Marine Biotechnology

, Volume 7, Issue 2, pp 95–103

Development of Molecular Probes for Dinophysis (Dinophyceae) Plastid: A Tool to Predict Blooming and Explore Plastid Origin


  • Yoshiaki Takahashi
    • School of Fisheries SciencesKitasato University
  • Kiyotaka Takishita
    • Marine Biotechnology Institute
    • School of Fisheries SciencesKitasato University
  • Tadashi Maruyama
    • Marine Biotechnology Institute
  • Takeshi Nakayama
    • Institute of Biological SciencesUniversity of Tsukuba
  • Atsushi Kobiyama
    • School of Fisheries SciencesKitasato University
  • Takehiko Ogata
    • School of Fisheries SciencesKitasato University

DOI: 10.1007/s10126-004-0482-5

Cite this article as:
Takahashi, Y., Takishita, K., Koike, K. et al. Mar Biotechnol (2005) 7: 95. doi:10.1007/s10126-004-0482-5


Dinophysis are species of dinoflagellates that cause diarrhetic shellfish poisoning. We have previously reported that they probably acquire plastids from cryptophytes in the environment, after which they bloom. Thus monitoring the intracellular plastid density in Dinophysis and the source cryptophytes occurring in the field should allow prediction of Dinophysis blooming. In this study the nucleotide sequences of the plastid-encoded small subunit ribosomal RNA gene and rbcL (encoding the large subunit of RuBisCO) from Dinophysis spp. were compared with those of cryptophytes, and genetic probes specific for the Dinophysis plastid were designed. Fluorescent in situ hybridization (FISH) showed that the probes bound specifically to Dinophysis plastids. Also, FISH on collected nanoplankton showed the presence of probe-hybridized eukaryotes, possibly cryptophytes with plastids identical to those of Dinophysis. These probes are useful not only as markers for plastid density and activity of Dinophysis, but also as tools for monitoring cryptophytes that may be sources of Dinophysis plastids.


Dinophysisfluorescent in situ hybridization (FISH)shellfish poisoningcryptophyteplastid


Some phytoplankton species are known to produce toxins that accumulate in plankton feeders. In particular, toxin accumulation in bivalves causes food poisoning in humans, and often leads to severe economic damage to the shellfish industry.

Diarrhetic shellfish poisoning (DSP) is a gastrointestinal syndrome caused by phytoplankton toxins, including okadaic acid, and several analogues of Dinophysistoxin (Yasumoto et al., 1985). These toxins are derived from several species of dinoflagellates belonging to the genus Dinophysis (Yasumoto et al, 1980; Lee et al., 1989). Despite extensive studies in the last 2 decades, little is known about the ecophysiology and blooming mechanisms of Dinophysis species because they are difficult to grow in culture.

Dinophysis species are divided into 2 groups, photosynthetic and nonphotosynthetic (heterotrophic) species, which are determined by the presence or absence of plastids, respectively (Lessard and Swift, 1986). The majority of the DSP-inducing species belong to the former group. Even in the photosynthetic species, food vacuoles are occasionally seen in the cells (Jacobson and Andersen, 1994; Koike et al., 2000), and heterotrophy is one mode of nutrition. Because plastid density in Dinophysis cells increases prior to blooming, photosynthesis is thought to be essential for the blooming process (Koike, 2002). Thus observation of the plastid density and understanding of the environmental conditions that cause increases in plastid density are necessary to predict blooming and subsequent outbreaks of DSP.

The plastid of Dinophysis is unique in dinoflagellates. It contains phycobilin-proteins as accessory pigments (Lessard and Swift, 1986; Hallegraeff and Lucas, 1988; Schnepf and Elbrächter, 1988; Geider and Gunter 1989; Hewes et al., 1998; Vesk et al., 1996) and a double thylakoid system with an electron-dense lumen (Schnepf and Elbrächter, 1988). Because these are characteristics of cryptophyte plastids, not of dinoflagellates, Dinophysis plastids are thought to be obtained through endosymbiosis with a cryptophyte. In addition, the plastid is considered a permanent organelle because there are no other remnants of a cryptophyte within the Dinophysis cell other than the plastids (Lucas and Vesk, 1990; Schnepf and Elbrächter, 1999).

We previously reported that 3 species of photosynthetic Dinophysis share a type of plastid containing identical plastid-encoded small subunit ribosomal DNA (pSSU rDNA) sequences, whereas their nuclear-encoded SSU rDNA sequences have species-specific base substitutions (Takishita et al., 2002). In general, the sequences from the fully established dinoflagellate plastids (containing–peridinin and fucoxanthin derivatives) have diverged substantially from the nuclear genes (Zhang et al., 1999, 2000; Barbrook and Howe, 2000; Tengs et al., 2000). We therefore suspect that the Dinophysis plastid is derived from the temporary acquisition of cryptophytes from the environment. This idea is supported by previous observations that the pigment concentrations and plastid morphologies of Dinophysis are extremely variable (Fukuyo, 1997; Koike, 2002) and that Dinophysis fortii can take up cryptophyte cells and maintain their plastids (Ishimaru et al., 1988). Hence cryptophytes with a plastid identical to that of Dinophysis should be crucial for plastid acquisition and blooming.

In this study we developed suitable genetic probes for pSSU rRNA and rbcL (encoding the large subunit of ribulose-1,5-bisphosphate carboxylase/oxygenase) messenger RNA in photosynthetic Dinophysis plastids. We describe the ability of the probes to bind to various cryptophytes and Dinophysis cells. We also describe the results of a trial for detecting environmentally occurring cryptophyte cells that are possible sources of Dinophysis plastids. These probes, along with fluorescent in situ hybridization (FISH), should be useful for (1) microscopic counting of DSP-inducing Dinophysis, (2) estimation of plastid density and photosynthetic activity, and (3) detection and enumeration of cryptophyte cells that could be the source of Dinophysis plastids.


Plastid-Encoded SSU rDNA and rbcL Gene Sequencing from Dinophysis and Cryptophyte Plastids

All of the DNA sequences used in this study and their GenBank accession numbers are listed in Table 1. Sequences of pSSU rDNA for 3 Dinophysis species (D. fortii Pavillard, D. acuminata Claparède and Lachmann, and D. norvegica Claparède and Lachmann) have been reported previously (Takishita et al., 2002). In addition, the rbcL gene from D. tripos Gourret and D. fortii and pSSU rDNA from D. tripos were sequenced for the first time in this study. The D. fortii and D. tripos cells were collected at Okkirai Bay, Iwate, Japan, on may 14 and 21, 2002, respectively. Two cryptophyte isolates collected from Tokyo Bay on May 15, 2003, tentatively identified as Plagioselmis sp. and Teleaulax sp. (Tsukuba University culture collections) on the basis of their nuclear SSU rDNA sequences, were used for pSSU rDNA and rbcL sequencing. Also, the rbcL gene sequences from Geminigera cryophila Hill (Marine Biotechnology Institute culture collection; MBIC10567) and Proteomonas sulcata Hill and Wetherbee (Provasoli-Guillard National Center for Culture of Marine Phytoplankton; CCMP 765) were determined.
Table 1

GenBank Accession Numbers Used in This Study

Gene and species name

Accession number

Plastid SSU rDNA


Chilomonas paramecium


Chroomonas placoidea


Cryptomonas ovata


Dinophysis acuminata


D. fortii


D. norvegica


D. tripos


Geminigera cryophila


Guillardia theta


Hemiselmis virescens


Palmaria palmata


Plagioselmis sp. (TUC-1)


Porphyra purpurea


Proteomonas sulcata


Pyrenomonas salina


Teleaulax sp. (TUC-2)




Chilomonas paramecium


Chroomonas sp. (SAG 980-1)


Dinophysis fortii


D. tripos


Geminigera cryophila


Guillardia theta


Palmaria palmata


Plagioselmis sp. (TUC-1)


Proteomonas sulcata


Pyrenomonas helgolandii


Teleaulax sp. (TUC-2)


DNA extraction, polymerase chain reaction (PCR) amplification of pSSU rDNA, cloning, and sequencing were performed according to Takishita et al. (2002). The rbcL gene was PCR-amplified with the following set of primers: GMRUBISCO1 and GMRUBISCO2 (Takishita et al., 2000).

Phylogenetic Analysis

The pSSU rDNA sequences from 4 species of Dinophysis (D. acuminata, D. fortii, D. norvegica, and D. tripos) were aligned with those from 10 species of cryptophytes and 2 species of rhodophytes (Table 1) by CLUSTAL W Version 1.8 (Thompson et al., 1994). Also using CLUSTAL W, we aligned the rbcL gene sequences from 2 species of Dinophysis (D. fortii and D. tripos), 8 cryptophytes, and one rhodophyte (Table 1). The pSSU rDNA sequences of D. acuminata, D. fortii, D. norvegica, G. cryophila, Chilomonas paramecium Ehrenberg, Cryptomonas ovata Ehrenberg, Guillardia theta Hill and Wetherbee, P. sulcata, Hemiselmis virescens Droop, Chroomonas placoides Butcher, Pyrenomonas salina (Wislouch) Santore, Porphyra purpurea (Roth) C. Agardh, and Palamaria palmata (Linnaeus) Kuntze, and the rbcL sequences of G. theta, P. salina, Chroomonas sp., C. paramecium, and P. purpurea were obtained from the DNA Data Bank of Japan (DDBJ). The generated alignments were visually inspected and manually edited. All ambiguous sites of the alignments were removed. The alignment data for pSSU rDNA and rbcL are available on request from the corresponding author.

The data sets of pSSU rDNA (16 taxa/1227 sites) and rbcL (11 taxa/996 sites) were tested for their optimal fit to various models of nucleotide evolution using MODELTEST Version 3.06 (Posada and Crandall, 1998). The proportion of invariable sites, a discrete γ distribution (4 categories), and base frequencies were estimated from the data set. Each maximum-likelihood (ML) tree was constructed under an optimal model. The data sets of pSSU DNA and rbcL were also subjected to analyses by the neighbor-joining (NJ) (Saitou and Nei, 1987) and maximum parsimony (MP) methods. The NJ tree was constructed using Kimura’s 2-parameter model (Kimura, 1980). Support for NJ branches was tested by bootstrap analysis of 1000 replicates. The MP tree was based on the tree-bisection-reconnection (TBR) branch-swapping algorithm with stepwise addition (the closest option) of taxa under the heuristic search method (50% confidence level). We conducted bootstrap analysis of 1000 replicates using the heuristic search method (50% confidence level) to assess the confidence of the branches in the MP tree.

For all phylogenetic analyses in this study, PAUP* Version 4.0 was used.

Genetic Probe Design

Probes for Dinophysis plastid SSU rRNA and rbcL mRNA were designed according to their specific regions of the sequences (Table 2). In addition to Dinophysis plastid-specific probes, we designed probes for Geminigera cryophila plastid (SSU rRNA and rbcL mRNA) to optimize the hybridization conditions. These oligonucleotides were synthesized with fluorescein isothiocyanate (FITC) conjugated to their 5′ ends (Espec Oligo).
Table 2

Probes for Dinophysis Plastid SSU rRNA and rbcL mRNA and Their Sequences

Probe name




Dinophysis spp., plastid SSU rRNA



Dinophysis spp., plastid rbcL mRNA



Geminigera cryophila, plastid SSU rRNA


aA probe for Geminigera cryophila plastid SSU rRNA used for optimizing hybridization and as a negative control.

Fluorescent In Situ Hybridization

Dinophysis cells and cryptophyte cultures used for FISH are listed in Table 3. Fixation and hybridization were performed essentially as described by Miller and Scholin (2000). Capillary-isolated Dinophysis cells in 20 μl of seawater or 500 μl of cryptophyte cultures in mid-logarithmic growth phase were fixed in 9 volumes of the standard saline ethanol fixative (a mixture of 25 ml 90% ethanol, 2 ml H2O, and 3 ml 25× SET buffer [3.75 M NaCl, 25 mM EDTA, 0.5 M Tris-HCl, pH 7.8]), after which the cells were stored at −20°C. To eliminate auto fluorescence from phycoerythrin pigment, which can overlap with FITC emission, fixed cells in a polystyrene tube were irradiated under a UV lamp (GL-15, Toshiba) for 1 hour. The UV-A intensity applied to the sample was less than 30 μW/cm2. Next, the cells were vacuum-filtered and trapped onto a 13-mm ø polycarbonate filter with a 0.8-μm pore size (K080A013A, Advantec). The filter was incubated for 30 minutes at room temperature in hybridization buffer (0.1% IGEPAL CA-630 [Sigma] and 25 μg/ml Poly-A [Sigma] in 5× SET). After removal of the buffer by vacuum filtration, the filter was incubated for 3 hours at 45°C in hybridization buffer containing 500 ng/ml of the FITC-labeled probe. After being washed 2 times with 5× SET, the filter was mounted onto a nonfluorescent slide-glass, immersed in nonfluorescent immersion oil (Olympus), and observed by fluorescence microscopy (BH2-RFC, Olympus). FITC fluorescence emitted from the bound probe was observed under narrow-blue light excitation (455–490 nm) with a band-pass emission guard filter (520–530 nm). The fluorescent micrographs were taken using a cooled digital camera (Penguin 600CL, Pixera)
Table 3

Dinophysis Cells (with collection sites and dates) or Cryptophyte Cultures (with strain names and culture sources) Used for FISH

Species name

Source (collection site and date, or strain name)

Dinophysis acuminata

Isolated from Kesennuma Bay, August 8, 2002

D. fortii

Isolated from Okirai Bay, May 21, 2002

D. norvegica

Isolated from Okirai Bay, May 6, 2003

D. tripos

Isolated from Okirai Bay, July 31, 2002

Geminigera cryophila

MBIC 10567 (Marine Biotechnology Institute Culture Collection)

Guillardia theta

CCMP 327 (Provasoli-Guillard National Center for Culture of Marine Phytoplankton)

Proteomonas sulcata

CCMP 327 (Provasoli-Guillard National Center for Culture of Marine Phytoplankton)

Plagioselmis sp

TUC-1 (Tsukuba University Culture Collection)

Teleaulax sp.

TUC-2 (Tsukuba University Culture Collection)

FISH Trial of Field-Collected Plankton

On June 19 and July 17, 2002, water samples were collected at depths of 10 and 20 m from a permanent station in Okkirai Bay, Iwate, Japan (see Koike et al., 2001) using a Van Dorn bottle. The water samples were immediately sieved through a nylon net with a 20-μm mesh size. A 5-ml portion of the filtrates was fixed with 45 ml of fixative and then stored at −20°C. Of this fixed sample, 25 ml was subjected to FISH using the D16P-1 probe. In addition, 4′,6-diamidino-2-phenylindole (DAPI; 0.5 μg/ml in 5× SET) was added to the filter after the final wash of the FISH procedure, then washed 2 times with 5× SET, and observed by fluorescence microscopy. Probe-hybridized particles on the filter were confirmed to be eukaryotes by the presence of a DAPI-labeled nucleus.

To determine the probe specificity, FISH using the D16P-1 probe was also conducted on Dinophysis cells within heterogeneous plankton. A plankton sample collected from the bay using a net with mesh size larger than 20 μm was fixed and processed for FISH and DAPI staining. In this case the sample was centrifuged (1500 g, 10 minutes) at each of the washing steps and mounted on a slide glass for observation.


The obtained sequences of pSSU rDNA from D. tripos and 2 cryptophytes (Plagioselmis sp. and Teleaulax sp.), as well as of rdcL from D. fortii; D. tripos, and 4 cryptophytes (Plagioselmis sp., Teleaulax sp., G. cryophila, and P. sulcata), were deposited in GenBank with the accession numbers listed in Table 1. The sequence of pSSU rDNA of D. tripos was identical to those of 3 previously reported Dinophysis (D. fortii, D. acuminata, and D. norvegica) (Takishita et al., 2002). The sequences of the rbcL gene were also identical in D. fortii and D. tripos.

Phylogenic trees based on pSSU rDNA and rbcL sequences are shown in Figures 1 and 2, respectively. In both trees Dinophysis spp. were positioned within the lineage comprising Plagioselmis sp., Teleaulax sp., and G. cryophyra with high bootstrap support (98%–100%). Specifically, Dinophysis spp. were closely related to Teleaulax sp. (94%–100% bootstrap supports). Alignments of partial pSSU rDNA and rbcL sequences from Dinophysis spp., Teleaulax sp., Plagioselmis sp., G. cryophila, G. theta, and P. sulcata are shown in Figure 3. The boxed regions in Figure 3 indicate the probe recognition sites. Compared with the sequences of D16P-1 (22 bp) and DrbcL-1 (20 bp), there were 1 to 5 and 5 to 7 substitutions, respectively, within the corresponding regions of cryptophytes other than Teleaulax sp.
Figure 1

ML phylogeny of pSSU rDNA from Dinophysis spp. and cryptophytes (boxed area). The rhodophytes Porphyra purpurea and Palmaria palmata were used to root the tree. Numbers at the nodes refer to the percentage (50% or more) of bootstrap support in NJ and MP analyses.
Figure 2

ML phylogeny of rbcL from Dinophysis spp. and cryptophytes (boxed area). The rhodophytes Porphyra purpurea and Palmaria palmata were used to root the tree. Numbers at the nodes refer to the percentage (50% or more) of bootstrap support in NJ and MP analyses.
Figure 3

Aligned nucleotide sequences of partial pSSU rDNA and rbcL for Dinophysis spp., Teleaulax sp., Plagioselmis sp., Geminigera cryophila, Guilardia theta, and Proteomonas sulcata. Boxed regions indicate the probe recognition sites of D16P-1 and DrbcL-1 (see Table 2).

The designed FITC-labeled genetic probes (Table 2; D16P-1 for pSSU rRNA and DrbcL-1 for rbcL mRNA) were used for FISH with 4 species of Dinophysis and 5 species of cryptophytes. FISH and observation protocols were optimized using the G. cryophila versus G16P-1 probe (Table 2), which was designed to specifically recognize the pSSU rRNA. This probe was also used as a negative control for Dinophysis. Fluorescent micrographs of Dinophysis spp. and cryptophytes treated with these 3 genetic probes are shown in Figure 4. Probes D16P-1 and DrbcL-1 specifically hybridized only to Dinophysis plastids, and G16P-1 hybridized only to G. cryophila plastids. Positive FITC signals of both the D16P-1 and DrbcL-1 probes were not observed for Teleaulax sp., which has corresponding gene sequences identical to those of Dinophysis. UV irradiation to reduce autofluorescence from phycobilin did not affect the fluorescent signal when using the G. cryophila versus G16P-1 probe or the Dinophysis spp. versus D16P-1 or DrbcL-1 probes. Thus the negative reaction of the probes to Teleaulax sp. may be due to the poor physiologic state (e.g., low ribosome density and low mRNA level) of this slow-growing culture.
Figure 4

Fluorescent micrographs from FISH of Dinophysis spp. and cryptophytes. The probe names are shown at the top. Results for Teleaulax sp., Plagioselmis sp., and Proteomonas sulcata were the same as those for Guillardia theta, and are therefore omitted. Bars = 20 μm.

FISH trials with field-collected samples revealed the occurrence of D16P-1 probe-hybridized nanoplankters. These plankters were eukaryotic microalgae with nuclear fluorescence due to DAPI staining, and appeared to be cryptophytes by their cell shapes and sizes (Figure 5 a, b). They could be differentiated from other DAPI-stained prokaryotic or eukaryotic nanoplankters by their obvious FITC fluorescence (Figure 5, b). Also, Dinophysis cells in the heterogeneous sample were clearly discerniable from other plankters by probe binding (Figure 5, c, d).
Figure 5

Fluorescent micrographs of FISH (D16P-1) for field-collected plankton. The left column shows DAPJ fluorescence under UV excitation (a, b) or transmitted light micrographs (c, d), and the right shows probe hybridized cells, indicated by FITC fluorescence. a: Probe-hybridized eukaryotic cells that appear to be cryptophytes. DAPI fluorescence shows the presence of a nucleus (arrow), and the probe binds to the plastidal area. b: Probe-hybridized cryptophytes can be distinguished from other DAPI-stained plankters. c: The probe specifically recognizes Dinophysis fortii cells (black arrows) from other dinoflagellate or protist cells. d: Recognition occurs even in the heterogeneous samples.


We have previously demonstrated that, on the basis of the phylogeny of pSSU rDNA, the plastids of Dinophysis have phylogenetic affinity with those of the cryptophyte G. cryophyra (Takishita et al., 2002). Recently, phylogenetic analyses of nuclear and nucleomorph SSU rDNA from cryptophytes have revealed that G. cryophyra constitutes a robust monophyletic group with 2 genera, Teleaulax and Plagioselmis (Deane et al., 2002; Hoef-Emden et al., 2002). Therefore, we investigated the evolutionary relationship among the plastids in Dinophysis, Geminigera, Teleaulax, and Plagioselmis by using phylogenetic analyses of pSSU rDNA and rbcL. We found that plastids of Dinophysis are more closely related to those of Teleaulax than those of Geminigera and Plagioselmis.

Although phototrophic Dinophysis species are now recognized as mixotrophic, it appears that photosynthesis contributes a great deal to their active growth because they contain numerous plastids and few food vacuoles during the blooming period (Koike, 2002). Thus monitoring plastid density and photosynthetic activity should allow prediction of Dinophysis blooming and subsequent outbreak of DSP. For this reason, in the current studies we developed molecular probes that specifically recognize pSSU rRNA and rbcL mRNA of the Dinophysis plastid. Using in situ hybridization (Miller and Scholin, 2000), we specifically identified plastids of toxic Dinophysis. Furthermore, the intensities of the hybridization signal for pSSU rDNA and rbcL can be used to estimate their transcriptional and photosynthetic activities, respectively.

These probes should help answer the question of whether the Dinophysis plastid is actually a genuine organelle of Dinophysis or one that is temporarily acquired from the environment. Despite the high evolutionary rate of sequences from fully established dinoflagellate plastids (Zhang et al., 1999, 2000; Barbrook and Howe, 2000; Tengs et al., 2000), the current results and our previous study (Takishita et al., 2002) show that 4 species of photosynthetic Dinophysis share a plastid with identical pSSU rDNA sequences. Moreover, the pSSU rDNA sequence was 99.8% identical with that of an environmental clone (OCS20) derived from ultra-sized (less than 10 μm in diameter) microalgae collected in the Pacific Ocean (Rappe et al., 1998). Because all known photosynthetic species of Dinophysis are more than 30 μm in length, it is unlikely that the OCS20 clone is from a Dinophysis species. Rather, OCS20 is more likely to be from a cryptophyte, many of which are less than 10 μm in size. This cryptophyte appears to be a likely source for Dinophysis plastids.

These molecular data support the hypothesis that the Dinophysis plastids are derived from temporarily acquired cryptophytes by a process termed kleptoplastidy (Larsen, 1992; Laval-Peuto, 1992; Schnepf and Elbrächter, 1992). Nonetheless, there have been many opposing opinions based on the differences in plastid morphologies. The possibility of kleptoplastidy is supported by other recent genetic information on the Dinophysis plastids. Hackett et al. (2003) reported that pSSU rDNA sequences from Dinophysis collected at different times of the year and from different locations were monophyletic but also showed significant variation. They claimed that these data support the contention that Dinophysis plastids are permanent. However, as they also observed, we cannot exclude the possibility that the polymorphism of pSSU rDNA indicates prey cryptophyte diversity.

Using FISH with the molecular probes developed here, we were able to distinguish plastids from Dinophysis and closely related cryptophytes. This technique should help clarify the biology of Dinophysis and the mechanism of DSP. In addition, we found probe-hybridized cryptophyte cells in field-collected plankters, which are likely the source of the Dinophysis plastid. To obtain more direct evidence of kleptoplastidy, investigations into the occurrence of source cryptophyte cells and the simultaneous increase of Dinophysis plastid density are needed. Ultimately, the isolation and establishment of the source cryptophyte culture and a feeding experiment for Dinophysis should be performed. Our probes should facilitate these explorations.


This work was supported in part by a grant-in aid (4704015 to K.K.) for Scientific Research from the Ministry of Education, Science, Sports, and Culture of Japan, and performed as a part of the Iwate Prefecturel Fund for Leading Investigation and Research. We also wish to acknowledge Dr. Shigetoh Miyachi, an international editor of this journal.

Copyright information

© Springer Science+Business Media, Inc. 2005