Growth, Toxin Content and Production of Dinophysis norvegica in Cultured Strains Isolated From Japan

1. Introduction

Harmful Algal Blooms (HABs) are caused by several species of microalgae in freshwater, marine, and brackish environments, which may lead toimportant ecosystematic and socioeconomic impacts as well as human illnesses [1,2,3]. There has been a marked increase in the occurrence of HABs worldwide over the past decades [1,2,3,4,5]. The rise has been associated with climate change and intensified anthropogenic activities, notably eutrophication, transport of species with maritime activities, alteration of natural habitats, and growth of the aquaculture industry [4,6,7,8,9,10,11,12,13,14].

In marine waters, dinoflagellates form the majority of toxin-producing HAB species and are responsible for several human poisoning syndromes, including Ciguatera Fish Poisoning (CFP), the Neurotoxic Shellfish Poisoning (NSP), Paralytic Shellfish Poisoning (PSP), and the Diarrhetic Shellfish Poisoning (DSP) [15]. For example, gastrointestinal poisoning in humans is caused by the consumption of shellfish contaminated with DSP toxins [16,17]. These toxins are produced by dinoflagellates from the genera Dinophysis, Phalacroma, and Prorocentrum. Ten species of Dinophysis and two species of Phalacroma are known to produce lipophilic Diarrheic Shellfish Toxins, or DSTs, i.e. Okadaic Acid (OA) and its analogues the dinophysistoxins (DTX) principally DTX1, DTX2 and DTX3; in addition to the bioactive pectenotoxins, the PTX [17,18,19,20,21,22,23,24,25,26,27,28].

Typically, Dinophysis spp. do not attain high cell densities but form dense patches of populations, which sets them apart from other HAB species and makes their monitoring and prediction of shellfish contaminations with DSTs more difficult, especially since molecular tools have been hard to develop due to insufficient resolution to differentiate between the species [21,29,30,31,32,33,34,35,36].

Despite the availability of extensive studies, little is known about the ecophysiology, bloom mechanisms, and toxin production of Dinophysis spp. due to difficulties in establishing and maintaining cultures [21,37,38,39]. The discovery of mixotrophy in Dinophysis spp. [37,38,39,40] and plastids of cryptophyte origin [40,41,42,43,44,45,46,47] led to the first success in establishing cultures of Dinophysis acuminata [48]. Seven species were subsequently cultured based on feeding Dinophysis spp. with the ciliate Mesodinium rubrum grown with the cryptophyte Teleaulax sp., namely D. fortii [49], D. acuta [50], D. sacculus [51], D. tripos [52], D. cf. ovum [53], D. caudata [54], and D. infundibulum [55]. Mainly growth and, in some cases, toxin production in the established cultures have been reported. A few studies have investigated the effects of temperature, prey, and irradiance on the growth and toxin production of Dinophysis spp. in these cultures [33,48,56,57,58,59,60,61,62,63,64,65].

There is evidence of the global expansion of Dinophysis species related to both climate change and aquaculture activities [14,66], resulting in hardship to fisheries and aquaculture industries through extended closure of shellfish production [67]. Among the toxigenic species, six of the toxic Dinophysis species have a wide, global distribution, including D. norvegica [36,68,69]. A boreal to cold-temperate species, D. norvegica is commonly reported from the Northern Hemisphere, for example from the coastal waters around Scotland and Norway, the Baltic Sea, and the Arctic Sea [29,30,71,72,73,74,75,76]. Recently, it was reported for the first time at very low occurrence in oceanic samples in the Southern Hemisphere, in southern Argentine Sea [77]. It forms dense blooms in the Baltic Sea and Eastern Canada with mild DSP outbreaks [78,79,80]. The earliest information based on cells picked from environmental samples showed the production of OA and DTX1 in Norway [17], and high content of OA in Eastern Canada [81]. More recently, LC-MS have shown the production of PTX2, PTX12, and traces of OA by strains from Norway [22]. In the Baltic Sea, D. norvegica produces OA, PTX2, and PTX2SA [82] leading to the contamination of blue mussels and flounders with OA [83,84]. One recent study reported the production of Dihydrodinophysistoxin-1 in picked cells from environmental samples and cultures of D. norvegica from the Gulf of Maine, USA [85], with a complete absence of OA, DTX1, and DTX2 following analyses with LC-MS/MS. In Japan, high levels of PTX2 have been reported for the first time in cells of D. norvegica picked from environmental samples [86]. In a later study, PTX2 was confirmed as the dominant toxin in D. norvegica, although some of the picked cells had trace levels of OA and DTX1 [87]. In the present study, we report the successful cultivation of D. norvegica isolated from Japanese waters for the first time. The toxin productions in seven strains of D. norvegica are provided as well as the information on the growth and toxin production of one strain during a 36-day culture experiment.

2. Results

2.1. Species identification

Cells of Dinophysis norvegica are generally large, ovoid, and robust. The posterior end tapers to a triangular shape (Figure 1). Dinophysis norvegica is very similar to D. acuta in morphology, therefore they may be misidentified. These species can be differentiated by their size (although it overlaps) and the location of thewidest position: D. acuta is larger and widest below the mid-section, whereas D. norvegica is smaller and widest in the middle region of the cell [88,89]. A phylogenetic analysis based on the D1/D2 region (735 bp) supported that the strains isolated from Funka Bay, Japan belong to the D. norvegica clade (Figure 2), and are closely related to the strains from Canada, Norway, and the USA from the Atlantic Ocean..

2.2. Feeding behavior and growth of Dinophysis norvegica in culture experiments

Only seven cultures of 48 single-cell isolates grew with the addition of the ciliate Mesodinium rubrum as the prey species. However, they were successfully established as clonal strains and the success rate for isolation was 14.6% (7/48). The cultures reached maximum cell densities of 1,057-3,050 cells ml^-1 (2,020 ± 702) (mean ± SD, n = 7) at end of the one-month incubation. Similar to the case of other Dinophysis species, Dinophysis norvegica was able to feed on M. rubrum and grew. The microphotographs showed the large nucleus which occupied the upper half of the cell and food vacuoles at the lower part (Figure 1A) and numerous chloroplasts (Figure 1B).

In the growth experiment, M. rubrum grew exponentially during the first 9 days, reaching until 5,600± 346 cells mL^-1 (mean ± SD, n = 3) (Figure 3). After that, cell abundances of M. rubrum decreased sharply and disappeared by Day 30, probably due to the active consumption by D. norvegica and natural death. Dinophysis norvegica grew, but it showed very slow growth for the first 12 days, suggesting that they have a long lag phase. After that they grew exponentially with a maximum growth rate of 0.56 divisions day^-1 (during Day 24-27), reaching a maximum concentration of 2,883 ± 104 cell mL^-1 at the end of incubation (Day 36), and the growth rates in every three days from Day 0 to Day 30 were -0.06 to 0.56 divisions day^-1 (0.18 ± 0.18) (Figure 3).

2.3. Toxin production

Low levels of DTX1 were found but only in cells of three strains of D. norvegica at 0.5 pg cells^-1 for DN05, 0.7 pg cell^-1 for DN08, and 1.2 pg cell^-1 for DN06. All seven strains, however, produced PTX2 with cell quotas varying from 60.6 pg cell^-1 to 152.4 pg cell^-1 (Table 1). All strains of Dinophysis norvegica did not produce OA, except for Stain DN08 which detected a trace level of OA. In cultures, the toxin concentration ranged from 0.705 ng mL^-1 to 3.55 ng mL^-1 in the case of DTX1 and from 126 ng mL^-1 to 375 ng mL^-1 in the case of PTX2 (Table 1).

Strain DN16062021FUN-06 of D. norvegica produced DTX1 and PTX2 throughout the 36-day culture experiment (Figure 4). Both productions of DTX1 and PTX2 started to increase from 0.092 ± 0.009 pg cell^-1 on Day 12 to 1.280 ± 0.185 pg cell^-1 on Day 36, and from 9.3 ± 0.5 pg cell^-1 on Day 12 to 154.7 ± 24.2 pg cell^-1 on Day 36, respectively. The specific toxin production rates for DTX1 and PTX2 during the exponential growth phase were 0.291 ± 0.020 pg mL^-1 day^-1 and 0.291 ± 0.023 pg mL^-1 day^-1, respectively. The net toxin production rate (R_tox) during the exponential growth phase was 0.001±0.0003 ng mL^-1 day^-1 for DTX1 and 0.13±0.03 ng mL^-1 day^-1 for PTX2 in the exponential phase. DTX2 concentrations that were above the detection limit were available from Day 18 to Day 36 and ranged from 0.132 to 1.28 ng mL^-1. For PTX2, the concentrations above the detection limit were measured from Day 3 until the end of the experiment Day 36, and were in the range of 3.4 to 154.7 ng mL^-1.

3. Discussion

Similar to the case of other Dinophysis species [49,50,51,52,53,54,55] , D. norvegica was able to feed on M. rubrum. Dinophysis norvegica swam around the prey and inserted the peduncle into the cell of M. rubrum, the ciliate became immobile and their cilia were shed from the cell within 1-5 minutes. The cytoplasm of the prey was actively ingested through the peduncle. It took 45-100 min until whole cell content of M. rubra was consumed in D. fortii and D. tripos [49,52]. Dinophysis acuminata, D. caudata, D. fortii and isolated from Japanese coastal waters displayed growth rates of 0.50-1.03 divisions day^-1, reaching maximum concentrations of 2,200-11,000 cells mL^-1 at temperature ranging from 18 to 25 °C [36] . In this study, D. norvegica showed a similar growth rate at a lower temperature of 12.5 °C than other species (Figure 2). The culture strains were maintained successfully at high densities (>2,000 cells mL^-1) for more than one year. This is an advancement compared to previous findings, where clonal cultures of D. norvegica isolated from Funka Bay and lake Notoroko in Hokkaido, Japan, grew well (>1,000 cell mL^-1) in the first incubation (24/96, 25%), however, when they were reinoculated into fresh Mesodinium cultures, no further growth was confirmed resulting in the failure to establish cultures (0/96, 0%) [52]. This led to the conclusion that the predator and the prey from different regions may be incompatible and cause of the failure in culturing [52]. In the growth experiment using the established cultures, D. norvegica had much longer log phase (12 days) (Figure 3) than those in D. acuminata, D. caudata and D. fortii (3-4 days) [49,54,55]. Perhaps, D. norvegica may become unstable when they are reinoculated into fresh Mesodinium cultures because of sudden environmental changes such as in pH levels. It is suggested that the reinoculation of D. norvegica cells with relatively high concentration (>250 cells mL^-1) would achieve the successful maintenance of the cultures for long periods. The ranges of water temperatures and salinity in which this species appears in southern Hokkaido including Funka Bay were 2-16 °C and 24.3-33.9 PSU during 2016-2020 (https://www.hro.or.jp/list/fisheries/research/central/section/kankyou/kaidoku/j12s220000000dgw.html). Influence of the temperature (12.5 °C) on the success of establishing cultures is still unclear in this study.

Establishment of culture strains in the toxic Dinopshysis species is important for investigating toxin production and how it is influenced by changes in the physical-chemical conditions. Until now, information on the concentration of toxins produced by D. norvegica has been available based on picked and pooled cells from field samples from two locations in Norway and Japan [19,22,87]. Similarly to the previous records based on isolated cells, OA concentrations in cultured strains remained mainly below the detection limit [19,22,87]. The maximum DTX1 concentrations in cultured strains were lower than reported based on cells from field samples (1.16 pg cell^-1 vs 14 pg cell^-1) [19,22,87]. Interestingly the PTX2 maximum concentrations were much higher in the cultured strains (152 pg cell^-1), than in the cells collected from the field in Japan (89 pg cell^-1) and 1.7 pg cell^-1 in Norway [19,22,87]. From 1987 until 2022, 43 HABs associated with D. norvegica have been recorded globally with the majority of those associated with DSP [90]. A study based on Scottish shellfish farms estimated that a 1% change in the production of toxins produced by Dinophysis spp. led to an annual loss of 1.37 million £ (GBP) [91]. Thus, further experiments with cultured strains can lead to an improved understanding of the variability in toxin production between the strains from the same and different geographic locations as well as the influence of changes in physical-chemical variables [69,92].

4. Materials & Methods

4.1. Isolation of clonal strains and establishment of clonal cultures

Mesodinium rubrum and Teleaulax amphioxeia were isolated from Inokushi Bay (32.7998 N, 131.8923 E) in Oita Prefecture, Japan, at the end of February 2007 [49]. The M. rubrum culture was maintained by mixing 50 ml of the culture (7.0–9.0 × 10³ cells mL^–1) with 100 ml of a modified f/2 medium [93,94], with the addition of 25-100 µL of T. amphioxeia culture (containing 0.5–2.0 × 10⁴ cells) as a food source. The culture medium was prepared with 1/3 nitrate, phosphate, metals, and 1/10 vitamins, plus any enrichment from the autoclaved natural seawater collected from Tokyo Bay (35.3460 N, 139.6570 E). Seawaters were filtrated through three cartridge filters with 5, 1, and 0.5 µm (STG-10-5, STG-10-1, STG-10-0.5, Kankyotechnos, Japan) at the sampling site and brought back to the lab. Salinity was adjusted to 30 practical salinity units (PSU) before autoclaving. A part of the ciliate culture was transferred into a fresh culture medium containing the food source once a week, and they were maintained at a temperature of 18 ^˚C under a photon irradiance of 100 μmol m^–2 s^–1, provided by cool-white fluorescent lamps, with a 12:12 h light: dark cycle.

Cells of Dinophysis norvegica (48 cells in total) were isolated by micropipetting from a seawater sample collected from Funka Bay, Japan (42.28N, 140.35E) in June 2021 and incubated in individual wells of a 48-well microplate (Iwaki, Japan). Each cell was grown in 1.0 mL of the culture medium prepared for the maintenance of the M. rubrum strain, containing ca. 1.0 × 10³ cells of the marine ciliate M. rubrum as the prey species. Dinophysis cells were incubated under the same light conditions as those for the M. rubrum culture, but at the temperature of 12.5 °C. After one month of incubation, several strains grew well and were established as clonal strains (7 strains in total). Small aliquots (0.1 mL) of the established cultures in each strain were inoculated into 2.9 mL of fresh M. rubrum culture (ca. 2 × 10³ cells mL^–1, just after reinoculation for the maintenance without adding Teleaulax culture) in 12-well microplates to maintain these strains, and they were incubated for one month under the same conditions as mentioned above. After one month of incubation, 1 mL of each culture was sampled for toxin analysis, and cell abundances in each stain were counted using a Nikon TE-300 (Tokyo, Japan) inverted microscope. Also, the orange auto-florescence derived from the chloroplasts of M. rubrum in D. norvegica cells was observed using an epifluorescence microscope under blue light excitation (Zeiss Axioskop 2 (Carl Zeiss, Göttingen, Germany) with digital camera Axiocam 305 color (Carl Zeiss, Göttingen, Germany).

4.2. Growth experiments

The Mesodinium rubrum culture grown until the late exponential growth phase (ca. 5.0 × 10³ cells mL^–1) was diluted with the fresh culture media to give initial concentrations of ca. 2.0 × 10³ cells mL^–1, and 7.5 mL aliquots of the mixed culture were inoculated into the wells of 6-well microplates (Iwaki, Japan). Next, 125 µL of a D. norvegica culture (strain, DN16062021FUN-06) containing 375 cells was added into the M. rubrum culture to give an initial concentration of 50 cells mL^-1. The growth experiment was conducted for 36 days under the same light and temperature conditions used for maintaining the culture of D. norvegica. In the growth experiment, 1 mL of the cultures (triplicate) was sampled every three days except for Day 33. The cell densities of D. norvegica and M. rubrum were counted using an inverted microscope. The specific growth rate (µ, divisions day^-1) of D. norvegica was determined during the exponential growth phase according to [95].

4.3. Sequences of 28S rDNA (D1-D2 region)

Genomic DNAs were extracted from several cells in each strain with 5% Chelex buffer in four of the established strains [96] . PCR amplification was carried out on a thermal cycler (PC-808, ASTEC, Fukuoka, Japan) with a reaction mixture consisting of 1 μL template DNA, 1 μM each of D1/D2 primer sets [97] , 0.2 mM of each dNTP, 1× PCR buffer, 1.5 mM Mg²⁺, 1U KOD-Plus-Ver.2 (TOYOBO, Osaka, Japan), and RNA-free dH₂O to bring up to 25 μL volume. The PCR cycling conditions were as follows: 2 min at 94 °C, 30 cycles at 94 °C for 15 sec, 55 °C for 30 sec, and 68 °C for 40 sec. The PCR products were transformed into DH5α cells (Promega, Madison, WI, USA) after ligation into the pGEM T-Easy Vector (Promega). The plasmid DNAs were purified after color selection. DNA sequences were determined using a Dynamic ET terminator cycle sequencing kit (GE Healthcare, Little Chalfont, UK) in combination with M13 Reverse and U19 primers and analyzed on a DNA sequencer (ABI3730, Applied Biosystems, Foster City, CA, USA). BLAST search was performed to determine closely related species and their GenBank sequences were obtained for phylogenetic analyses. The sequences were aligned using AliView [98] ( and identical sequences were compiled into a single sequence. All newly obtained sequences were deposited into the DDBJ databank (accession numbers: LC634028- LC634030).

A phylogenetic tree was constructed based on maximum likelihood (ML) using MEGA version 10 [99] with the best substitution model selected: Kimura 2-parameter model plus gamma distribution (G = 0.86). Bootstrap support (BS) values of ML and neighbor-joining (NJ) analyses for the trees were estimated using 500 replicates each. For the posterior probabilities (PP) of Bayesian inference, the best model substitution calculated by Akaike information criterion in jModelTest version 2.1.10 [100] was TIM plus gamma (G = 0.9710), and the effective sample size was calculated using Bayesian Evolutionary Analysis Sampling Trees (BEAST) and Tracer. Bayesian inference was conducted using MrBayes version 3.2.5 [101] based on the Bayesian information criteria calculated by jModelTest. 3,077,000 Markov chain Monte Carlo generations was with four chains and trees sampled every 1,000 generations and PP were estimated with 25% generations burn-in. Convergence of the chains was reconfirmed when the average standard deviations of the split frequencies were below 0.01 after calculations. Sequences from Prorocentrum micans and Prorocentrum minimum were used as outgroup.

4.4. DSP toxin analysis

The samples were frozen at –30°C until the toxins were extracted by solid-phase extraction (SPE). The SPE of toxins was modified compared to a previous method [86,102,103]. The 1-mL frozen and thawed samples were applied to the MonoSpin C18 centrifuge cartridge column (GL Science Inc., Tokyo, Japan) equilibrate with 1.0-mL each methanol and distilled water. The SPE column was washed with 0.5-mL distilled water, and the toxins were eluted with 0.1-mL methanol. The methanol elutes were directly analyzed by LC-MS/MS. LC-MS/MS analysis of the toxins was carried out according to a previous method [104]. A Nexera-20XR series liquid chromatograph (Shimadzu, Kyoto, Japan) was coupled to a QTRAP 4500 mass spectrometer (SCIEX, MA, USA) of hybrid triple quadrupole/linear ion trap. Separations were performed on LC columns (internal diameter [i.d.], 100 mm × 2.1 mm) packed with 1.9-μm Hypersil GOLD C8 (Thermo Fisher Scientific Inc., Waltham, MA, USA) and maintained at 30°C. Eluent A was water, and eluent B was acetonitrile water (95:5), containing two mM ammonium formate and 50 mM formic acid. Toxins were eluted from the column with 50% B at a flow rate of 0.3 mL·min^–1. Multiple reaction monitoring (MRM) LC-MS/MS analysis with negative-mode ionization was carried out using the target parent ions and the fragment ions in Q1 and Q3 for each toxin as follows: OA, m/z 803.5 > 255.1; DTX1, m/z 817.5 > 255.1; PTX2, m/z 857.5 > 137.0; PTX1 and PTX11, m/z 873.5 > 137.0; PTX2 Seco acid (PTX2 SA), m/z 875.5 > 137.0. The lowest detection limits of OA/DTX1 and PTX2 were 0.1 and 1.2 ng·mL^–1. These levels are equivalent to 0.1 pg·cell^–1 of OA (and DTX1) and 1.2 pg·cell^–1 of PTX2, when 100 cells of the toxic plankton were analyzed using our LC-MS/MS method. During the growth experiments, the specific toxin production rate (µtox, pg cell^-1 day^-1) and net toxin production rate (R_tox) were calculated for the exponential growth phase, using the previously published equations [105].