resultant plasmid was digested with BamHI and ligated with a BamHI-cut kanamycin resistance gene cassette (Alexeyev 1995). The resulting plasmid was transferred into the WT and ΔPixE. ΔCheA was isolated from a plate containing agar infused with BG11 medium and kanamycin, and ΔCheAPixE was isolated from a plate containing agar, kanamycin and spectinomycin. Complete gene segregation was validated by PCR using primers cheAupF and cheAdownR.

Δsll2003 and Δsll2003PixE were constructed as follows. First, the 500 bp regions upstream and downstream of sll2003 were amplified with primer pairs sll2003-500bp-F/sll2003-500bp-R and sll2003downF/sll2003downR, respectively. The two fragments were mixed and inserted into BamHI-cut pUC18 using In Fusion Cloning kit reagents. As with the vector construction for ΔCheA, the BamHI site was situated between the two inserted fragments. The resultant plasmid was digested with BamHI and ligated with a BamHI-cut kanamycin resistance gene cassette (Alexeyev 1995). The resulting plasmid was transferred into the WT and ΔPixE. Δslll2003 was isolated from a plate containing agar infused with BG11 medium and kanamycin, and Δsll2003PixE was isolated from a plate containing agar infused with BG11 medium, kanamycin and spectinomycin. Complete gene segregation was validated by PCR with primers sll2003-500bp-F and sll2003downR.

Single-cell motility and tracking measurements

When the Synechocystis culture had an OD₇₃₀ of 0.3−0.5, the cells were collected and suspended in fresh BG11 medium at an OD₇₃₀ of 0.05−0.1. Aliquots of the cell suspensions (5 μl) were spotted onto 3 ml of 0.4% (w/v) solidified agar containing BG11 medium in a 35 mm glass-bottomed plate. The spotted cultures were dried for 15 min and then incubated for 18 h at 30 °C under white light. Before microscopy, the plates were covered with black plastic tape leaving widows at the top and bottom and on one side to allow for lateral light illumination, blue (λ_max = 470 nm) or red light λ_max = 660 nm). Light intensity was measured with a LI-COR LI-250 light meter. Time-lapse images were captured at one frame every 5 s and room temperature (∼25 °C) for 10 min using a Zeiss AxioObserverZ1 inverse microscope equipped with a ×40 objective lens (LD40XPh/0.6). Particle Track and Analysis (https://github.com/arayoshipta/projectPTAj) that runs on the ImageJ software package (https://imagej.nih.gov/ij/) was used for single-cell tracking analysis. To determine the motility bias of the phototaxis response, the velocity of each single cell was calculated by dividing the total displacement path length by the displacement time (Chau et al. 2015). Gaussian fitting used IGOR Pro (WaveMetrics, Inc.).

Genome re-sequencing

Genomic DNAs of the WT and suppressor mutants were isolated and purified using a Puregene Yeast/Bact. Kit (QIAGEN) in accordance with the manufacturer’s instructions. The library was prepared following the protocol of the Illumina TruSeq DNA PCR-Free Library Preparation Kit. Sequencing for 69 bp single reads was performed on the Illumina GAIIx platform. Reads were mapped against the PCC6803 reference genome (NCBI Reference Sequence: NC_020286.1) using BWA 0.7.12 (Li and Durbin 2009), and pileups of the read alignments were produced by SAMtools version 1.2 (Li et al. 2009). Mutations found in this study were compared with re-sequencing data of substrains GT-Kazusa, GT-S, GT-I, PCC-N, PCC-P (Kanesaki et al. 2012) and PCC-M (Trautmann et al. 2012) deposited previously (Supplementary Table S1). Re-sequencing results have been submitted to the DDBJ with accession numbers (DRA Accession) DRA004890 (WT), DRA004892 (Mutant 1) DRA004891 (Mutant 2) and DRA005140 (Mutant 3) under BioProject Accession PRJDB5012.

Supplementary data

Supplementary data are available at PCP online.

Funding

This work was supported by the Japan Society for the Promotion of Science [KAKENHI (25117508) and (16K14694), (16H03280) to S.M.].

Acknowledgments

We thank Dr. Yoshiyuki Arai (Osaka University) for suggestions and for use of Particle Track and Analysis, Professor Naoki Mizusawa (Hosei University) for providing the Synechocystis PCC-P strain, and Dr. Daisuke Nakane and Professor Takayuki Nishizaka (Gakushuin University) for suggestions concerning single-cell microscopy.

Disclosures

The authors have no conflicts of interest to declare.

Abbreviations

Abbreviations
 
• BLUF

  blue light-using flavin

 
• LED

  light-emitting diode

 
• WT

  wild type. Re-sequencing data have been submitted to the DDBJ with accession numbers DRA004890, DRA004892, DRA004891 and DRA005140 under BioProject Accession PRJDB5012