Development of a Dinoflagellate-Oriented PCR Primer Set Leads to Detection of Picoplanktonic Dinoflagellates from Long Island Sound

We developed dinoflagellate-specific 18S rRNA gene primers.PCR amplification using these oligonucleotides for a picoplanktonicDNA sample from Long Island Sound yielded 24 clones, and allbut one of these clones were dinoflagellates primarily belongingto undescribed and Amoebophrya-like lineages. These resultshighlight the need for a systematic investigation of picodinoflagellatediversity in both coastal and oceanic ecosystems.

INTRODUCTION

Top

Introduction

Dinoflagellates have received considerable attention due totheir ecological and economical significance and their remarkablecytological and genetic features (6, 7, 20). However, our knowledgeof the species diversity of these organisms remains limitedeven though novel living (2, 3, 10, 14) and fossil lineages(4, 19) have been discovered. Knowledge of the diversity of"small" dinoflagellates is particularly deficient. The recentdiscovery of ultraplanktonic (<5-µm) and picoplanktonic(<3-µm) dinoflagellates in Antarctica and the PacificOcean (13, 15), respectively, is the first demonstration ofa rich biodiversity of small dinoflagellates that have escapedroutine microscopic detection. A better understanding of dinoflagellatebiodiversity requires targeted approaches, particularly forpicoplanktonic species.

Development of dinoflagellate-oriented primers.

Top

Development of dinoflagellate...

Based on a large database of nuclear small-subunit (18S) rRNAgenes, we designed PCR primers that target dinoflagellates.A total of 140 18S rRNA gene sequences, including sequencesfrom dinoflagellates, diatoms, chlorophytes, haptophytes, cryptophytes,and other algae, were obtained from GenBank and were alignedusing ClustalW1.8; 11 of the dinoflagellate species were sequencedin this study (GenBank accession no. DQ388456 to DQ388466).Regions unique to dinoflagellates were used to design threeforward and two reverse PCR primers (Table 1), which were pairedwith previously described eukaryotic 18S rRNA gene universalprimers (22) for DNA amplification.

View this table:
[in this window]
[in a new window]
TABLE 1. 18S rRNA gene PCR primers designed in the present study

The primers were tested with 33 genera of cultured dinoflagellates(35 species, including Oxyrrhis marina), as well as nine othertaxa (Table 2). Algal cultures were grown in f/2 medium (28 ${per thousand}$ or 15 ${per thousand}$ salinity), cells were harvested, and DNA was purifiedas previously described (23). Briefly, after cell lysis in 1ml DNA buffer (100 mM EDTA [pH 8.0], 0.5% sodium dodecyl sulfate,200 µg ml^–1 proteinase K), DNA was purified usingDNA Clean and Concentrator columns (Zymo Research, Orange, CA).With these DNA samples as templates, PCR was performed usingfive combinations of the primers, as follows: primers 18ScomF1and Dino18SR1 (expected product size, 0.65 kb), primers 18ScomF1and Dino18SR2 (0.92 kb), primers Dino18F1 and 18Scom R1 (1.60kb), primers Dino18F2 and 18Scom R1 (0.92 kb), and primers Dino18F3and 18S com R1 (0.90 kb). All primer sets except the Dino18SF2-18ScomR1set exhibited specificity for dinoflagellate 18S rRNA genes,which allowed amplification from most taxa examined (Table 2).The only exceptions were O. marina (often referred to as anancestral dinoflagellate [17] or a predinoflagellate [18]) andExuviaella cassubica, for which all primers failed, likely becausetheir 18S rRNA gene sequences are significantly divergent. Ofthe four pairs of dinoflagellate-specific primers, 18ScomF1-Dino18SR1and Dino18F1-18ScomR1 showed superior sensitivity and were ableto detect 1 to 10 cells/reaction mixture for most of the dinoflagellatestested. Primers Dino18F1 and 18Scom R1 was chosen for furtherstudy because they spanned a larger 18S rRNA gene region (1.6kb).

View this table:
[in this window]
[in a new window]
TABLE 2. Specificity of primers (in five pairs) with various protists

Detection of picodinoflagellates in Long Island Sound.

Top

Detection of picodinoflagellates...

Three water samples collected on 8 September 2005 along theboat dock of the University of Connecticut Avery Point campus(Long Island Sound) were combined and mixed. Microscopic examinationof a subsample revealed that phytoplankton lineages such asNitzschia, Navicula, Chaetoceros, and Eucampia were dominant.A 2-liter subsample was prescreened (100-µm mesh), andthis was followed by passage through a 3-µm polycarbonatemembrane (Nuclepore, Pleasanton, CA) under a low vacuum pressure(<10 lb/in²). One liter of the filtrate was collected andfiltered onto a 0.2-µm-pore-size, 47-mm-diameter polycarbonatemembrane (Nuclepore). The filter membrane was cut into smallpieces using sterile scissors, placed in a 1.5-ml microcentrifugetube, and stored at –80°C until DNA extraction. Toexamine whether any large plankton were present in the sample,100 ml of the 3-µm filtrate was filtered onto a 0.2-µm-poresize, 25-mm-diameter black Poretics polycarbonate membrane (OsmonicsInc., Minnetonka, MN), fixed with 1% paraformaldehyde, and stainedwith 0.005% acridine orange. Observation with an epifluorescencemicroscope (Olympus BX51) revealed only small organisms exceptfor one large cell (length, $~$ 12 µm) that appeared to bea Heterocapsa cell.

DNA extracted from the <3-µm plankton sample as describedabove was used for PCR amplification with primers Dino18F1 and18Scom R1. The PCR was performed using Takara ExTaq DNA polymerase(Takara Mirus Bio, Madison, WI) following the manufacturer'sinstructions, with an annealing temperature of 58°C. Theamplicon was cloned into a TA vector, and 24 of the clones wererandomly selected and sequenced (GenBank accession no. DQ386737to DQ386760). A BLAST search showed that all 24 clones wererelated to dinoflagellates. These sequences were aligned usingCLUSTAL W (1.8) with a Kimura two-parameter model (11). A maximumlikelihood tree was inferred using PHYML (5) based on the GTR+ I + ${Gamma}$ model of nucleotide substitution, which was identifiedas the best-fit model by Modeltest3.7 (16). Species of chlorophytes(as an outgroup), cryptophytes, diatoms, ciliates, apicomplexa,and dinoflagellates (a total of 101 taxa) were initially includedin the analysis to obtain a global view of the distributionof the dinoflagellate-like clones. One (clone 12) clusteredwith the recently recognized parasitic ciliate Cryptocaryonirritans (21) with robust bootstrap support, whereas the other23 clones clustered with dinoflagellates. In further analyseswe focused on dinoflagellates with Toxoplasma gondii (apicomplexa)as the outgroup. The results indicated that 12 of the 23 dinoflagellate-likeclones (clones 3, 5 to 7, 9 to 11, 16, 18, 19, 23, and 24) weremembers of a clade of undescribed eukaryotes comprising previouslyisolated picoeukaryotes from the Pacific Ocean (Fig. 1). Thisclade received strong bootstrap support and diverged beforeall known dinoflagellates (Fig. 1), suggesting that it couldbe an ancestral clade of dinoflagellates or a lineage that isintermediate between dinoflagellates and apicomplexa. Of theremaining 11 clones, 7 formed a clade with the parasitic dinoflagellateAmoebophrya. Four of these seven clones (clones 4, 8, 17, and21) clustered tightly with Amoebophrya sp., which infects variousdinoflagellate species, and the other three (clones 13, 15,and 20) appeared to be more ancestral. These assignments hadmoderate to strong support. One of the four other clones (clone1) was related to a Gymnodiniales-dominated clade and clusteredtightly with Gymnodinium simplex and Symbiodinium microadriaticum,and two clones (clones 12 and 14) clustered with Gymnodiniumcatenatum/Lepidodinium viride. The last clone (clone 2) appearedto be a Heterocapsa species, likely related to the large cellobserved under the microscope (see above). To our knowledge,this is the first documentation of the presence of picodinoflagellatesin coastal waters.

View larger version (41K):
[in this window]
[in a new window]
FIG. 1. Maximum likelihood tree of dinoflagellates constructed using Phyml V2.4.4. T. gondii was used as the outgroup. The parameters of the GTR + I + ${Gamma}$ model of nucleotide substitution were estimated as follows: the frequencies of nucleotides were 0.23931 for A, 0.31386 for T, 0.18944 for C, and 0.25739 for G; the rate parameters were 1.1979 for A ${leftrightarrow}$ C, 3.3746 for A ${leftrightarrow}$ G, 1.3953 for A ${leftrightarrow}$ T, 0.9178 for C ${leftrightarrow}$ G, 5.1569 for C ${leftrightarrow}$ T, and 1.0 (fixed) for G ${leftrightarrow}$ T; the fraction of invariant sites was 0.2824; the shape parameter ( ${alpha}$ ) was 0.6838; and the likelihood value (loglk) was –34037. The robustness of species groups was assessed using the bootstrap approach with 100 resamplings. The thickest lines indicate bootstrap values of >80%, the thick lines indicate bootstrap values of >50%, and the thin lines indicate bootstrap values of <50%. Scale bar = 0.1 substitution per base.

An increasing number of studies have indicated that field DNAsamples derived from mixed microbial assemblages are prone toformation of PCR chimeras by cDNA strands from different species,thus creating artificial, novel genes (1, 9). To examine whetherany of the sequences obtained in this study was a chimera ofdifferent species in the water sample, sequences of the field-derivedclones, as well as previously reported dinoflagellate sequencesused in this study, were analyzed using the program Bellerophon(8). The details of the analysis are shown in Table S1 and Fig.S1 in the supplemental material. The results indicated thatmost clones appeared to be nonchimeric sequences; the only exceptionwas one clone (clone 20) that was ambiguous.

Wide distribution of diverse and predominantly unknown picodinoflagellates.

Top

Wide distribution of diverse...

Strikingly, one-half of the 24 clones identified in this studywere closely related to undescribed picodinoflagellate lineagesfrom the Pacific Ocean. Furthermore, the seven clones that clusteredwith the parasitic dinoflagellate Amoebophrya were allied withlineages from the Pacific Ocean. Separate analyses that includedthe shorter sequences from Antarctic deep water (13) revealedthat these sequences also were closely related to the Amoebophrya-likelineages detected in the Pacific Ocean and Long Island Sound(results not shown). In addition, lineages close to Gymnodinium/Symbiodiniumhave also been found in the Pacific Ocean. Therefore, the distributionof picodinoflagellates extends from the open ocean to coastalwaters, suggesting that small dinoflagellates are cosmopolitan.Moreover, the similarity between the dinoflagellate speciescomposition in Long Island Sound surface water and the dinoflagellatespecies composition in the aphotic zone of Antarctica and thePacific Ocean (13, 15), especially the presence of Amoebophrya,suggests that heterotrophic and parasitic dinoflagellates maybe more common than currently thought. The ease with which novellineages were isolated also suggests that dinoflagellate diversityhas been underestimated. However, this observation should notbe surprising given that even in the larger cell size rangenew species have frequently been discovered (e.g., Stoeckeria,Takayama, Polarella, and Pfiesteria [2, 3, 10, 14]). The dearthof information regarding the lineages of small dinoflagellatesis a result of a lack of targeted analyses.

Interestingly, dinoflagellates have normally been categorizedas larger phytoplankton. Recently, LaJeunesse et al. (12) suggestedthat Symbiodinium (4 to 13 µm) is the lineage containingthe smallest dinoflagellates. It is clear now that smaller dinoflagellatesare present in both oceanic and neritic waters and that thesmallest organism is likely yet to be described. Once the widespreaddistribution of these picodinoflagellates is verified by cultureand morphological analyses, their role in the microbial loopin the world's ocean can begin to be assessed. In addition,given the pressing need for a complete dinoflagellate genomesequence, free-living picodinoflagellates like those describedhere are likely to be the best candidates for such an attempt.The "normal-size" taxa have human-size (or much larger) genomes(12), which makes the use of current sequencing approaches withthese organisms infeasible.

Concluding remarks.

Top

Concluding remarks.

This is the first report of dinoflagellate-oriented primersand the first documentation of the presence of picodinoflagellatesin coastal waters. Although the number of clones sequenced waslimited, our findings nevertheless revealed the high level ofdiversity and the dominance of novel lineages of picodinoflagellates.Therefore, larger-scale targeted analyses of different oceanicregions are essential for determining the true biodiversityof these taxa.

Nucleotide sequence accession numbers.

Top

Nucleotide sequence accession...

The nucleotide sequences of the 24 field-retrieved picoplanktonclones have been deposited in the GenBank database under accessionnumbers DQ386737 to DQ386760, and the nucleotide sequences ofthe 11 dinoflagellate cultures have been deposited in the GenBankdatabase under accession numbers DQ388456 to DQ388466.

ACKNOWLEDGMENTS

We are grateful to Terry Gaasterland of Scripps Genome Centerfor access to her computing facility for the phylogenetic analysisand to Christopher Dungan of the Maryland DNR Cooperative OxfordLaboratory for kindly providing Perkinsus species.

This work was supported by National Science Foundation grantsDEB-0344186 (to S.L. and H.Z.) and DEB-0107754 and EF-0431117(to D.B.).

FOOTNOTES

* Corresponding author. Mailing address: Department of Marine Sciences, University of Connecticut, Groton, CT 06340. Phone: (860) 405-9168. Fax: (860) 405-9153. E-mail: senjie.lin{at}uconn.edu.

Supplemental material for this article may be found at http://aem.asm.org/.

REFERENCES

Top