Light Enhances Survival of Dinoroseobacter shibae during Long-Term Starvation

Aerobic anoxygenic phototrophs (AAPs) as being photoheterotrophs require organic substrates for growth and use light as a supplementary energy source under oxic conditions. We hypothesized that AAPs benefit from light particularly under carbon and electron donor limitation. The effect of light was determined in long-term starvation experiments with Dinoroseobacter shibae DFL 12T in both complex marine broth and defined minimal medium with succinate as the sole carbon source. The cells were starved over six months under three conditions: continuous darkness (DD), continuous light (LL), and light/dark cycle (LD, 12 h/12 h, 12 µmol photons m−2 s−1). LD starvation at low light intensity resulted in 10-fold higher total cell and viable counts, and higher bacteriochlorophyll a and polyhydroxyalkanoate contents. This coincided with better physiological fitness as determined by respiration rates, proton translocation and ATP concentrations. In contrast, LD starvation at high light intensity (>22 µmol photons m−2 s−1, LD conditions) resulted in decreasing cell survival rates but increasing carotenoid concentrations, indicating a photo-protective response. Cells grown in complex medium survived longer starvation (more than 20 weeks) than those grown in minimal medium. Our experiments show that D. shibae benefits from the light and dark cycle, particularly during starvation.

Comparable to purple bacteria, AAPs are capable of lightdriven and respiratory electron transport for their energy metabolism. However, this activity is different in several aspects. AAPs use light energy under oxic conditions and contain much less bacteriochlorophyll a (BChl a), with sometimes higher carotenoid than BChl a concentrations [18]. Increased amounts of carotenoids may possibly be a protective response against reactive oxygen species (ROS), which are produced in the light under oxic conditions [19]. As the biosynthesis of BChl a is known to be sensitive to ROS, AAPs synthesize BChl a only in the dark [20][21] and require a day and night cycle to sustain their photosynthetic activity. In accordance with the low BChl content, light is only used as a supplementary energy source [22][23]. Furthermore, AAPs are not autotrophs, but heterotrophs. Their photosynthetic activity is not used for CO 2 fixation, but prevents the oxidation of organic substrates.
Light-driven proton translocation and ATP formation by AAPs have been investigated in several studies [24][25]. Recently, Tomasch et al. [26] studied the transcriptional response of D. shibae under different light regimes. The substrate-saving effect of light energy was demonstrated by lower respiration rates in the light [23,25]. In chemostat cultures, the light-dependent increase of growth yields [27], [28] was reversely correlated to the growth rate and increased at low rates up to 110% [29].
The purpose of the present study was to identify the conditions under which AAPs benefit the most from their photosynthetic capacities. We hypothesized that AAPs benefit from photon energy under conditions of carbon and electron donor limitation, as proteomic responses to starvation and light conditions among AAPs have been reported [30]. Here, we performed long-term starvation experiments under different light regimes, and investigated survival and physiological fitness of Dinoroseobacter shibae [31] a representative of the globally abundant marine Roseobacter clade [32][33][34].

Cultivation and Starvation Conditions
Batch cultures were incubated at 23uC on a shaker (Innova 42-R, New Brunswick, 125 rpm) in the dark (DD), under continuous illumination (LL; 12 mmol photons m 22 s 21 ) or light and dark cycles (LD, 12 h/12 h, 12 mmol photons m 22 s 21 ). For determining the optimum light intensity, GRO-LUX fluorescent light bulbs were used as light source. Growth was monitored by measuring the optical density (OD) at 650 nm. For starvation experiments, stationary phase cells were harvested by centrifugation at 33306g in a Sorvall RC-2 refrigerated centrifuge for 15 min at 5uC and washed with salt solution containing 20 g NaCl and 0.5 g KCl per liter. Cells were resuspended in salt solution (sea water medium without carbon source 250 ml in 500 ml sterile Erlenmeyer flasks) and incubated as described above.

Cell Counts and Viability
Total cell counts were analyzed by SYBR Green staining and epifluorescence microscopy (Olympus BX51). For live counts, serially diluted samples were plated onto MB agar plates incubated at 25uC and counted after 4-8 weeks. To study morphological changes under different conditions, suspensions of starved cells were placed on a Formvar copper grid (Plano) for 5 min. Adsorbed cells were stained with 0.5% aqueous uranyl acetate for 1 min followed by a distilled water rinse and examined with a Zeiss EM 902A transmission electron microscope. A Proscan High Speed SSCCD camera system with iTEMfive software was used for image acquisition. At least fifty pictures were analyzed to compare changes in cell morphology.

Estimation of Dry Biomass, Protein Content, PHA and Photosynthetic Pigments
To determine the dry biomass, starved cells were harvested and washed with 50 mM ammonium acetate buffer and dried overnight at 80uC. Lowry's method with Folins reagent was used to determine the total protein concentration [36]. The in vivo photo pigments were analyzed by recording the absorption spectrum of whole cells in a UV/VIS spectrophotometer (Perkin Elmer, Lambda 2S) with a resolution of 1 nm from 350 to 950 nm. Polyhydroxyalkanoate (PHA) content was determined using Nile blue staining [37]. For pigment analysis, cells were centrifuged at 6000 g for 20 min and pigments were extracted from the pellet with 1 ml acetone:methanol (7:2) for 1 hour in the dark. BChl a absorption was determined at 772 nm with an extinction coefficient of 75 mM 21 cm 21 [38]. Carotenoids were quantified at 482 nm using an extinction coefficient of 123.6 mM 21 cm 21 [39].

Determination of Respiration Rates
Washed cells were resuspended in HEPES buffer (10 mM, pH 7.75, supplemented with NaCl 20 g per liter, KCl 0.5 g per liter). Oxygen concentrations were measured with a Clark-type oxygen electrode (Bachofer, Reutlingen, Germany) while the chamber was maintained at 30uC. The influence of light on the respiration was checked by illuminating the cells in the reaction chamber using a halogen lamp at 400 mmol photons m 22 s 21 .

Determination of ATP
Washed cell suspensions (3 ml) were incubated in Hungate tubes sealed with rubber stoppers and flushed with nitrogen at room temperature. The physiological responses were analysed by incubation of cell suspensions under anoxic conditions and studying the reaction upon oxygen and organic substrate addition in the dark and in the light. The energy content of the cells was determined using ATP bioluminescence Assay Kit CLS II (Boehringer Mannheim) and the extraction of ATP was carried out as described in [25].

Proton Translocation Measurements
Washed cells were resuspended in 2.5 ml salt solution containing 2% NaCl and 0.05% KCl and treated with 500 ml of 0.5 M KSCN, thus providing the membrane-permeable anion SCNin order to destroy the membrane potential. The cell suspensions were then flushed with nitrogen for 20 min. Known amounts of oxygen (16 nmol O 2 ) in KCl solution, were injected into the measuring chamber equipped a with pH electrode (Mettler Toledo pH Electrode, Inlab Micro).

Optimum Light Intensity for Starvation Experiments
Cells pre-grown with succinate at different light intensities were harvested, washed, resuspended in medium without carbon source, and incubated under light and dark cycles (LD, 12 h/ 12 h) for four weeks. Different light intensities were tested to determine the optimum illumination for survival. Under all conditions, total and viable cell counts decreased However, cultures incubated at medium light intensity (12 mmol photons m 22 s 21 ) showed higher survival rates (Table 1), reaching tenfold higher cell counts than those incubated at high (23 mmol photons m 22 s 21 ) or low (3 mmol photons m 22 s 21 ) light intensities. Protein to biomass ratios did not show any significant differences between the light conditions. In contrast, at medium and low light intensities carotenoid concentrations decreased whereas bacteriochlorophyll a concentrations remained constant. At high light intensity, bacteriochlorophyll a concentrations decreased, while carotenoid concentrations increased by a factor of 5 indicating their photo-protective function.

Morphological and Physiological Adaptations upon Nutrient Limitation
Starvation experiments were performed at the optimum light intensity (12 mmol photons m 22 s 21 ) under different light regimes, i.e. light and dark cycle (LD), continuous light (LL), and continuous dark (DD). LD conditions resulted in the highest survival rates. During the first week of starvation, total and viable cell counts of the DD cultures decreased faster than those in LL and LD cultures (Figure 1). After four weeks of starvation, LD cultures resulted in 10-fold higher total and viable cell counts. Cell morphology was affected both in LL and DD cultures, as cells appeared irregular and wrinkled and flagella were detached anymore ( Figure 2). In contrast, cells in LD cultures had a less irregular shape, and some cells still carried flagella after four weeks. Nile-blue staining showed the presence of polyhydroxyalkanotes (PHA) during the first days of starvation ( Figure S1). After four weeks, PHAs disappeared in LL and DD cells, but were still visible in LD cells.
The physiological fitness of cells was assessed by respiration rates, respiration-driven proton translocation, and the ability of ATP regeneration. Respiration rates served as a measure of physiological activity and light utilization. The slowdown of respiration in the light indicates saving of electron donors and light-driven energy conservation. Independent of the starvation conditions, endogenous respiration rates were 10% higher in the   Table 2). Utilization of light was also shown in proton translocation measurements, when oxygen and light were supplied simultaneously to cells incubated under N 2 ( Figure 3 a and b). The highest ratio of translocated protons per added oxygen atom (H + /O) in the dark was 2.3, independent of the starvation conditions. When cell suspensions were additionally illuminated and charged with oxygen pulses (16 nmol), the H + /O ratios increased to 2.8. After two weeks of starvation, H + /O ratios of LL cells had dropped by a factor of 4, while only decreasing by half in LD cells.
The cytoplasmic ATP concentrations showed a steady decrease upon starvation (Table 3). Initial ATP concentrations were 3 * 10 218 mol ATP per cell, decreasing by 90% in both LL and LD cells after 3 weeks. When the cells were washed and incubated anoxically, the ATP levels decreased close to zero within 2 hours. With the addition of oxygen and light, ATP was regenerated by more than 50% within two minutes.

Survival of Cells Grown in Complex Compared to Defined Medium
Under DD conditions, total cell counts in complex media were stable over six weeks ( Figure S2), whereas with succinate they decreased more than tenfold after 1 month. Unfortunately, the light intensity for LD and LL conditions was high (53 mmol photons m 22 s 21 ) since these experiments were carried out prior to the determination of the optimum light intensities. Here, cell counts declined faster under LD conditions than in the dark, which underlines the detrimental effects of high light intensity (Table S1, Table S2 and Table S3).

Discussion
Our study demonstrated that day and night cycle substantially increases the survival of Dinoroseobacter shibae during long term starvation. However, high light intensity may be detrimental and induce protective reactions.

The Diurnal Light and Dark Regime Increases Cell Fitness
All measured parameters show that Dinoroseobacter shibae benefits from exposure to a diurnal light and dark regime. Under this condition, the cells survived starvation with better physiological fitness. As cells were shown to preserve intracellular PHA, known to be the main storage material in D. shibae [28], cells maintained increased dry mass to protein ratios. Cell pigmentation, morphology and motility of cells grown under the diurnal light cycle were less affected than those in the dark or under continuous light. Viability and fitness of cells decreased in similar manner both in the dark and under continuous light, which implies that light inhibits BChl a synthesis and prevents photosynthetic activity over longer periods. The increased fitness of LD-starved cells was also found by quantifying respiration, proton translocation and ATP regeneration. The beneficial effects became more pronounced with prolonged starvation. These observations could be explained by one underlying rationale: Light utilization appears rather as a mechanism to survive starvation than as a growth-promoting factor, which improves survival and physiological fitness severalfold.

Light as a Stress Factor
Not only continuous light, but also high intensities during day and night cycles had negative effects on the cells. The harmful light effects were also recognized from the pigment analysis, as cells increased their carotenoid contents at high light intensities likely as a protective response. By low pigment concentrations, AAPs might prevent formation of reactive oxygen species [19]. This was corroborated by the low optimum light intensity of 12 mmol photons m 22 s 21 , which might additionally minimize ROS exposure. The optimum light intensity of 12 mmol photons m 22 s 21 is much lower than mid-day intensities in, e.g., the North Pacific Gyre, which average 50-150 mmol photons m 22 s 21 at 20-150 m water depth [40]. Thus, low-light adaption of Dinoroseobacter shibae minimizes ROS exposure and might explain why none of the AAPs is able to grow purely phototrophic.

Ecological Implications
The light-supported survival of starvation of D.shibae has also been reported for other groups of bacteria. In a bacteriochlorophyll a-containing gammaproteobacterium, expression of the photosynthesis genes depended on the type of carbon source [41]. Also some proteorhodopsin-carrying bacteria such as the abundant Cand. Pelagibacter ubique [42], Dokdonia sp. strain MED134 [43] and Vibrio sp. strain AND4 [44], benefit from light during starvation.
Considering the specific advantage of the AAPs in their natural environment, it becomes clear that light is not the overall limiting factor for their distribution. Cottrell et al. found that AAPs are distributed over the whole photic zone [40]. In a particle-rich estuary, the BChl a concentrations of AAPs varied in response to particles but not to light limitation [45]. In a recent study, Č uperová et al. showed that the DOC concentrations influenced the AAP abundance in alpine lakes [46]. In agreement with these findings, our study suggests that the limitation of organic substrates might promote the competitiveness of AAPs. Nutrient limitation is a predominating feature in most oceanic regions. Biomass of AAPs in the South Pacific Ocean was on average two-fold higher than that of other prokaryotic cells [11,16]. Obviously, AAPs can profit from light utilization by conserving instead of oxidizing organic substrates or their storage compounds.