Phenotypic characterization of Synechocystis sp. PCC 6803 substrains reveals differences in sensitivity to abiotic stress

Synechocystis sp. PCC 6803 is a widely used model cyanobacterium, whose substrains can vary on both genotype and phenotype levels. Previously described phenotypic variations include ability of mixotrophic growth, ability of movement on agar plates and variations in pigments composition or cell size. In this study, we report for the first time significant variation among Synechocystis substrains in complex cellular traits such as growth rate, photosynthesis efficiency, cellular dry weight and cellular composition (including protein or carbohydrates content). We also confirmed previously reported differences in cell size. Synechocystis cultures were cultivated in controlled environment of flat panel photobioreactors under red, blue and white light of intensities up to 790 μmol(photons) m-2 s-1, temperatures 23°C–60°C, input CO2 concentrations ranging from 400 to 15 000 ppm and in BG11 cultivation medium with and without addition of NaCl. Three Synechocystis substrains were used for the comparative experiments: GT-L, GT-B (Brno, CZ) and PCC-B (Brno, CZ). Growth rates of Synechocystis GT-B were inhibited under high intensities of red light (585–670 nm), and growth rates of both substrains GT-B and PCC-B were inhibited under photons of wavelengths 485–585 nm and 670–700 nm. Synechocystis GT-B was more sensitive to low temperatures than the other two tested substrains, and Synechocystis GT-L was sensitive to the presence of NaCl in the cultivation media. The results suggest that stress sensitivity of commonly used Synechocystis substrains can strongly vary, similarly as glucose tolerance or motility as reported previously. Our study further supports the previous statement that emphasizes importance of proper Synechocystis substrains selection and awareness of phenotypical differences among Synechocystis substrains which is crucial for comparative and reproducible research. This is highly relevant for studies related to stress physiology and development of sustainable biotechnological applications.


Introduction
Cyanobacteria, autotrophic prokaryotes, are widely used as model organisms for studying basic principles of photosynthesis, stress responses or evolutionary processes. They also possess great promises for utilization in biotechnology, mostly for their capacity to generate biomass on short timescale and to produce wide range of valuable compounds [1,2]. Among The results of this study bring new insight into our understanding to the complexity of phenotypic variability in the model cyanobacterium Synechocystis sp. PCC 6803. The results emphasize importance to consider specific cultivation requirements and capabilities of stress tolerance of each selected Synechocystis substrain. This can be expected to represent a crucial aspect in studies that address effects of environmental stress on Synechocystis physiology. Additionally, our results can have large consequences for biotechnological applications that consider Synechocystis as a production strain.

History of the strains
Three Synechocystis substrains were used for the comparative growth experiments. Substrain GT-L was kindly provided by Prof. D. A. Los (Timiryazev Institute of Plant Physiology, Russian Academy of Sciences, Moscow, RU) in 2013. Growth capacities of this strain were described in detail previously [19]. To compare phenotypic signature of this substrain with another glucose tolerant Synechocystis, a substrain kindly provided by Dr. Patrik R. Jones was used (Imperial College London, GB, when providing the strain (early 2012), Dr. P. R. Jones  Rippka et al. (1979), and Williams (1988) [7,16,50], as well as in accordance to personal communication with members of laboratories in Jerusalem and in Moscow. was staying at the University of Turku, FI). This substrain was designated as GT-B (Brno, CZ). As a reference, Synechocystis sp. PCC 6803 regularly ordered from Pasteur Culture Collection (Paris, FR) in late 2013 was used, denoted as PCC-B (Brno, CZ). Full history of all three substrains used in this work represented in Fig 1. After receiving, all substrains were resuspended in fresh cultivation medium BG11 [6] supplemented with 17 mM HEPES (Carl Roth, Karsruhe, Germany) and cultivated on standard orbital shaker in the cultivation chamber as described in the next section. After reaching late exponential / early linear phase, the cells were gently centrifuged (2000 g, 10 min, 25˚C), resuspended in 10 ml of fresh cultivation medium, supplemented with methanol to a final concentration of 5%, and long-term cryopreserved in -80˚C until further work.

Inoculum cultures conditioning
Before comparative growth experiments, the frozen cultures were thawed on ice in the dark. After thawing, the cultures were resuspended in fresh BG11 medium (supplemented with 17 mM HEPES) in 250 ml Erlenmeyer flask, placed on orbital shakers (120 rpm) and covered with a paper tissue for at least 24 hours to protect the cells from the excess of light. The inoculum cultures were cultivated on orbital shakers in a cultivation chamber that was tempered at 31˚C under illumination of intensity 110 μmol(photons) m -2 s -1 (provided by warm white light LEDs) and under ambient air atmosphere. Each strain was maintained on orbital shakers for a period of one to three months; during that time all cultures were periodically resuspended in a fresh cultivation medium at least twice a month.
All cultures were periodically checked for contamination using standard LB agar plates. In addition, the cultures were routinely checked for contamination prior to their inoculation into the photobioreactors, either using standard LB agar plates or using ImageStream X MkII imaging flow cytometer (Amnis Corp., Seattle, WA, USA). The cytometric detection of bacterial contamination was based on SYBR 1 Green I nucleic acids gel staining (Thermo Fisher Scientific, Waltham, MA USA) that marks DNA of both cyanobacteria and bacteria cells. Nucleic acids were marked by addition of 5 μl of SYBR 1 Green I solution (diluted 1:100 in DMSO) to 500 μl of moderate dense cells suspension. The samples were incubated for 10 min in dark at laboratory temperature. During the cytometric analysis, 488 nm argon laser was used to excite both SYBR 1 Green I and chlorophyll a, and 642 nm laser was used to excite phycobilisomes. To identify Synechocystis cells within all measured objects, gating of the measured populations was applied to discriminate: a) focused objects, b) objects with both pigments as well as with SYBR 1 Green fluorescence, c) single cell and round objects (width/ length ratio between 0.8-1.0) and d) objects of reasonable size (2-5 μm in diameter). The bacteria cells were discriminated as following: a) focused objects, b) objects with SYBR 1 Green fluorescence and no chlorophyll fluorescence, and c) object smaller than 2 μm in diameter.
For comparative growth experiments, only inoculum cultures with no bacteria proliferation as observed with the use of the standard LB agar plates were used. In addition, both inoculum and running cultures in the photobioreactors that contained more than 2% of bacteria (as detected by ImageStream MkII imaging flow cytometer) were always discarded and not used for any further work. Contamination in the selected range (0-2% bacteria) did not have any impact on the growth rates or any other measured parameter-within the selected contamination range, the results were always highly reproducible for all Synechocystis substrains.

Photobioreactor
Comparative growth experiments were performed in flat panel photobioreactors that were described in detail previously [17]. Two types of photobioreactor illumination panels were used. The first panel was designed as a chessboard configuration of red and blue LEDs, and the second panel was designed as a chessboard configuration of red and white LEDs, respectively (red: λmax % 633 nm, λ1/2 % 20 nm, Luxeon LXHLPD09; blue: λmax % 445 nm, λ1/2 % 20 nm, Luxeon LXHL-PR09; white: Luxeon LXHL-PW09; all manufactured by Future Lighting Solutions, Montreal, QC, Canada). Spectral characteristics of the LEDs are shown in [19]. The photobioreactor continuously measured optical density (OD) by an inbuilt densitometer and steady-state pigment fluorescence emission yield by a build-in fluorometer (both described in [17]). Dissolved O 2 was monitored by InPro6800 electrodes (Mettler-Toledo, Inc., Columbus, OH, USA). Culture homogenization (mixing) was secured by inflow gas bubbling with rate of 200 ml min -1 . The bubbling was complemented by rotation of magnetic stirrer bar (ø5 × 35 mm, 210 rpm) in a vertical plane. Culture temperature and pH were monitored by InPro3253 electrode (MettlerToledo, Inc.). All other photobioreactor accessories were the same as described in [19].

Experimental setup
The growth characterization was performed in two regimes: in a quasi-continuous regime and in a batch regime. The quasi-continuous cultivation setup (turbidostat) was operated in accordance with   [19]. Briefly, the exponentially growing cells were maintained in a defined optical density range (measured at 680 nm, OD 680 ) by controlled dilution by fresh cultivation medium. The optical density was measured by the photobioreactor optical detector, and the OD 680 range was set to 0.52-0.58, which corresponded to approximately 2-4 x 10 7 cells ml -1 (Fig A in S1 File). Starting OD 680 of all cultures was 0.1-0.2, which corresponded to approximately 2-4 x 10 6 cells ml -1 . Once the culture density reached OD 680 0.58, the quasi-continuous cultivation setup was initiated by starting automated culture dilution within the selected OD 680 range. During the quasi-continuous regime, the cultures were growing under each specific cultivation condition for at least 24 hours. This period was long enough to reach growth stability, i.e. to acclimate the cells to specific cultivation conditions. Growth rates were calculated from dynamics of OD 680 increase in the specified OD 680 range. Growth stability was evaluated by measuring at least three reproducible successive growth rates, as calculated from independent growth periods. At 680 nm, light is absorbed by both cells structure (e.g. cell count and morphology) and pigments content (mostly chlorophyll a in the case of Synechocystis). The use of OD 680 for growth evaluation [19] allowed us to detect stabilization of both combined cells characteristics. Moreover, in the quasi-continuous characterization setup, biochemical composition of the cells (including pigment content) kept stable for each examined condition after growth stabilization and thus OD 680 readouts could be reliably used for growth rate determination with more favorable signal-to-noise ratio when compared to OD 720 . In the course of quasi-continuous experiments, the tested Synechocystis substrains were cultivated under red light of intensities 25-660 μmol(photons) m -2 s -1 , blue light of intensities 0-220 μmol(photons) m -2 s -1 , and white light of intensities 0-790 μmol(photons) m -2 s -1 , and in the temperature range 23˚C-38˚C.
Under higher temperatures, the cultures were not able to grow in the quasi-continuous regime any more, and it was only possible to introduce the heat shock for short time during a batch cultivation. The batch regime was operated according to [20]. The batch conditions were set to be saturating for growth of all tested Synechocystis substrains from at least 95%, i.e. 220 μmol(photons) m -2 s -1 of red light, 25 μmol(photons) m -2 s -1 of blue light, and 32˚C (see Fig 2 and Results for further details). Synechocystis sensitivity to NaCl was tested in the batch regime under the same cultivation conditions. During all batch experiments, Synechocystis was supplemented with CO 2 concentration of 5 000 ppm, which was saturating for growth of all tested substrains (Fig B in S1 File).

Analytical methods
The oxygen evolution/respiration rates (photosynthetic-irradiance curves) were measured within the photobioreactor cuvette by turning off aeration for 10 min, during a 5 min dark period followed by a 5 min light period, according to [20]. Pigment fluorescence curves (O-J-I-P-S-M) were measured by Fluorometer FL-100 (Photon System Instruments Ltd., Brno, CZ) which was equipped with 635 nm LED of intensity c. 2 000 μmol(photons) m -2 s -1 for pigments fluorescent excitation (the LED provided both actinic and measurement light). Carbon uptake was evaluated using Gas Analyzing System (Photon System Instruments Ltd., Brno, CZ), described in detail previously [18]. The mixture of air and CO 2 was bubbled for 5 min through the photobioreactor cuvette in a closed loop, where the only carbon loss was possible due to the photosynthetic fixation.
Content of chlorophyll a and total carotenoids was measured spectrophotometrically following protocol of   [21]. Content of phycobilisomes was measured by modified method of Bennett et Bogorad (1973) [22]. Briefly, 10 ml of culture suspension was collected in 15-ml falcon tubes and centrifuged (4 000 g, 10 min, 4˚C), supernatant was  [19]). All experiments were carried out in a quasi-continuous regime as described in the main text under growth saturating CO 2 concentration of 5 000 ppm. Each point represents average of at least four independent experiments, error bars represent standard errors. The dashed lines in panels A and B represent fitting of the data points by the function derived by Platt et al. (1980) [51]. The dashed lines in panels C and D represent linear fitting of the data points by the least squares method. Legend in panel D represents particular combinations of white light (W) and red light (R) in units of μmol(photons) m -2 s -1 .g. https://doi.org/10.1371/journal.pone.0189130.g002 Phenotypic variation in Synechocystis sp. PCC 6803 substrains under abiotic stress discarded and pellet was stored in -80˚C for 2 months until further processing. After thawing on ice, the pellets were resuspended in PBS buffer (pH 7.4) and centrifuged (4 000 g, 10 min, 4˚C). Supernatant was partially discarded and the cells were transferred to Eppendorf tubes on ice. The Eppendorf tubes were centrifuged again (15 000 g, 10 min, 4˚C), supernatant was discarded and the tubes were put to -80˚C for 30 min. Frozen samples were lyophilized overnight. After lyophilization, glass beads were added to dry cellular pellets and the samples were homogenized on Silamat S6 homogenizer (Ivoclar Vivadent AG, Schann, LI) for 30 sec. After homogenization, PBS buffer (4˚C) was added to the tubes for the phycobilisomes extraction, the samples were vortexed for 3 sec and put on ice in the dark for 30 min. After extraction, the samples were centrifuged (15 000 g, 10 min and 4˚C) and concentration of phycobilisomes was measured with the use of spectrophotometer UV 2600 (Shimadzu Scientific Instruments, Inc., Columbia, MD, USA) according to following equations: Content of cellular proteins was measured by Bicinchoninic Acid Protein Assay Kit (Sigma-Aldrich, St. Louis, MO, USA), using bovine serum albumin as a standard, and following instructions of the manufacturer. Content of cellular saccharides was measured according to [23]. Briefly, after harvesting, 1 ml of cellular suspension was centrifuged (20 000 g, 15 min, 25˚C) and the pellet was resuspended in 1 ml of distilled water. 8 x 30 μl of each sample was transferred to 96-well plate, and 30 μl of 5% phenol, followed by 150 μl of 96% H 2 SO 4 was added to each well (after each step, the suspension was mixed by gentle pipetting). The plates were covered with the original plate lid and the samples were incubated at laboratory temperature for 60 min. The concentration of cellular saccharides was measured by Multiskan™ GO Microplate Spectrophotometer (Thermo Fisher Scientific, Waltham, MA, USA) at 490 nm, using D-glucose as a standard.
Content of glycogen was measured according to [20]. After harvesting from the photobioreactor, 10 ml of culture suspension was centrifuged (4 000 g, 15 min), supernatant was discarded and pellet was resuspended in 5 ml of distilled water. Aliquots of 480 μl were transferred to eight Eppendorf tubes. The samples were centrifuged (20 000 g, 10 min, 4˚C), supernatant was discarded and 480 μl of cold methanol was added to each tube. The samples were homogenized by gentle pipetting, placed to +4˚C for 20 min and centrifuged again at 4˚C. The methanol supernatants were used for analysis of chlorophyll a and carotenoids as described above and the pellets were stored in -80˚C until further processing for up to 3 months. After thawing on ice, the pellets were resuspended in 200 μl of 30% KOH and they were incubated at 95˚C for 90 min. After cooling down at laboratory temperature, the samples were mixed with 1.2 ml of cold ethanol (+4˚C) and stored at -20˚C overnight. The precipitated carbohydrates were then separated by intensive centrifugation (20 000 g, 70 min, 4˚C). The supernatant was discarded and the pellet was dried at 60˚C for 30 min in Vacuum Concentrator 5305 (Eppendorf, Hamburg, DE). Dry pellets were diluted in 480 μl of deionized water and transferred to 96-well plate. Content of carbohydrates in the samples was measured spectrophotometrically according to [23], following the same procedure as for determination of total cellular saccharides.
Cell size was determined using ImageStream MkII imaging flow cytometer (Amnis Corp., Seattle, WA, USA). Right after harvesting from the photobioreactor, 500 μl of the culture suspension was centrifuged (4 000 g, 4 min, 25˚C), supernatant was discarded, pellet was resuspended in 0.25% glutaraldehyde solution and the samples were incubated for 10 min at laboratory temperature. The fixed cells were stored in -80˚C until further processing (up to 1 month in total). For further analysis, the samples were thawed on ice for 2 hours, and they were kept at laboratory temperature for additional 30 min after thawing. During the cytometric analysis, only bright field images were collected. Gating of the measured populations was applied to discriminate: a) focused objects (using combination of both RMS gradient and Treshold Mask features of IDEAS1 software), and b) round objects (width/ length ratio between 0.9-1.0). The imaging flow cytometer was calibrated with nonfluorescent microspheres (1-15 μm, Thermo Fisher Scientific, Waltham MA, USA) and the results were validated with light microscope Axio Imager 2 (Carl Zeiss, Oberkochen, DE). Cellular dry weight was measured using XA105DR analytical balances (Mettler Tolledo, Greifensee, CH). Cell count was measured with Cellometer Auto M10 (Nexcelom Bioscience, Lawrence, MA, USA).
To test statistical differences among Synechocystis substrains under the tested conditions, one-way ANOVA followed by Tukey's HSD post-hoc test was applied, using STATISTICA 1 software (Dell Inc., Tulsa, OK, USA).

General features of the tested Synechocystis substrains
All Synechocystis substrains used in this study were tolerant to the presence of 5 mM glucose in liquid cultivation medium as well as in solid BG11 agar plates. Synechocystis substrains GT-B and GT-L were not motile, whereas substrain PCC-B showed negative phototaxis on agar plates (top illumination by warm white light through a 3 cm circular area in the middle of standard 10 cm agar plate). This is in agreement with previously published motility distribution among PCC-derived and ATCC-derived Synechocystis substrains [10,13]. All Synechocystis substrains used in this study were not sensitive to the presence of shear stress introduced by rotation of the cylindrical stirrer bar inside the cuvette up to 210 rpm or maximal circular velocity 0.4 m s -1 . Also intensity of bubbling between 50-200 ml min -1 had no effect on the growth of all three tested Synechocystis substrains (tested with CO 2 input concentration of 5 000 ppm, Fig C in S1 File).

Synechocystis sensitivity to high light
Red light (585-670 nm). During quasi-continuous cultivation under 25-660 μmol(photons) m -2 s -1 of red light, the cultures were supplemented with 25 μmol(photons) m -2 s -1 of blue light (Fig 2A and 2B), following the same setup as introduced previously [19]. The red light sensitivity experiments were performed at 32˚C (temperature that saturated growth of substrain GT-L from 95%), and at 35˚C (optimal growth temperature for this substrain [19]). Growth rates of both substrains GT-L and GT-B were not photoinhibited even under 660 μmol(photons) m -2 s -1 of red light. However, growth rates of substrain GT-B started to decrease under red light of intensities higher than 220 μmol(photons) m -2 s -1 (Fig 2A). This growth was even more pronounced when the cultivation temperature was increased from 32˚C to 35˚C (Fig 2B).
To provide additional insight into operation of photosynthetic electron transport chain, a series of fluorescence induction curves (O-J-I-P-S-M) was measured under low and high red light (Fig 3E-3G) as well as under high white light (Fig 3H). Initial part of the fluorescence induction curve (O-J-I-P, the time frame <1 sec) was almost identical for all Synechocystis substrains, which was confirmed also by observing minimal differences in derived O-J-I-P curve parameters (Fig E  in S1 File). Also P-S fluorescence dip (c. 1-20 sec of illumination) was similar in all tested Synechocystis substrains. However, the amplitude of S-M fluorescence rise (c. 20 -120 sec of illumination) was significantly reduced under both high red and white light in substrain GT-B (Fig 3), which suggests differences in photosynthesis induction in this substrain.
Under high intensities of red light, content of cellular pigments (including chlorophyll a, total carotenoids, phycocyanin and allophycocyanin) was similar in the cells of all tested Synechocystis substrains (Fig F in S1 File). On the other hand, cellular composition of the tested Synechocystis substrains (including content of cellular carbohydrates, Fig 4, and proteins, Fig G in S1 File) varied under high red light, similarly as cell size and cellular dry weight (Fig 4). The substrain GT-L was having cells of the largest size under 220 and 660 μmol(photons) m -2 s -1 (c. 2.4  7-2.0 μm, Fig  4A). Cellular dry weight under high red light (660 μmol(photons) m -2 s -1 ) was the highest in substrain GT-L (7443±356 fg cell -1 ) and the lowest in substrain GT-B (4836±1267 fg cell -1 , Fig  4B). Cellular dry weight expressed as mg l -1 was not varying among Synechocystis substrains (Fig H in S1 File), following similar cellular concentration in all cultures (Fig A in S1 File). Content of intracellular polysaccharides (glycogen) under the same conditions was also the highest  Fig  4C). Similarly, content of total cellular carbohydrates under high red light was the lowest in substrain GT-B (802±168 fg cell -1 ) when compared to the other two substrains (1870-3011 fg cell -1 , Fig I in S1 File). Protein content under low and high red light (25 and 660 μmol(photons) m -2 s -1 ) was interestingly higher in substrain PCC-B (2516±147 fg cell -1 ) than in the other two substrains (1804-2178 fg cell -1 , Fig G in S1 File).
The above mentioned results suggests that substrain GT-L was able to utilize red photons with the best efficiency. On the other hand, substrain GT-K was clearly photoinhibited under high intensities of red light, which also affected its ability to accumulate sugars or proteins.
Blue light (405-485 nm). In order to test the effect of other than red wavelengths on Synechocystis growth, the three Synechocystis substrains were cultivated also under constant red light of 220 μmol(photons) m -2 s -1 (light intensities saturating growth of all substrains from at least 95% at 32˚C , Fig 2A) with addition of blue light of intensities 25-220 μmol(photons) m -2 s -1 . As shown in Fig 2C, none of the tested Synechocystis substrains were sensitive to blue light up to 220 μmol(photons) m -2 s -1 . On the contrary, growth of all tested Synechocystis substrains slightly increased after addition of blue photons to growth saturating red light.

Photons of wavelengths 485-585 nm and 670-700 nm
Additionally, the three Synechocystis substrains were cultivated under high intensities of white light. White light spectrum (405-700 nm) contained 27% of red photons (585-670 nm), 18% of blue photons (405-485 nm) and 55% of photons of other wavelengths (485-585 nm and 670-700 nm). To test effect of non-red and non-blue photons on Synechocystis growth, the red photons (as a combination of the whole spectrum of red light and a portion of white light spectrum) were kept on the constant level of 220 μmol(photons) m -2 s -1 (intensity saturating growth of all substrains from more than 95% at 32˚C , Fig 2A) by sequential stepwise increase of white light (up to 790 μmol(photons) m -2 s -1 ) followed by simultaneous stepwise decrease of red light (up to 8 μmol(photons) m -2 s -1 ). Particular red and white light combinations (as listed in Fig 2D) were set such that the blue photons (as part of the white light spectrum) were increasing up to 150 μmol(photons) m -2 s -1 and non-red and non-blue photons (485-585 nm and 670-700 nm) were increasing up to 435 μmol(photons) m -2 s -1 .
Growth of substrain GT-L was enhanced by photons of 485-585 nm and 670-700 nm, whereas growth of substrains GT-B and PCC-B was inhibited under these wavelengths ( Fig  2D). We performed again a series of additional measurements to reveal nature of the white light sensitivity of substrains PCC-B and GT-B. The maximal photosynthetic capacity of both substrains PCC-B and GT-B slightly decreased under high white light (526 μmol(photons) m -2 s -1 of white light complemented by 72 μmol(photons) m -2 s -1 of red light, the white light was containing 290 μmol(photons) m -2 s -1 of wavelengths 485-585 nm and 670-700 nm, Fig 3). On the contrary, photosynthetic capacity of substrain GT-L slightly increased under high white light. Content of cellular carbohydrates under high white light, (both total carbohydrates and glycogen) followed the same trends as under high red light-the lowest concentrations were measured in substrain GT-B (774±55 fg(total saccharides) cell -1 and 146±30 fg(glycogen) cell -1 , respectively) when compared to the other two substrains (1368-2691 fg(total saccharides) cell -1 and 398-828 fg(glycogen) cell -1 , respectively, Figs I and J in S1 File). Interestingly, Synechocystis substrain PCC-B was able to increase content of cellular saccharides under high intensities of white light even though its growth and photosynthesis were both inhibited under these conditions (Figs I and J in S1 File).
To test whether the photons of 485-585 nm and 670-700 nm had an instant photoinhibition effect, we cultivated substrain PCC-B under low intensities of white light. Addition of 56 μmol(photons) m -2 s -1 of white light (which contained only 31 μmol(photons) m -2 s -1 of 485-585 nm and 670-700 nm photons) already caused inhibition of growth rates of this substrain, and with stepwise increase of white light intensity, the photoinhibition effect was intensifying (Fig K in S1 File). The immediate growth inhibition of the substrain PCC-B under the photons of wavelengths 485-585 nm and 670-700 nm also suggests distinct photoinhibition mechanism under these photons and red photons.
Synechocystis sensitivity to suboptimal temperature Low temperature. Growth of the three Synechocystis substrains was also tested at temperatures below and above the growth optimum. During the low temperature experiments, Synechocystis cultures were cultivated in the quasi-continuous regime in a temperature range 23-35˚C (Fig 5A) under 220 μmol(photons) m -2 s -1 of red light (with addition of 25 μmol(photons) m -2 s -1 of blue light) as a light intensity that saturated growth of all Synechocystis substrains from more than 95% at 32˚C (Fig 2A). Optimal growth temperature of 35˚C was identical for all tested Synechocystis substrains. However, Synechocystis GT-B was more sensitive to temperatures below the growth optimum (23-32˚C) than the other two substrains. At 32˚C, where growth of both substrains GT-L and PCC-B was already saturated from more than 93%, the substrain GT-B reached only 84% of growth rates maximum. This is consistent with the observation that under 220 μmol (photons) m -2 s -1 of red light (with addition of 25 μmol(photons) m -2 s -1 of blue light) and at 32˚C, growth rates of substrain GT-B were reduced when compared to the other two substrains (Fig 2A). Sequential stepwise temperature reduction from 32˚C to 23˚C resulted in very similar response of substrains GT-L and PCC-B: growth rates of both these substrains were decreasing with similar temperature coefficients Q 10 1.69 and 1.76, respectively. However, the temperature coefficient Q 10 of substrain GT-B was measured as 2.90 (Fig 5A), which indicates increased sensitivity to low temperature in this substrain.
High temperature. Additionally, the effect of temperature higher than the growth optimum on Synechocystis growth was tested. At first, all three Synechocystis substrains were cultivated in the quasi-continuous regime at 38˚C, where only minimal variability in growth rates was observed (Fig 5A). At higher temperatures it was not possible to cultivate Synechocystis in the quasi-continuous regime anymore. Therefore, at higher temperatures, the cultures were cultivated in a batch regime at 32˚C and only short term (four hours) heat shock was introduced by increasing temperature from 32˚C to 52˚C or to 60˚C, respectively, once the cultures reached optical density OD 680 0.7 (approximately 4 x 10 7 cells ml -1 ).
All three Synechocystis substrains were able to recover from both four hours heat shocks at 52˚C and 60˚C. After the 52˚C treatment, all tested substrains recovered growth after c. 72 hours (Fig 5B). After the 60˚C treatment, the substrain GT-L recovered growth after c. 86 hours, whereas the other two substrains recovered growth after c. 100 hours (Fig 5C).

Synechocystis sensitivity to salt stress
Synechocystis was also cultivated in a batch regime in the presence of 0.5 M NaCl in the cultivation medium BG11. Synechocystis substrains GT-B and PCC-B showed only slight differences in growth in standard BG11 medium and in BG11 enriched with NaCl (Fig 6). On the and after introduction of heat shock of 52˚C (B) and 60˚C (C) during the batch growth. A-The Q 10 coefficient was calculated according to [19], the dashed lines represent linear fitting of the data points by the least squares method in the temperature range 23˚C-32˚C (23˚C-35˚C in case of the substrain GT-B). Each contrary, substrain GT-L was not able to grow in the presence of NaCl. Concerning overall robustness of GT-L substrain, this was quite surprising since this substrain was not sensitive to any previously tested condition, including high intensities of red, blue and white light and low and high temperatures. On the contrary, under high light, this substrain showed significantly better fitness than the other two tested substrains.
The sensitivity of substrain GT-L to the presence of NaCl in the cultivation media was further investigated. Salt stress in cyanobacteria affects, among other processes, homeostasis of inorganic ions [24]. Therefore, in order to test growth restoration in the presence of NaCl, growth of Synechocystis GT-L was tested in the batch regime in BG11 media enriched with MgSO 4 , K 2 HPO 4 , and CaCl 2 (or with combinations of these nutrients). Full grow curve of Synechocystis GT-L in the presence of NaCl was restored after addition of CaCl 2 together with K 2 HPO 4 in the concentration five times higher than in standard BG11 medium (Fig L in S1 point represents average of at least four independent experiments, error bars represent standard errors. B, C -Batch growth was performed at 32˚C and the heat shock was introduced by increasing temperature for four hours after the cultures reached OD 680 0.7 (the temperature record is marked by dashed line). The batch experiments were performed at least in three biological replicates with quantitatively similar results, data from one representative experiment are shown.
https://doi.org/10.1371/journal.pone.0189130.g005 File). However, growth in the quasi-continuous regime always ceased soon after the first dilution by the cultivation media enriched by salts. Therefore, to reach stable growth in the quasicontinuous regime, it was necessary to cultivate Synechocystis GT-L in full 5xBG11 medium that contained each single nutrient at concentrations five times higher than regular BG11. Even in 5xBG11, growth rates of Synechocystis GT-L in the presence of NaCl reached only 60% of the growth rates as in standard BG11 medium (0.048±0.002 h -1 and 0.079±0.011 h -1 , respectively, Fig M in S1 File).

Discussion
All Synechocystis substrains were not sensitive to the presence of glucose in both liquid and solid cultivation media during mixotrophic cultivation on light. Both GT-L and GT-B substrains were expected to be glucose-tolerant since they represent relatives of the first derived glucose-tolerant Synechocystis substrain "GT" (Williams (1988) [16], Fig 1). However, Synechocystis 6803 in Pasteur Culture Collection was reported previously to be glucose sensitive [7], even though glucose tolerance was observed in another substrain derived from Pasteur Culture Collection, PCC-M [10]. The newly discovered glucose tolerance of the PCC-B substrain reported in this study can be related to a spontaneous glucose tolerance switch as already described previously [7]. The ability of mixotrophic growth in Synechocystis is a complex process that can be defined by several genes (such as pmgA [25], infA or glcP [10]), transcription factors (such as Sll0822 [10]) but also by post-translational modifications or modulation of enzyme activities [26]. To reveal nature of glucose tolerance in the substrain PCC-B, further investigation is needed that would take into consideration these factors.
Under modest cultivation conditions (low light, temperature 32˚C and increased carbon concentration), only small differences in almost all measured parameters were detected among the tested Synechocystis substrains, including growth rates (Fig 2), photosynthesis (Fig 3, Fig E  in S1 File), content of cellular saccharides, cell size and cellular dry weight (Fig 4, Fig I and J in S1 File), CO 2 fixation (Fig D in S1 File), and pigments content (Fig F in S1 File). This is mostly consistent with results of the previous study where only minor differences in growth rates, oxygen evolution and total chlorophyll levels were detected among Synechocystis substrains GT-Kazusa, GT-O1 and PCC-Moscow under "standard" laboratory conditions (low light, 30˚C, ambient CO 2 ) [15]. On the other hand, even under modest conditions as used in this study, content of cellular proteins was elevated in substrain PCC-B (Fig G in S1 File). Similarly, variations in cell size, whole-cell absorption and 77 K fluorescence was reported previously under "standard" conditions [15]. This suggests that some physiology variation among Synechocystis substrains can be expected even under "standard" conditions, however, the phenotypic differences as detected in this study were much more pronounced under intensive and stress cultivation conditions (such as high light, low temperature or in the presence of salt in the cultivation medium).
In general, the most sensitive Synechocystis substrain identified in this study was the substrain GT-B. Growth rates of this substrain were inhibited under high intensities of red light (585-670 nm) and even under low intensities of non-red and non-blue part of white light spectra (485-585 nm and 670-700 nm) after addition of these photons to growth saturating red light (similarly as PCC-B substrain, Fig 2). Growth reduction of GT-B during sequential stepwise temperature reduction from 32˚C to 23˚C was also pronounced more than in the other two tested substrains (Fig 5). Under high red light, Synechocystis GT-B reduced photosynthetic capacity (Fig 3), and the changes in cell size, dry weight and content of both total and storage carbohydrates were much less pronounced when compared to the other two tested substrains (Fig 4, Figs I and J in S1 File). In addition to other changes, the S-M fluorescence rise was reduced in this substrain under high red light (Fig 3). The S-M rise was shown previously to be connected with state 2 to state 1 transition and it was also suggested that it reflects a protective mechanism for excess energy dissipation [27]. Based on reduction of the fluorescence rise during the S-M transition in substrain GT-B, it can be expected that the decreased efficiency of red photons utilization (as reflected by the previously mentioned GT-B characteristics) was connected with reduced capacity of energy transfer from phycobilisomes antenna to photosystem II and/or with reduced photoprotection capacities in this substrain.
On the contrary, Synechocystis GT-L and PCC-B used red photons with better efficiency when compared to GT-B. Under high red light, growth of Synechocystis GT-L and PCC-B was not inhibited and these substrains also accumulated higher amount of carbohydrates in the cells (Fig 4 and Fig I in S1 File). Content of glycogen in Synechocystis was determined previously up to 11.1±0.3% (of cellular dry weight) after 15 days of cultivation in another glucose tolerant substrain [28], or up to 67.5±3.7 μg(glycogen) μg(chl) -1 after 48 hours under nitrogen starvation [29]. In this study, substrain GT-L was able to accumulate up to 15.8±2.5% or 15.9 ±1.8 μg(glycogen) μg(chl) -1 of glycogen during exponential growth. The other two tested substrains accumulated always below 10% of glycogen in the cells. Even though the substrain GT-L looks as promising candidate substrain for glycogen production, comparison with the previous studies is difficult since the content of glycogen was strongly dependent on light availability (Fig 4 and Fig J in S1 File) that was lower in the previous studies [28,30].
The tested Synechocystis substrains varied also in growth rates (Fig 2), photosynthetic rates (Fig 3), cellular dry weight (Fig 4), proteins content (Fig G in S1 File) and carbohydrates content (Fig 4, Figs I and J in S1 File). Such variation on Synechocystis substrain level is reported for the first time here. The overall carbohydrates content (up to c. 40% of dry weight for GT-L substrain under high red light) was higher than reported previously in another Synechocystis substrain (up to c. 13%, [31]). On the other hand, cellular protein content (32-57% of cellular dry weight for all three Synechocystis substrains tested here) was lower than reported previously (61-73%, [31]), even though the photosynthetic rates measured here were comparable with the previous study (0.03-0.21 mol(O 2 ) mol(Chl -1 ) s -1 in this study vs. 0.066-0.081 mol (O 2 ) mol(Chl -1 ) s -1 in Touloupakis et al. (2015) [31]). We also observed differences in cell size. Substrain GT-L that was having cells of the biggest size under high red light (the difference in cell size between GT-L and GT-B under the highest tested red light was 48%). Differences in cell size were observed previously among Synechocystis substrains GT-O1 and GT-Kazusa and PCC-Moscow [15]. The cell size variation as observed here is consistent with variations in the content of cellular saccharides and cellular dry weight (Fig 4, Figs I and J in S1 File).
The observed phenotypic variation shows that the effectivity of cellular energy management can vary among Synechocystis substrains, which can lead to substantial changes in cellular composition. Therefore, considering Synechocystis sp. PCC 6803 as a candidate strain for optimization of synthesis of targeted compounds (such as glycogen or proteins, but probably also for other products) can be tricky since 1) some Synechocystis substrains are more sensitive to specific cultivation conditions than others, 2) some Synechocystis substrains can produce specific compounds with higher rates than others, and 3) it cannot be excluded that Synechocystis substrains can "spontaneously" change phenotype in time.
Growth rates of all Synechocystis substrains increased only slightly after addition of blue photons along to the saturating red light (Fig 2). This fact can be connected with activation of non-photochemical quenching in Synechocystis under blue light [32]. On the other hand, addition of non-blue and non-red photons (485-585 nm and 670-700 nm) to saturating red light inhibited growth of substrains GT-B and PCC-B but increased growth rates of substrain GT-L (Fig 2). Interestingly, the fluorescence induction had similar trends as under red light only (Fig 3), which suggests distinct mechanisms of growth inhibition under red photons and photons of wavelengths 485-585 nm and 670-700 nm. Since addition of even small amount of non-blue and non-red photons resulted in reduction of growth rates in substrain PCC-B ( Fig  K in S1 File), it can be expected that sensitivity to these wavelengths had a signaling character. Light quality has been previously connected with stress responses activation in Synechocystis [33]. Under increasing light intensities, Synechocystis can be expected to change intracellular redox state or increase formation of reactive oxygen species. Both these processes can then act as stress response triggers [34]. Moreover, it has been shown that genotypic differences among Synechocystis substrains can include modification of various histidine kinases [9][10][11]13]. Changes of stress signal transduction cascade together with "activated" stress response under high light can then have diverse consequences. Although further experiments are needed to map probable variations in the signaling network among Synechocystis substrains, it is demonstrated here that sensitivity to specific wavelengths can vary in Synechocystis. This should be taken into consideration in studies that address effect of light on Synechocystis physiology or in studies and applications that aim on phototrophic production optimization.
Growth rates of substrain GT-B between 23˚C-32˚C were more reduced in comparison to the two other two substrains (Fig 5). Synechocystis sensitivity to cold stress was previously connected with saturation status of membrane lipids [35]. However, fatty acids composition was almost identical for all tested Synechocystis substrains at both temperatures 32˚C and 23˚C (Table B in S1 File). The fatty acids ratio at both 32˚C and 23˚C was similar as reported previously, even though some fatty acids (including 16:0, 16:1, 18:0, 18:1, 18:2, 18:3 and 18:4) were detected in slightly different concentrations when compared to the previous studies [35][36][37]. Change of membrane fluidity is not the only way how cyanobacteria can react to low temperatures-also formation of reactive oxygen species, changes in DNA supercoiling or cellular redox status can be detected as part of cyanobacteria cold stress response [38]. Spectrum of cold-induced genes is broad and it covers genes involved in signal perception and transduction, transcription and translation, RNA-binding proteins genes, and some genes of other functions. Moreover, several signaling proteins were described to be involved in cyanobacteria cold stress response, including universal stress histidine kinases such as Hik33. This protein can sense also red light [39] and it was even suggested that it can work as an additional red light sensor [33]. From this perspective, it can be expected that the increased sensitivity to low temperature of substrain GT-B (Fig 5) is connected with early red light photoinhibition of this substrain, with Hik33 being involved as a common element in both stress responses. This is consistent with our suggestion of reduced photoprotection in substrain GT-B under high red light.
Under high temperatures (52˚C and 60˚C) it was not possible to cultivate Synechocystis cultures in the quasi-continuous regime anymore and the heat stress was applied only for period of four hours during the batch cultivation (Fig 5). It has been previously shown that Synechocystis can survive heat shock of 44˚C applied for 24 hours [40] or even for four days [41]. However, it was also reported that even five minutes of incubation at 52˚C and 54˚C is lethal for Synechocystis cells [42], even though in another study Synechocystis was found to survive for 30 minutes at 50˚C [43]. Our finding of successful growth recovery of all three tested substrains from four hours heat shock at both 52˚C and 60˚C are not in contradiction with the previous findings. In fact, the cultures recovery after the 5 min heat shock in the previous study at both 52˚C and 54˚C was tested only for up to 35-40 hours [42], whereas here, the cultures recovery was followed for at least 72 hours. Successful growth recovery after 60˚C heat shock as reported here shows for the first time that Synechocystis is able to survive such high temperature treatment.
Interestingly, the substrain GT-L (which was in general the most robust substrain under high lights and suboptimal temperatures, Figs 2-5) was not able to grow under increased salinity in standard BG11 medium (Fig 6). Long-term salt stress (in range of hours or days) in Synechocystis was reported to initiate production of compatible solutes, alteration in fatty acids composition [44] and ion transport [45], induction of specific genes [46], and accumulation of specific proteins [47]. Successful growth recovery in salt media after addition of both CaCl 2 and K 2 HPO 4 (Fig L in S1 File) to the substrain GT-L culture suggests that the ions management was affected under increased concentration of NaCl-both calcium and potassium ions participate in the salt stress response in cyanobacteria [24,48], and calcium was even found to be released from the cells as a consequence of membrane depolarization [49]. The finding that salt tolerance in Synechocystis (which was originally isolated from fresh water lake) can vary on the substrain level can have consequences for large scale production systems that consider use of sea water for cyanobacteria cultivation.
To conclude, the results presented in this study bring new insight into our understanding to Synechocystis sp. PCC 6803 stress sensitivity and tolerance, and undertake strongly the previous recommendation of taking into consideration specific lineages of this model strain. Based on the distribution of stress sensitivity among tested Synechocystis substrains, it is unlikely that stress tolerance is conserved in each specific Synechocystis lineage. It can be rather expected that the stress sensitivity can change spontaneously in Synechocystis, similarly as glucose tolerance or motility. Mechanisms of stress sensitivity "switch" on the molecular level still remain to be understood. However, the strong phenotypic variation on the substrain level represents important component of Synechocystis complexity that should be taken into account for keeping this model strain robust for the diverse portfolio of already started research activities, and also for its utilization in industrial applications.