Cell Cycle Regulation of the Mixotrophic Dinoflagellate Dinophysis acuminata: Growth, Photosynthetic Efficiency and Toxin Production

Abstract

The mixotrophic dinoflagellate Dinophysis acuminata is a widely distributed diarrhetic shellfish poisoning (DSP) producer. Toxin variability of Dinophysis spp. has been well studied, but little is known of the manner in which toxin production is regulated throughout the cell cycle in these species, in part due to their mixotrophic characteristics. Therefore, an experiment was conducted to investigate cell cycle regulation of growth, photosynthetic efficiency, and toxin production in D. acuminata. First, a three-step synchronization approach, termed “starvation-feeding-dark”, was used to achieve a high degree of synchrony of Dinophysis cells by starving the cells for 2 weeks, feeding them once, and then placing them in darkness for 58 hours. The synchronized cells started DNA synthesis (S phase) 10 hours after being released into the light, initiated G2 growth stage eight hours later, and completed mitosis (M phase) 2 h before lights were turned on. The toxin content of three dominant toxins, okadaic acid (OA), dinophysistoxin-1 (DTX1) and pectenotoxin-2 (PTX2), followed a common pattern of increasing in G1 phase, decreasing on entry into the S phase, then increasing again in S phase and decreasing in M phase during the diel cell cycle. Specific toxin production rates were positive throughout the G1 and S phases, but negative during the transition from G1 to S phase and late in M phase, the latter reflecting cell division. All toxins were initially induced by the light and positively correlated with the percentage of cells in S phase, indicating that biosynthesis of Dinophysis toxins might be under circadian regulation and be most active during DNA synthesis.

1. Introduction

A least twelve species within the genus Dinophysis are known to synthesize toxins associated with diarrhetic shellfish poisoning (DSP) events worldwide (Reguera et al., 2014; Reguera et al., 2012). These toxins, i.e., okadaic acid (OA) and dinophysistoxins (DTXs) can accumulate in filter-feeders and become a potential threat to public health and the viability of shellfish fisheries (Suzuki et al., 2001a; Suzuki et al., 2001b). Pectenotoxins (PTXs), although not directly involved in the diarrhetic syndrome, are also considered toxins through their alteration of the actin-based cytoskeleton (Fernandez 2014), and often co-exist with OAs and DTXs. The toxin profile and content of different geographical populations and/or isolates of Dinophysis species. differ greatly, leading to inter-regional variation in the impacts of DSP outbreaks worldwide (Campbell et al., 2010; Fux et al., 2010; Gao et al., 2017; Hackett et al., 2009; Kamiyama and Suzuki, 2009; Suzuki et al., 1998; Tong et al., 2015b).

Studies of Dinophysis toxin production have flourished since the discovery that D. acuminata can be successfully maintained in the laboratory by providing the ciliate Mesodinium rubrum as a food source (Park et al., 2006). A number of external factors have been considered critical in controlling the toxin production of Dinophysis cells: temperature (Gao et al., 2017; Kamiyama et al., 2010; Tong et al., 2010), dissolved nutrients (Hattenrath-Lehmann and Gobler, 2015; Hattenrath-Lehmann et al., 2015; Nagai et al., 2011; Tong et al., 2015a), light intensity (Nielsen et al., 2012; Nielsen et al., 2013; Tong et al., 2011), food availability and quality (Gao et al., 2017; Kim et al., 2008; Nielsen et al., 2012; Nielsen et al., 2013; Smith et al., 2018; Tong et al., 2015b) and even the co-occurring bacterial community (Gao et al., 2018; Hattenrath-Lehmann and Gobler, 2017). Dinophysis toxin production typically varies during different growth stages, with the highest production rates occurring in the exponential phase and attainment of the maximum cellular toxin content upon entering stationary phase (Smith et al., 2012; Tong et al., 2011; Tong et al., 2015a). Furthermore, specific toxin production rates were generally close to the specific growth rate of Dinophysis during exponential growth (Basti et al., 2018; Smith et al., 2018; Tong et al., 2011), suggesting that toxin synthesis may be linked to the cell cycle.

Knowledge of cell cycle regulation of dinoflagellate metabolism and growth is fundamental to understanding the dynamics of blooms and toxin production. Dinoflagellate cells follow a typical eukaryotic G1-S-G2-M cell cycle (Bhaud et al. 2000). The first growth stage is G1, in which growing cells exhibit high metabolic activity, followed by the S stage during which DNA synthesis occurs. The second growth stage is G2, a temporal gap between the S phase and the subsequent division in which cells accumulate the nutrients needed for the division process. M (mitosis) represents the stage during which nuclear division occurs, followed by cytokinesis (cell division; Pan et al., 1999). The DNA division cycle and intracellular trigger for the production of DSP toxins in Dinophysis cells have not been investigated, possibly due to the mixotrophic characteristics of these cells (Hansen et al., 2013; Park et al., 2006) and uncertainties associated with their life cycle (Escalera and Reguera, 2008; Koike et al., 2006; Reguera and González-Gil, 2001; Reguera et al., 2007).

Mixotrophic members of the Dinophysis genus can acquire and utilize chloroplasts of their ciliate prey that had previously been obtained from a cryptophyte microalga prey (Hansen et al., 2013). This feeding process and the use of its kleptochloroplastid by Dinophysis make it difficult to isolate the role of autotrophy and heterotrophy in toxin synthesis. Field studies have shown that the maximum frequencies of dividing Dinophysis cells were found at onset of the sunlight or ca. two hours after sunrise, similar to many other dinoflagellates (Campbell et al., 2010; Chang and Carpenter, 1991; Pizarro et al., 2008; Reguera et al., 2003). Reguera et al. (2003) reported that the highest frequencies of binucleated cells were detected at ca. one hour before lights went on and decreased sharply during the following 60 min, indicating that the Dinophysis cell cycle was associated with light: dark patterns. Field studies found that Dinophysis acuta attained a 3.5 times higher maximum OA content during the dark period than in daytime (Pizarro et al., 2008), suggesting that the biosynthesis of DSP toxins might be associated with the circadian rhythm of Dinophysis cells. Up to now, a lack of laboratory studies of toxin production during the cell cycle have not provided the detailed data needed to explore diel toxin dynamics of Dinophysis populations.

Synchronization of a dinoflagellate population is an important prerequisite to cell cycle experiments in the laboratory, especially for slow-growing dinoflagellates (Taroncher-Oldenburg et al., 1997). A homogeneous population can be achieved following release from external restrictive conditions (Pardee et al., 1978). Therefore, in the present study, several constrained approaches were employed to induce the synchronization of mixotrophic Dinophysis acuminata cells. A diurnal laboratory experiment was conducted to characterize the relationship among DNA synthesis, photosynthetic activity, and toxin production during the cell cycle of Dinophysis acuminata. Our focus is therefore on the asexual reproduction of Dinophysis. Results are discussed in the context of other synchronization studies in dinoflagellates, and a potential mechanism linking specific cell cycle stages to the production of toxin is proposed.

2. Materials and methods

2.1. Culture maintenance

The Dinophysis acuminata strain (DAYS01) was isolated from Xiaoping Island, Yellow Sea of China (121.53E, 38.83N) in July 2014 (Gao et al., 2017), and maintained in the laboratory with biweekly feeding on the ciliate Mesodinium rubrum (strain AND-A0711) from southern Spain (Rodríguez et al., 2012) in a 1:5 Dinophysis: Mesodinium ratio until concentrations of Dinophysis reached ~5,000 cells mL⁻¹. The ciliate was delivered weekly in a 1:1 ciliate: cryptophyte ratio by mixing cells from a stock culture of the ciliate Mesodinium rubrum (ca. 10,000 cells mL⁻¹) with those of the cryptophyte Teleaulax amphioxeia (strain AND-A0710 stock, ca. 150,000 cells mL⁻¹). All cultures were maintained in f/6-Si medium (i.e., all materials were one third of f/2-Si medium concentrations, Guillard and Ryther, 1962) at 15°C at a light intensity of 54 μmol photons m⁻² s⁻¹ and a 14h:10h light: dark cycle started at 6AM.

2.2. Synchronization Experiment

To obtain healthy and synchronized D. acuminata cells, dark exposure and/or starvation approaches were attempted. Three treatments were conducted: 1) feeding in the light, then dark for 58 h (D_ark); 2) starvation for two weeks, followed by feeding, and dark for 58 h (Starve_2w); 3) starvation for four weeks, followed by feeding, and dark for 58 h (Starve_4w). For the D_ark treatment, D. acuminata cells in early stationary phase were first inoculated into triplicate 3 L flasks and fed M. rubrum at a ratio of 100 Dinophysis cells: 2000 Mesodinium cells mL⁻¹. Once no prey remained in the medium, the D. acuminata culture, with the population in late exponential phase, was then maintained in the dark for 58 h. The durations of the experiments were limited to early stationary phase to avoid possible sexual reproduction which might happen when population enters late stationary or decline phases (Escalera and Reguera, 2008). In Starve_2w and Starve_4w treatments, both starvation and dark-restricted methods were used sequentially by allowing starvation for two or four weeks respectively, followed by the D_ark treatment. To collect the samples in the dark, a low-light (<30 μmol photons m⁻² s⁻¹) red flashlight (H-R6-A1, Aoliliya Company, Hangzhou, China) was used to provide illumination. All processing was completed within 5 minutes.

Triplicate samples were taken daily for cell counts during the feeding and dark periods at 08:00 and every two hours during the diel cycle study. Samples were preserved with 4% formalin (v/v) and enumerated in a Sedgewick-Rafter chamber using a light microscope.

The average growth rate of D. acuminata (μ) was calculated as:

$$\text{μ}=\frac{\text{ln}\left({\text{C}}_2/{\text{C}}_1\right)}{{\text{t}}_2-{\text{t}}_1}$$

where C₁ and C₂ are the cell concentrations (cells mL⁻¹) at times 1 and 2, respectively; t is the time (days) of exponential growth and μ (day⁻¹) is the average growth rate (Guillard, 1973).

2.3. Cell cycle study

DNA division, photosynthetic efficiency and intracellular toxin production of D. acuminata were determined every two hours for 24 h from the onset of the light (06:00) until the lights go on again within the three-step synchronization treatments (Starve_2w and Starve_4w). The light: dark cycle was set at 14: 10, therefore, dark was initiated at 20:00. All samples were taken in triplicate.

For determination of DNA division, 50 mL subsamples were collected by centrifugation under 2795 × g, 5min, 4°C (Sorvall ST 8R, ThermoFisher, Germany). The pellet was resuspended in 10 mL pre-cooled (4 °C) methanol (AR, 99.5%; SCR, Shanghai, China), and stored at 4°C for 12 h. Next, the cells were washed and resuspended in 1 ml PBS (phosphate buffered saline, pH = 7.1; Hyclone, Beijing, China) in a 2 mL microcentrifuge tube, and then washed and stained with 10 μg mL⁻¹ propidium iodide (PI; Aladdin, Shanghai, China) PBS solution containing 0.1 mg mL⁻¹ RNase A (Aladdin, Shanghai, China) and 0.1% Triton X-100 (Aladdin, Shanghai, China), and maintained at 37°C for 1 h. The PI-stained samples were stored overnight in the dark at room temperature for analysis using a CytoFLEX flow cytometer (Beckman Coulter Inc., USA) with the laser set at 488 nm for excitation, and 585/BP for fluorescence of the DNA-bound PI. The cell cycle profile of D. acuminata was analyzed using Modfit LT 5.0 software (Windows version, Verity Software).

For photosynthetic efficiency (F_v/F_m) analysis, 50 mL subsamples were concentrated by centrifuging at 2795 × g and 15°C for 5 min (Sorvall ST 8R, ThermoFisher, Germany). Next, 30 μL of the pellet were transferred into a 96-well plate, and exposed in the dark for 20 min to relax the reaction centers of photosystem II. Then, F_v/F_m of D. acuminata was analyzed using an IMAGING PAM fluorometer (WALZ, Effeltrich, Germany). The F_v/F_m ratio was assessed by fluorescence quenching analysis and application of a saturating light flash (saturation pulse), and computed using the following equations:

$${\text{F}}_{\text{v}}={\text{F}}_{\text{m}}-{\text{F}}_0$$

$${\text{F}}_{\text{v}}/{\text{F}}_{\text{m}}=\left({\text{F}}_{\text{m}}-{\text{F}}_0\right)/{\text{F}}_{\text{m}}$$

where F_v is the maximum variable chlorophyll-a fluorescence yields in a dark-adapted state. F₀ is the minimal fluorescence level when all antenna pigment complexes associated with the photosystem are assumed to be open. F_m is the maximal fluorescence level when a high intensity flash has been applied when all antenna sites are assumed to be closed (Kolber et al., 1998; Murchie and Lawson, 2013).

For intracellular toxin (OA, DTX1 and PTX2) analysis, subsamples of 15 mL were collected, each containing ~60,000 cells. Cells were separated from the medium using a 15-μm Nitex sieve, washed 3 times with fresh seawater and rinsed into a pre- weighed 15-mL centrifuge tube. The collected cells were enumerated by pipetting 0.5 mL of the mixed sample and 0.5 mL of filtered seawater with 4% v/v of formalin solution. The 15-mL tube was reweighed to determine the volume of harvested D. acuminata cells (sample weight divided by the density of seawater, 1.03 g mL⁻¹) and frozen at −20°C. For toxin extraction, cells were thawed at room temperature and sonicated with an Ultrasonic Homogenizer (JY92-IIN, SCIENTZ, China) for 10 min (6.5 s work / 3.5 s interval, 65W, 220V). Then all samples were passed through solid phase extraction (SPE) cartridges (Oasis HLB 60 mg; Waters, Milford, MA) pre-conditioned with 3 mL methanol and 3 mL Milli-Q water, followed by washing with 3 mL Milli-Q water. Toxins retained on the SPE cartridges were eluted with 1 mL methanol into 1.5-mL HPLC vials (SI-0715, IMT, Hangzhou, China). Finally, eluates were heated at 40°C and dried under a stream of N₂(N-EVAP 111, Organomation, USA) generated by a nitrogen evaporator (XYN-30, XiYou Analytical, Shanghai, China), and resuspended in 1 mL of methanol for toxin analysis.

Toxins were analyzed using liquid chromatography coupled with tandem mass spectrometry performed on a Dionex UltiMate 3000/API 4000 LC-MS/MS (Thermo ScientificTM DionexTM, Waltham, MA, USA) with electrospray ionization, using a Waters X-Bridge™ C18 column (3.0×150 mm, 3.5 μm particle size) at 40 °C for the negative mode and Waters X-Bridge™ C18 column (2.1×50 mm, 2.5 μm particle size) at 25 °C for the positive mode. The toxins OA and DTX1 were analyzed in negative mode, with two mobile phases: A) 0.05 v/v % ammonium hydroxide (NH₄OH) in water and B) 0.05 v/v % NH₄OH in 90% acetonitrile. Linear gradient elution was employed, starting with 10% B for 1 min, increasing to 90% B for 9 min, then held in 90% B for 3 min, 10% B for 2 min, and 10% B for 4 min. The flow rate was 0.4 mL min⁻¹. Pectenotoxin (PTX2) was analyzed in positive mode, consisting of phase C (water) and phase D (95% acetonitrile) both buffered with 2 mM ammonium formate and 50 mM formic acid. A linear gradient from 10% to 80% D was run for a total of 9 min, with the flow rate set at 0.3 mL min⁻¹. The transitions in multiple reaction monitoring (MRM) mode were (precursor > fragment): ESI negative [M−H]⁻: OA (803.5 > 255.1); DTX1 (817.5 > 255.2); ESI positive [M+H or M+NH4] ⁺: PTX2 (876.5 > 823.5). Toxin reference standards for OA, DTX1, and PTX2 were purchased from the National Research Council, Halifax, Canada.

The specific intracellular toxin production rate μ_tox (pg toxin h⁻¹) during the cell cycle was determined by the ln differences in toxin concentrations T (pg toxin mL⁻¹) over time (2h) as follows (Anderson et al., 1990):

$${\text{μ}}_{\text{tox}}=\frac{\text{ln}\left({\text{T}}_2/{\text{T}}_1\right)}{2}$$

where T₁ and T₂ were the toxin concentrations (pg mL⁻¹ culture) at times 1 and 2, respectively.

2.4. Statistical Analysis

The triplicate data were checked for normality. Photosynthetic efficiency (F_v/F_m), Comparisons of percentages of each cell cycle stage and intracellular toxin content (in pg cell⁻¹) were analyzed by two-way repeated measures ANOVA (SPSS, version 21.0) with Holm-Sidak pairwise comparisons. The correlations between cellular toxin content or specific toxin production rate and different cell cycle stages were determined by Pearson correlation analysis using “psych” package in R (http://www.Rproject.org, v. 3.3.3). Prior to the calculation, cell toxin concentrations were first standardized using Z-score transformation to a mean value of 0 and standard deviation of 1 using the “vegan” package in R. The alpha level was set at 0.05 for all analyses.

3. Results

3.1. Growth rates

Well-fed D. acuminata grew exponentially for 14 days (Fig. 1A), while the exponential phase lasted 10 (Fig. 1B) and 13 days (Fig. 1C) for 2 and 4-week starved Dinophysis, respectively. Average growth rate (± standard error, SE) in the Starve_2w treatment was 0.28 ± 0.01 d⁻¹, which was significantly greater than that in D_ark (0.18 ± 0.01 d⁻¹) and Starve_4w (0.20 ± 0.01 d⁻¹) treatments, leading to a significant (P = 0.002) difference in the maximum cell concentration (8,323 ± 170 cells mL⁻¹) in the Starve_2w treatment vs that in D_ark and Starve_4w treatments (6,710 ± 262 cell mL⁻¹ and 5,485 ± 249 cells mL⁻¹, respectively).

Fig. 1.

Cell concentration (in cells mL⁻¹) of Dinophysis acuminata and Mesodinium rubrum in the following treatments: (A) feeding-dark for 58 h, (B) 2-week starvation-feeding-dark for 58 h, and (C) 4-week starvation-feeding-dark for 58 h. (D, E and F) D. acuminata cell concentrations over the diel cycle following the synchronization process. The blue and grey shaded areas indicate starvation and dark periods, respectively. Plotted values indicate the mean ± SE (n = 3).

Prey were efficiently consumed (prey cell concentration was < 50 cells mL⁻¹) for 14 days in all three treatments, at the time when all cultures entered the dark synchronization process (dark for 58 h). Ten percent, 12% and 19% of cells were lost during the sustained dark stage in Starve_2w, D_ark and Starve_4w treatments, respectively. The growth response of D. acuminata in the 24 h diurnal cycle was similar, showing no detectable growth in the light but significant growth in the dark (T-test, P < 0.05), i.e., 0.015 ± 0.005, 0.011 ± 0.003 and 0.008 ± 0.001 cells h⁻¹ in Starve_2w, D_ark and Starve_4w treatments, respectively (Fig. 1D, E and F).

3.2. Synchrony in relation to cell cycle stages

Degree of synchrony was measured as the percentage of G1 cells and by determining whether the cell population followed the typical cycle stages of G1-S-G2-M. Percentages of cells in each stage are shown in Fig. 2. Dinophysis acuminata synchrony was successfully achieved in Starve_2w and Starve_4w treatments (Fig. 2A and 2B). A better result was achieved in treatment Starve_2w, which yielded the highest percentage of G1 cells (average of 65%) and the clearest pattern of changing dynamics of G1-S-G2M cells of D. acuminata (Fig. 3). In the Starve_2w treatment, 63% of the total cell population was found in G1 phase at the onset of the light cycle (06:00). Thereafter, the percentage of G1 cells consistently increased for 10 h (until 16:00), reaching a maximum of 90% (Fig. 2A). The percentage of cells in G1 phase, decreased sharply thereafter, from 16:00 to middle-dark period at 24:00. Concurrently, the increase in G1 cells during light time and decrease until middle-dark time led to the opposite pattern in the relative abundance of S-phase cells, which decreased to 10% at 16:00 and increased to a maximum of 35% at 12:00. The G2+M cells appeared early in the morning (06:00 to 10:00) and late at dark cycle (24:00 to 06:00), peaking two hours before lights were turned on (04:00) at 30%. The nuclear/cell division, M phase (karyokinesis/mitosis) started at 02:00 and 04:00 based on microscopic observation.

Fig. 2.

Diel cell cycle pattern (A, B), PTX2, OA and DTX1 cell toxin content (C, D) of Dinophysis acuminata over a 24 h diel cycle in the 2 weeks starvation-feeding-dark (A, C) and 4 weeks starvation-feeding-dark treatments (B, D), respectively. Grey shading indicates the dark period.

Fig. 3.

Daily flow cytograms of propidium iodide-stained D. acuminata cells in the 2-week starvation-feeding-dark treatment. The colored area indicates the fitted cell cycle phase: blue shows cells in G1 phase, pink is for cells in S phase and grey for cells in G2+M phase. Plotted values indicate the mean ± SE (n = 3).

Cells in G1 phase in the D_ark treatment generally contributed ~65% of the total population except during the period 10 to 16 h (16:00 to 22:00), when they reached 83% on average (data not shown). Cells in S phase during this same period declined to a minimum of 17%. Similarly, G2+M cells appeared early in the morning (06:00 to 10:00) and late at dark cycle (02:00 to 06:00) but with a much lower contribution (2–13%). In contrast, no middle-dark period (24:00 to 02:00) shift of S and G2+M cells was observed in the D_ark treatment.

The dynamics of cell cycle stages exhibited a similar pattern in the Starve_4w treatment (Fig. 2B), with cells in G1 phase fluctuating around 60% during the entire diel cycle. Maximum % G1 cells and a concurrent decrease in % S cells were observed during the light cycle at 16:00 (Fig. 2B). A decrease in % S cells and a concurrent increase in G2+M cells were observed at middle-dark period (24:00 to 02:00).

3.3. Photosynthetic Efficiency

The photosynthetic efficiency (F_v/F_m) of D. acuminata cells over the 24 h diel cycle was examined in Starve_2w and Starve_4w treatments (Fig. 4). The value of F_v/F_m varied significantly over the 24 h cell cycle in the Starve_2w treatment, but not in the Starve_4w treatment. In the former, the photosynthetic efficiency increased significantly by 13% (T-test, P < 0.05) when the light was turned on for two hours, and remained constant at 0.69 for four hours. A sharp decline occurred during the following 4 h until 16:00, then increased until sundown (20:00). In the dark cycle, F_v/F_m remained constant at a level of 0.65 until 04:00 and then dropped again until the light was turned on again (Fig. 2A and 4A). The average level of F_v/F_m was significantly (P = 0.0002) higher in the Starve_4w treatment (Fig. 2B and 4B), with a mean value of 0.71 during the 24-h diel cycle.

Fig. 4.

Cell concentration and photosynthetic efficiency, F_v/F_m of Dinophysis acuminata over a 24 h diel cycle in the 2-week starvation-feeding-dark (A) and 4-week starvation-feeding-dark treatments (B), respectively. Grey shading and plotted values as in Fig. 2.

3.4. Toxin production and the cell cycle

The intracellular toxin content of PTX2, OA and DTX1 was quantified every two hours during the 24 h diel cycle in Starve_2w (Fig. 2C) and Starve_4w (Fig. 2D) treatments. In the former, all three toxins increased gradually for 8 h during the first light/dark cycle, decreased markedly from 14:00 to 16:00 when the percentage of G1 cells increased and S cells decreased, and increased thereafter until the middle of the dark interval (24:00, Fig. 2C). The cellular toxin content of D. acuminata dropped when the percentage of G1 cells increased and that of S+G2+M cell decreased (Fig. 2A and 2C). The temporal patterns of the intracellular content of PTX2, OA and DTX1 were comparable, with the lowest values of 25.49, 0.58 and 0.31 pg cell⁻¹ at 16:00 (10 h of light exposure), and highest values of 63.17, 3.21, and 2.09 pg cell⁻¹ for PTX2, OA and DTX1, respectively, inthe middle of the dark cycle interval (24:00, 4 hours after the initiation of darkness) (Fig. 2C). Overall, toxin cell contents of OA, DTX1 and PTX2 averaged over the 24 h cycle were significantly lower, 0.65, 0.45 and 43.93 pg cell⁻¹, in the Starve_4w treatment (Fig. 2d) than those in the Starve_2w treatment (2.22, 1.29 and 48.38 pg cell⁻¹) (T-test, P < 0.05). Additionally, a slight (non-significant) increase in the intracellular content of all toxins was observed in the middle-dark period in the Starve_4w treatment (two-way ANOVA, P > 0.05).

To evaluate toxin production of D. acuminata cells during the cell cycle, the intracellular toxin quota (= cell toxin content) of PTX2, OA and DTX1 are shown in relation to the percentage of each cycle stage in Starve_2w and Starve_4w treatments (Fig. 5). The OA and DTX1 toxin content values in the starved treatments were Z-score transformed (mean value of 0 and standard deviation of 1 using the “vegan” package in R) due to the unpaired low values (Fig. 2C and D). All three dominant toxins shared similar patterns during the period of DNA synthesis in D. acuminata. In general, cell toxicities showed a negative relationship with the percentage of G1 cells (Fig. 5A, B and C), but a positive relationship with cells in the S phase (Fig. 5D, E and F). These correlations were however not significant (P > 0.05). Cells in the G2+M phase showed no clear relationship with the toxin quota (Fig. 5G, H and I).

Fig. 5.

Z score-transformed intracellular toxin content of OA (A, D and G), DTX1 (B, E and H) and PTX2 (C, F and I) in Dinophysis acuminata over the 24 h diurnal cycle in the 2-week starvation- feeding-dark treatment (open circle) and 4-week starvation-feeding-dark treatment (dark circle) as a function of the percentage of cells in G1 (A, B and C), S (D, E and F) and G2M (G, H and I) phases; r = Pearson’s correlation coefficient and P = probability of T-test for each correlation.

Specific toxin production rate (μ_tox) was calculated using the cell toxin content of D. acuminata. In the Starve_2w treatment, variations in μ_tox with the percentage of cells in different stages were highly dynamic, such that the specific toxin production rates were positive from 8:00 to 14:00, and from 16:00 to 24:00, but negative from 14:00 ~16:00, and 24:00~8:00. To understand how toxin production varies during the cell cycle, changes in μ_tox as a function of the percentage of cells in different cell cycle stages in Starve_2w and Starve_4w treatments were examined (Fig. 6). The specific toxin production rates of PTX2, OA and DTX1 were positive and significantly correlated with the percentage of cells in S phase (P < 0.05; Fig. 6D, E and F). The production of PTX2 showed a negative correlation with the percentage of cells in G2M (Fig. 6I).

Fig. 6.

Toxin production rates (μ_tox, pg mL⁻¹ h⁻¹) of OA (A, D and G), DTX1 (B, E and H) and PTX2 (C, F and I) of Dinophysis acuminata over a 24 h diel cycle in the 2-week starvation-feeding-dark treatment (open circles) and 4-week starvation-feeding-dark treatment (dark circles) as a function of the percentage of cells in G1 (A, B and C), S (D, E and F) and G2M (G, H and I) phases; r and P and boldfaced values as in Fig. 5.

4. Discussion

There have been several cell cycle studies in phototrophic dinoflagellates e.g., (Dapena et al., 2015; Pan et al., 1999; Siu et al., 1997; Taroncher-Oldenburg et al., 1997; Wang et al., 2013b), but relatively few studies have focused on heterotrophic (Bhaud et al., 1991; Kwok and Wong, 2003; Wong and Whiteley, 1996) or mixotrophic species (Chang and Carpenter, 1991; Reguera et al., 2003). In the present study a three-step restrictive approach, “starvation-feeding-darkness”, achieved a high degree of synchrony of Dinophysis acuminata cells. This allowed determination of the variability in toxin production and photosynthetic activity of D. acuminata during individual cell cycle stages, leading to improve understanding of the mechanism of toxin biosynthesis.

4.1. Synchronization

To understand the metabolic processes that dinoflagellates undergo during different cell cycle stages, many prior attempts have been made to synchronize an experimental dinoflagellate cell population (summarized in Table 1). These attempts, include: 1) manipulation of the length of the light/dark cycle (Galleron, 1976), 2) size-specific cell filtration (Homma and Hastings, 1988), 3) release of motile cells adhered to the surface of vegetative cysts (Bhaud et al., 1991); and 4) exposure to prolonged darkness. Among these synchronization approaches, darkness has been established as a critical factor that can arrest cells at a particular stage of the cell cycle (Chisholm, 1981; Chisholm and Brand, 1981). Phototrophic dinoflagellates, e.g., Prorocentrum and Alexandrium spp., were synchronized by prolonged dark exposure, with the length of the dark period ranging from 36 h to 17 days (Table 1). Differences in the synchronization efficiency of varying dark period durations are thought to relate to the growth cycle of the dinoflagellate, i.e., sustained dark exposure needs to be long enough to span several cell division cycles if synchrony is to be achieved (Taroncher-Oldenburg et al., 1999). Nutrients are considered potential modulators of cell cycle phases (Wong and Kwok, 2005), and dinoflagellates are apparently able to rapidly adapt to the available source of nutrients, manipulating their cell-cycle progression by increasing the length of S and G2 phase in response to the number of cells in G1 phase that can reach the minimum threshold required to enter the S phase (Dapena et al., 2015).

Table 1.

Summary of previous studies on cell cycle synchronization indicating the duration of S and G2M phases and dinoflagellate growth rates.

Nutrition type	Species	Method of Synchronization	S (h)	G2M (h)	Growth Rate (d−1)	Reference
Autotrophy	Amphidinium carterae	Light/dark cycle	-	-	-	Galleron (1976)
		Dark for 48 h	6	8	0.52	Li et al. (2016)
Autotrophy	Amphidinium operculatum	-	4	6	0.77	Leighfield and Van Dolah (2001)
Autotrophy	Gonyalaux polyedra	Sieving technique	-	-	-	Homma and Hasting (1988)
Heterotrophy	Crypthecodini cohnii	Swarmers released	1.5	1.5	-	Bhaud et al. (1991)
		Swarmers released	5	10	-	Wong and Ehiteley (1996)
		Swarmers released	3	3	-	Kwok and Wong (2003)
Autotrophy	Karenia mikimotoi	Dark for 96 h	16	12	0.26	Lei and Lu (2011)
Autotrophy	Prorocentrum lima	Dark for 17 d	6	8	0.10	Pan et al. (1999)
Autotrophy	Prorocentrum donghaiense	Dark for 36 h	8	8	0.69	Wang et al. (2013b)
		Suspension transfer	8	6	0.76	Shi et al. (2013)
		Dark for 48 h	12	10	0.37	Li et al. (2015)
		Suspension transfer	8	6	0.75	Shi et al. (2017)
Autotrophy	Alexandrium catenella	Dark for 36 h	10	8	0.29	Siu et al. (1997)
		-	8	8	-	Harlow et al. (2007)
		Dark for 36 h	4	4	0.57	Wang et al. (2013 a)
		Dark for 48 h	6	4	-	Zhang et al. (2014)
Autotrophy	Alexandrium catenella (=A. fundyense)	Dark for 82 h	12	10	0.59	Taroncher-Oldenburg et al. (1997)
		Dark for 82 h	6	8	-	Zhuang et al. (2013)
Autotrophy	Alexandrium minutum	Dark for 48 h	3.5	5.5	0.69	Llaveria et al. (2009)
		Dark for 66 h	12	10	0.66	Dapena et al. (2015)
Autotrophy	Alexandrium tamarense	Dark for 82 h	-	-	0.25	Cho et al. (2011)
		Dark for 36h	6	4	0.62	Gao et al. (2012)
		Dark for 82 h	-	-	0.10	Cho et al. (2014)
Mixotrophy	Dinophysis acuminata	Dark for 48 h	-	-	0.24	Data not shown (Failed)
		Dark for 58 h	-	-	0.21	Present study (Failed)
		Starved for 2 weeks & Dark for 58 h	6~8	10	0.26	Present study (Succeeded)
		Starved for 4 weeks & Dark for 58 h	8~10	10	0.19	Present study (Unclear)

Manipulation of the light/dark cycle combined with nutrient/food limitation were used in the present study to achieve cell synchronization in the mixotrophic Dinophysis acuminata. Based on the hypothesis of Taroncher-Oldenburg et al. (1997), in mixotrophic cells, deprivation of their minimal prey can be a block point and if the period of starvation is equal to or greater than the division cycle, as most of the cells should be blocked at the same point. Thus, to achieve the synchrony of a Dinophysis population, it is probably not enough to use only dark exposure or only starvation, given the mixotrophic and kleptochloroplast characteristics of Dinophysis, in which cell division is strongly associated with prey availability.

Growth stages of Dinophysis are known to be affected by factors such as the duration of starvation (Hansen et al., 2016; Minnhagen et al., 2008; Minnhagen et al., 2011; Park et al., 2008), prey concentration or type of prey (Gao et al., 2017; Gao et al., 2019; Smith et al., 2018; Tong et al., 2010), and the duration of the dark treatment (García-Portela et al., 2018; Smith et al., 2012; Tong et al., 2011). In the present study, 2-week starvation was the most effective method to synchronize D.acuminata cells, possibly due to the “sustainable” loss of kleptochloroplasts during the prolonged starvation period (Kim et al., 2008; Riisgaard and Hansen, 2009). Minnhagen et al. (2011) found that the decrease in plastid DNA division rate (plastid DNA cell⁻¹ day⁻¹) was negatively correlated with population growth, which may be restored by refeeding the ciliate M. rubrum (Minnhagen et al., 2011). This indicates that synthesized plastids are insufficient to support cell division during prolonged (4-week) starvation.

4.2. Cell cycle and growth

The cell cycle of typical eukaryotes is comprised of G1-S-G2-M phases, yet species of dinoflagellates differ in their cell cycle patterns (Table 1, Fig. 7). Synthesis of DNA (S phase) of most phototrophic dinoflagellates usually occurs in the dark due to the need to accumulate sufficient energy and metabolites for this process (Chisholm, 1981; Pan et al., 1999; Pittendrigh, 1993). Alexandrium catenella, for example, initiated DNA synthesis ~3 to 8 h after being exposed to the dark (Siu et al., 1997; Wang et al., 2013a; Zhang et al., 2014). Alexandrium minutum entered S phase two hours after the onset of a dark cycle (Dapena et al., 2015). Prorocentrum donghaiense started DNA synthesis at the light to dark transition point (Shi et al., 2017; Wang et al., 2013b), whereas Prorocentrum lima started DNA synthesis four hours after that transition (Pan et al., 1999). Amphidinium carterae (Galleron and Durrand, 1979; Li et al., 2016), Amphidinium operculatum (Dolah Van and Leighfield, 1999; Leighfield and Dolah, 2001), Karenia brevis (Dolah Van et al., 2007) and Alexandrium fundyense (Taroncher-Oldenburg et al., 1997; Zhuang et al., 2013), all replicated DNA early during the light period.

Fig. 7.

Cell cycle stages in a representative dinoflagellate species. Dotted, solid and double line arrows indicate the end of the G1 phase, start of the S phase and start of the G2+M phase, respectively. Black bars mark the dark period.

A hypothetical diagram of cell cycle events during synchronous growth of Dinophysis acuminata on a 14:10 L:D cycle (with the sunrise set up at 6:00) is shown in Fig. 8. Although most cells are synchronized in the G1 phase after the three-step approach prior to entering this cell cycle, only a relatively small percentage (~15–30%) of the population proceed through cytokinesis over the 24-h period. Dinophysis acuminata initiated DNA synthesis at 16:00, following 10 h in the light, and DNA doubled starting at 22:00 in the dark (Starve_2w, Fig. 3, 7 and 8A) or 24:00 to ~10:00 (Starve_4w, Fig. 7 and 8B), whereas the end of the G1 phase occurred at middle time of dark cycle (24:00). This indicates that starvation had a significant effect on the percentage of cells entering the S phase or the duration of the S phase. Specific intracellular toxin production rates (μ_tox) were fitted in the hypothetical cell cycle diagram of D. acuminata (Starve_2w, Fig. 8). Specific toxin production rates were positive throughout G1 and S phases, but negative during the transition from G1 to S phase, and late in M phase, the latter reflecting cell division.

Fig. 8.

Schematic representation of cell cycle events in Dinophysis acuminata during synchronous growth on a 14:10 L:D cycle following the 2 weeks starvation-feeding-dark (A) and (B) 4 weeks starvation-feeding-dark treatments. Open bar = light period; black bar = dark period; blue = G1 phase; red = S phase; gray = G2 phase; dark gray = M phase. Positive and negative signs indicate positive vs. negative toxin production rate, respectively. Sizes of the + and − signs are proportional to the specific toxin production rate μ_tox.

The duration of the S (DNA synthesis) + G2 (DNA doubling) + M (binucleated) phases (or the terminal event) is considering a factor that controls cell division or growth rates of a dinoflagellate. For example, Taroncher-Oldenburg et al. compared cell cycle phase durations in Alexandrium catenella at different growth rates in semi-continuous culture and found that a decrease in growth rate was associated with an increase in the duration of the different cell cycle stages (Taroncher-Oldenburg et al., 1999). For Prorocentrum donghaiense, a longer terminal event was associated with a lower growth rate (Table 1; Li et al. 2015; Shi et al. 2013; Shi et al. 2017; Wang et al. 2013b). Chang and Carpenter (1991) estimated the duration of S + G2 + M phases in Dinophysis acuminata to be 11 to 13 h; therefore, the estimated in situ growth rate was ~0.54 and 0.67 d⁻¹ depending on the curve fitting used. Similarly, the duration of the S + G2M phase of D. acuminata in the present study in the Starve_2w treatment was ~13 to 15 h (Fig. 2A and 8A), and the estimated growth rate was 0.36 d⁻¹ (0.015 h⁻¹ in the dark). In contrast, the terminal period of 17 to 19h in the Starve_4w treatment (Fig. 2B and 8B), led to a lower growth rate of 0.19 d⁻¹ (0.008 h⁻¹ in the dark). Results of the present study indicate that for the slow-growing, mixotrophic dinoflagellate Dinophysis acuminata, stress either from exposure to the dark and/or starvation, which contribute to growth limitation, may also influence the duration of life cycle phases.

4.3. Cell cycle and photosynthetic efficiency

Photosynthetic efficiency (F_v/F_m) is one of the key parameters used to evaluate the photophysiological performance of algal species (Maxwell and Johnson, 2000; Parkhill et al., 2001; Suggett et al., 2009). In the present study, F_v/F_m in D. acuminata ranged from 0.56 (16:00) to 0.69 (10:00) in the Starve_2w treatment, and averaged 0.71 over 24 h in the Starve_4w treatment. These values are within the range of those published for well-fed D. acuta (0.5–0.8, Hansen et al. 2016), and are much higher than those of the phototrophic Prorocentrum lima (0.35–0.44), another DSP toxin producer (Aquino-Cruz et al., 2018), and starved D. caudata (Park et al., 2008). Other studies (García-Portela et al., 2018) also showed that photosynthetic efficiency of D. acuminata and D. acuta vary with light quality and intensity. High F_v/F_m values (~0.6–0.67) were observed in low light intensity (~10 and 40 μmol photons m² s⁻¹) and in blue or green light.

A decrease in the value of F_v/F_m implies that the cells suffered from chloroplast stress due to intrinsic causes (Baker, 2008) or external conditions leading to growth limitation (Burns et al., 2013; Maxwell and Johnson, 2000), such as nutrient limitation or suboptimal temperature (Aquino-Cruz et al., 2018; Suggett et al., 2009). During the present diel study, three significant changes in F_v/F_m occurred in the Starve_2w treatment at the end of G2+M, start of S, and end of G1, i.e., two increases (06:00–08:00 and 16:00–20:00) and one decrease (12:00–16:00), respectively. The increase of photosynthetic efficiency indicated relative enhanced maximum quantum yield of PSII primary photochemistry of D. acuminata cells, thus the improved photosynthetic capability (Baker, 2008). The minimum F_v/F_m value was measured at 16:00, when the percentage of cells in G1 was maximized and that of cells in S was minimal, suggesting that cells were stressed at this stage. The reasons for decreases in F_v/F_m are often complex, and the reduction may also be caused by conditions other than stress (Adams and Demmig-Adams, 2004; Adams et al., 2006).

4.4. Cell Cycle and toxin Production

In the present study, intracellular toxins were quantified only. Previous studies have shown that the total toxin content (intracellular + extracellular) in the early plateau phase of the D. acuminata strain (DAYS) was proportional to the number of cells collected for toxin extraction (Gao et al., 2017; Gao et al., 2018). Gao et al. (2019) used Mesodinium cell lysate to stimulate toxin release in another isolate of D. acuminata (DAYS), and the total toxin content was significantly higher only in the prey lysate treatment during late plateau phase, while time span of the experiments covered only exponential and early plateau phase. This indicated that the overall toxin production and partitioning potentials were relatively steady, especially for the isolates of DAYS. In the present study, the cells were examined at a time when intracellular toxin probably represents the bulk of the produced material, due to the evidence of no active extracellular release by the healthy D. acuminata isolate (DAEP01, Smith et al., 2012) and the same isolate (DAYS, Gao et al., 2017), and thus extracellular toxins that were not measured would not likely change the conclusions about the timing of toxin production.

Toxin production in some dinoflagellates can be induced by light (Pan et al., 1999; Taroncher-Oldenburg et al., 1997) and is potentially coupled to specific cell cycle stages. The present study was designed to identify the potential toxin production mechanism either associated with cell cycle stages or activated by photon irradiance in D. acuminata, here used as a model for toxic Dinophysis species in general.

The pattern of variation in cell toxin content of D. acuminata suggested that toxin biosynthesis in this species was strongly associated with the S stage of the cell cycle, especially in the Starve_2w treatment (Fig. 5 and 6). The toxin content (intracellular toxin cell quota) and intracellular toxin production rate (μ_tox) exhibited a positive relationship with the percentage of S phase-cells. In the DSP producer Prorocentrum lima, DTX4 synthesis occurred during the G1 and S phases, whereas that of OA and DTX1 coincided with S and G2 phases (Pan et al., 1999). This supports the hypothesis that plastids are the essential organelle for OA production, as described in P. lima cells (Zhou and Fritz, 1994). A similar diel pattern in toxin production was reported in Dinophysis acuta (Fux et al., 2010; Pizarro et al., 2008), with the highest toxin content occurring at middle-dark period. Batch culture studies also indicated that the highest toxin production rate occurs during exponential growth of Dinophysis populations (Nielsen et al., 2013; Smith et al., 2012; Tong et al., 2011; Tong et al., 2015a), again suggesting a potential temporal association between DNA synthesis and toxin production.

The marked decrease in toxin content observed in the present study at 16:00 (start of the S phase, Fig. 2C) might also result from the release or degradation of toxin. In other studies, detectable toxin levels of OA, DTX1 and PTX2 have been found in the culture medium of Dinophysis spp. due to cell lysis, indicative of the passive and active release of toxins (Nagai et al., 2011; Smith et al., 2012). Further evidence, such as the identification of genes or enzymes involved in toxin production and their regulation pathways in Dinophysis spp. during the cell cycle, or measurement of extracellular or total toxins, are needed to further support the hypothesis that DSP toxins are predominantly produced concomitantly with DNA synthesis.

Little is known about the relationship between photosynthesis and toxin production of Dinophysis spp. The toxin concentrations of D. acuminata cells remained low during a 40 day-dark treatment (Tong et al., 2011), and light had a strong effect on DSP toxin production (Nielsen et al., 2012; Nielsen et al., 2013; Tong et al., 2011). In the present study, the continuous increase in toxin production of D. acuminata in the light over four hours (Fig. 2 and 8) suggests that toxin production may be initiated by exposure to light. Okadaic acid (OA) was suggested to be synthesized and stored in the chloroplast (Zhou and Fritz, 1994), perhaps as a direct product of photosynthesis, or it may be stored in peripheral vacuoles of the dinoflagellate Prorocentrim lima (Barbier et al., 1999). However, OA in the mixotrophic D. acuminata, may be linked to photosynthetic activity, but would not be solely controlled with the “kleptoplast” because there is no evidence for toxins in Mesodinium or the cryptophyte.

Combined results of the present and prior studies suggest that temporal variation in toxin content of D. acuminata is likely related to cell cycle events, light, temperature, genetics, ammonium concentration, prey and growth phase. Any factor affecting cell cycle stages or cell division will also alter cell toxicity, so many of these factors overlap in this regard. Although the mechanism of DSP synthesis in Dinophysis spp. is not yet fully understood, the present study makes an important contribution by demonstrating a clear role of cell cycle stages in the mixotrophic Dinophysis acuminata, as evidenced by the variation of photosynthetic efficiency, toxin content and toxin production rate over the cell cycle.

5. Conclusions

The three-step synchronization approach, “starvation-feeding-dark”, used in the present study yielded a high degree of synchrony in a mixotrophic Dinophysis species. Furthermore, results indicated that the length of starvation may affect the timing and/or duration of cell cycle stages. The isolate of Dinophysis acuminata used exhibited a comparable G1-S-G2-M transition to that described in other dinoflagellates. Synchronization of cells in the population started in conjunction with DNA synthesis (S phase) 10 h after exposure to light, was followed by the second growth stage (G2 phase) in the middle of the dark period and completed mostly (G2M cells) 2 h before the lights were turned on. Over a diel cycle the toxin content of OA, DTX1 and PTX2 followed a similar pattern of increase during the G1 phase, significant reduction upon entering the S phase, and then re-increase during the S phase and decrease during G2M. Specific toxin production rates were positive throughout the entire single G1 and S phases, but were negative (indicating less toxin production) when cells transitioned from G1 to S phases, and during the late G2M phase. All three toxins were initially induced by light exposure and intracellular toxin production was positively correlated with the percentage of cells in S phase, indicating that biosynthesis of DSP toxins might be under circadian regulation and most active during DNA synthesis. Thus, the current study provides new data on regulation of growth, photosynthetic efficiency and toxin production during the cell cycle of a mixotrophic Dinophysis species.

Highlights.

• A successful three-step synchronization approach, termed “starvation-feeding-dark”, has achieved a high degree of synchrony of Dinophysis cells.

• The synchronized Dinophysis cells started DNA synthesis (S phase) 10 hours after being released into the light, initiated G2 growth stage eight hours later, and completed mitosis (M phase) 2 h before lights were turned on.

• Toxin variability during the cell cycle of D. acuminata were first investigated, showing that okadaic acid (OA), dinophysistoxin-1 (DTX1) and pectenotoxin-2 (PTX2), were initially induced by the light and biosynthesis of Dinophysis toxins might be most active during DNA synthesis.

Abbreviation

Fv/Fm

the maximum photochemical quantum yield of PSII

F0

the minimum in vivo relative chlorophyll-a fluorescence yields in a dark-adapted state

Fv

the maximum variable chlorophyll-a fluorescence yields in a dark-adapted state

Fm

the maximum in vivo relative chlorophyll-a fluorescence yields in a dark-adapted state