Index References

Ribosomal DNA (rDNA) genes, including the internal transcribed spacer (ITS) regions, are the most frequently used target for phylogenetic studies, because not only do they occur in all living organisms, but they typically are present in several copies which are distributed throughout the genome (Harmsen and Karch, 2003). Particularly, eukaryotic nuclear rDNA is tandemly organized, with copy numbers up to the order of 10,000 (Schlötterer, 1998). Because of their microscopic sizes and fragile bodies, species of microalgae, like bacteria, are commonly identified using these gene sequences (Ki et al., 2004; Ki and Han, 2006). At present, more than 20,000 strands of rDNA sequences from dinoflagellates have been identified and deposited into public databases (i.e. EMBL/DDBJ/GenBank).

The ichthyotoxic Cochlodinium polykrikoides Margalef is a harmful dinoflagellate that is one of the most frequent causes of fish kills (Kim et al., 1999; Ahn et al., 2006). Over the last two decades, outbreaks of C. polykrikoides in Korean coastal waters have significantly increased in frequency, severity, and duration (Kim et al., 1999). These dinoflagellates closely resemble the related dinoflagellates Akashiwo sanguinea, Gymnodinium catenatum and G. impudicum. Because of this close resemblance, Cho et al. (2001) tried to discriminate C. polykrikoides from others using DNA sequences. They suggested that the ITS1 region was valuable for this taxonomic discrimination. Independently, we have analyzed the sequence of C. polykrikoides to identify nuclear DNA markers that will facilitate the taxonomic identification of this harmful microalgae. Recently, we reported that the partial 18S rDNA was sufficient to discriminate C. polykrikoides from other related dinoflagellates (Ki et al., 2004). In the course of analyzing dinoflagellate rDNA sequences, we found consistent patterns of extremely long PCR fragments in the ITS and 5.8S rDNA sequences of C. polykrikoides compared to other dinoflagellates. This study reports the complete nucleotide sequences of the ITS and 5.8S rRNA genes from C. polykrikoides and its close relatives. In addition, we report the characterization of the long ITS in C. polykrikoides and our prediction of the secondary structure of ITS1.

Clonal cultures of Cochlodinium polykrikoides (CcPk02, 03, 05, 06), including Akashiwo sanguinea (GnSg03) were kindly provided by Dr. M. Chang of the Korean Ocean Research and Development Institute. An additional strain (CCMP414) of Gymnodinium catenatum was obtained from a commercial source (the Provasoli-Guillard National Center for Culture of Marine Phytoplankton, CCMP). All cultures were grown in f/2 medium pH 8.2 at 15°C, with a 12:12-h light:dark cycle and a photon flux density of ca. 65 μmol photons/m²/s. Mid-exponential batch cultures were harvested by centrifugation at 3,000g for 10 min. The concentrated cells were mixed with 100 μL of 1X TE buffer (10 mM Tris-HCl, 1 mM EDTA; pH 8.0), and stored at –20°C until DNA extraction. Genomic DNA was isolated from the stored cells using the DNeasy Plant mini kit (Qiagen, Valencia, CA) according to the manufacture’s instructions.

PCR was used to amplify complete ITS regions including 5.8S rDNA with a set of PCR primers (18F1641, 28R01) as shown in Fig. 1. In addition, three Cochlodinium polykrikoides-specific PCR primers (Fig. 1) were designed based on the 18S and 28S rDNA sequences determined previously (Ki et al., 2004), by comparing other dinoflagellate sequences available in GenBank. Overall specificities were tested by phylogenetic analyses of most dinoflagellate sequences retrieved from the database. Additional tests were performed on BLAST search by comparing these sequences against all publicly available data. Fifty-microliter PCR reactions were carried out in 1X PCR buffer (10 mM Tris-HCl, 50 mM KCl, 1.5 mM MgCl₂, 0.001% gelatin; pH 8.3) with < 0.1 μg genomic DNA template, 200 μM each of the four dNTPs, 0.5 μM of each primer and 1 Unit Taq polymerase (Promega, Madison, WI, UAS). Using a UNO-II Thermoblock (Biometra, Göttingen, Germany), PCR thermocycling parameters were as follows: 95°C for 5 min; 35 cycles of denaturation at 95°C for 20 s, annealing at 55°C for 30 s and extension at 72°C for 60 s and a final extension at 72°C for 5 min. The PCR products (2 μL) were analyzed by 1.5% agarose gel electrophoresis according to standard methods.

View Details

Fig. 1. A map of Cochlodinium polykrikoides ITS and 5.8S rDNA and PCR products. (A) diagrammatic presentation of C. polykrikoides ITS and 5.8S rDNA, indicating the long internal repeats in ITS1. Solid boxes refer to rDNA genes and open boxes are represented repeat units, respectively. The R1 to 6 in ITS1 refer to six internal repeats, to which individual sequences are aligned and represented below. Arrows indicate the position of primers used in this study. Forward primers, Cp18-3F3 (5’-TTAGATGTTC TGGGCTGCACG-3’, positioned at 1446-1466), 18F1641 (5’-GAGGAAGGAG AAGTCGTAACAAGG-3’, 1622-1641), ITSF01 (5’-TCCCTGCCCTTTGTACACAC-3’, 1747-1770), are shown above 18S rDNA, near the priming site. Reverse primers, ITSR01 (5’-TCCGCTTACTTATATGCTTAAATTCAGC-3’, 29-56), Cp28R2 (5’-GTTGGCGT TGCATTTCGAGAC-3’, 148-168), Cp28-3R2 (5’-ACGGAGGGCTGCAGATTGAC-3’, 227-246) and 28R01 (5’-AAACTTCGGAGGGAACCAGCTAC-3’, 1019-1041), are shown below 28S rDNA. Each primer position is based on the complete 18S and 28S rDNA sequence of C. polykrikoides (Accession no. AY347309). (B) PCR products amplified with a C. polykrikoides genomic DNA (CcPk06) and four different primer sets. Each PCR product is expected to be, in order, 1579 bp, 1691 bp, 1769 bp and 2564 bp in length, respectively. (C) PCR products amplified with various combinations of both different templates and two primer sets (ITSF01 + ITSR01, Cp18-3F3 + ITSR01). Genomic DNAs used here are from each strain of C. polykrikoides (CcPk05: Lane 1, 5; CcPk06: lane 2, 6), Akashiwo sanguinea (GnSg03: lane 3, 8) and Gymnodinium catenatum (CCMP 414: lane 4, 9), respectively. M1 and M2 indicate 100-bp and lamda hind III DNA markers.

ITS1 in Cochlodinium polykrikoides contained long repetitive sequences, which apparently formed strong hairpin structure in sequencing reactions. Consequently, the helix structure might hinder elongation of the enzyme at the base of the hairpin and cause the enzymatic reaction to stop there or render unreadable, weakened signals. Hence, we could not analyze the repetitive regions with the dye terminator sequencing method. In the present study, using the dye-labeled primer sequencing method, we were able to determine the complete sequences of the ITS regions in C. polykrikoides. The sequencing protocol has been described in detail (Ki et al., 2004). In brief, DNA sequencing was performed using PCR products and near infrared dye (IRD)-labeled primers. Identical PCR and nested primers were used for sequencing reactions on each amplicon containing entire ITS, 5.8S and partial LSU rDNAs. PCR products (3–6 μL) were subjected to DNA cycle sequencing using a ThermoSequenase^TM version 2.0 Cycle Sequencing Kit (USB, Cleveland, OH, USA) in the presence of 1.5 picomoles of sequencing primers. The four base-specific reactions were initially subjected to 94°C for 1 minute, followed by 40 cycles consisting of 94°C for 20 s, 50°C for 30 s, and 72°C for 60 s in the UNO-II Thermoblock. When complete, reactions were stopped by adding 4 μL of IR2 stop/loading buffer (Licor®, NE, USA), and the products were heat-denatured and analyzed on a Model 4200 Dual Dye Automated Sequencer (Licor), according to the manufacturer’s instructions. All of the sequences obtained here were completely assembled with the Sequencher3.0 software (Gene Codes, MI, USA). All DNA sequences reported here have been deposited in GenBank database as accession numbers AY347309, AY831411, DQ779984-6 and DQ779990.

A secondary structure of ITS1 was estimated using the program Mfold, version 3.2 (http://www.bioinfo.rpi.edu/applications/mfold/old/rna/), according to Zuker and Turner (Zuker, 2003). With the default option (e.g. temperature setting, T = 37°C), Mfold predicted 7 secondary structures from ITS1 sequence including a single repeat unit and 27 secondary structures from the entire ITS1 sequence, respectively. Using different parameter settings of, for example, temperature (T = 10°C, T = 20°C, T = 37°C), did not affect the general architecture, but did result in different energy levels for the secondary structures. The secondary structure models inferred here were redrawn with the software RNAViz (De Rijk and De Wachter, 1997), a versatile program developed to draw secondary structures of molecules in a fast and user-friendly way. Among them, we finally selected a general model of secondary structure for C. polykrikoides ITS1.

PCR amplification following the typical PCR protocol generated relatively long PCR fragments (approximately 1,200 bp) from entire ITS rDNA regions of Cochlodinium polykrikoides. DNA sequencing revealed the PCR products to be 1,166 nucleotides in length. Cryptically, this contained six long internal repeat sequences (Fig. 1A). To find general ITS characteristics and to determine the relationship between the ITSs of C. polykrikoides and its relatives, we determined the complete sequences of ITS1, 2, and 5.8S rDNA from Gymnodinium catenatum and Akashiwo sanguinea, which are considered phylogentically as members of the closest Cochlodinium relative (Cho et al., 2007). With these sequences and available GenBank data (e.g. A. sanguinea, AY831412; G. corii, AF318226; G. maguelonnense, AF318225; G. mikimotoi, AF318224), we found that the armoured dinoflagellate species have around 200 bp in each ITS1 and ITS2, and all species had either 160 bp or 161 bp of 5.8S rDNA. The complete sequence length of the ITS region was, therefore, less than 560 bp in most dinoflagellates. In contrast, the ITS regions of C. polykrikoides was considerably longer. The ITS of C. polykrikoides consisted of 813 bp of ITS1, 160 bp of 5.8S rDNA, and 193 bp of ITS2. Among these regions, the ITS1 was extremely long and thus contributed substantially to the length of the PCR fragments. Additional database searches (e.g. EMBL/DDBJ/GenBank) revealed that the ITS1 sequence of C. polykrikoides was the longest among dinoflagellate sequences recorded to date. The majority of dinoflagellate ITS1 sequences deposited in the database were less than 300 bp in length, and many of them were shorter than 250 bp.

The extreme length of the PCR products was confirmed through the application of C. polykrikoides specific-primers (Fig. 1A) to be PCR-amplified the complete ITSs regions. PCRs, employing combinations of the primers, apparently generated bands of each expected size (Fig. 1B). In addition, when a universal primer set (ITSF01 + ITSR01) was applied to the closely related dinoflagellates A. sanguinea and G. catenatum, the PCR product patterns differed comparatively in size (Fig. 1C). However, PCR, employing C. polykrikoides specific primers, did not generate any fragments from the relatives (Fig. 1C).

For intraspecific variation of ITS1, we investigated four Cochlodinium polykrikoides stains isolated from geographically different coastal waters in Korea (CcPk02, Tongyoung; CcPk03, Narodo; CcPk05, Hakdong; CcPk06, Sarangdo). Upon comparison, the sequences from these strains were 100% identical to one another. This suggested that the long sequences in ITS1 might be a common feature of populations of C. polykrikoides rather than a mutation of particular strains. Previously, however, Cho et al. (2001) reported 243 bp of full ITS1 sequences (GenBank access. no. AF208248; isolate, CP-1) from C. polykrikoides. By comparison, this sequence was completely identical to the sequences reported here, indicating that all strains were apparently the same species of C. polykrikoies. Although the reasons for the disparity with our results are unclear, one possible reason is that the previous sequence represented the incomplete sequence of ITS1. Indeed, the GenBank record for the sequence shows that it is not full-length sequences of the ITS regions. Another possible explanation might be intragenomic rDNA diversity in ITS1 of C. polykrikoides. That is, the 18S–26S rDNA as a multigene family is subject to concerted evolution (Elder and Turner, 1995), and few variations have been reported. These variations are thought to stem from an individual unequal crossing over and subsequent gene conversion (Hugall et al., 1999). Although the chances of gene conversion are low and few cases have been reported (Hugall et al., 1999), the finding that the PCR of the ITS from C. polykrikoides does not amplify two different-sized bands (see Fig. 1B) unsupports this possibility.

The ITS1 of Cochlodinium polykrikoides contains a tract of 101 nucleotides, which occurs six times in tandem, as noted previously. The six repeated elements were exactly identical, except for two nucleotides, in complete sequences (Fig. 1A). The ITS1 of C. polykrikoides can therefore be separated into three distinct regions: the 5’ end (122 nucleotides), the six parallel repeats, and the 3’ region (85 nucleotides). In addition, the tandem repeats in C. polykrikoides were present in the middle of the spacer positioned between the 3’ and 5’ ends of ITS1. This is consistent with the previous studies (van Herwerden et al., 1999; von der Schulenburg et al., 2001). von der Schulenburg et al., (2001) reported that long repetitive elements were always confined to the middle of the spacer in coccinellid species. They suggested that the reason for this might be that the 39- and 59-end regions of ITS1 might have functional importance, whereas the lack of such constraints in the middle of the spacer might have favored the rise of repetition in this region. From a structural perspective, Bakker et al. (1995) reported that the ITS is believed to have few evolutionary constraints and might be expected to evolve at or near the neutral level.

A predicted model of the secondary structure of the ITS1 pre-RNA transcripts conforms well to the hairpin structure present in C. polykrikoides (Fig. 2). In addition, the predicted secondary structures clearly indicated that there are three distinct regions within ITS1. Each single unit of each repeat sequence represents a palindromic DNA sequence (see DeBoer and Ripley, 1984), as shown in the dotted box of Fig. 2. Thus these sequences can form hairpins as a double helix (Fig. 2) when pre-rRNA genes express these regions. In addition, the long ITS1 containing the entire repeats should form a completely helical structure as a long chain (Fig. 2). In fact, these hairpin structures are believed to play a role in processing signals. Recently, Campbell et al. (2005) pointed out that small hairpins, such as those formed by these regions, are functional and that natural selection would favor GC-rich sequences to stabilize the stems. Actually, C. polykrikoides has considerably high G/C content (61.4%) in ITS1, particularly in repeat units (63.9%) when compared to other dinoflagellates: A. anguinea, 56.2%; G. catenatum, 50.0; G. impudicum, 53.7%. These findings suggest that the hairpin in C. polykrikoides might be structurally stable and play an important role in ribosome biogenesis through the processing of the pre-RNA.

View Details

Fig. 2. A putative secondary structure of the ITS 1 pre-RNA transcripts. The dotted box marks the structure predicted with single-repeat unit, which forms palindrome-like sequences. The sequence is written clockwise from 5’ to 3’ and alphabet “R” represents each repeat unit used in Fig. 1.

To date, many studies have investigated ITS1 variations (van Herwerden et al., 1999; Harris and Crandall 2000; von der Schulenberg et al., 2001; Campbell et al., 2005). Commonly, short tandem repeats (STR) are considered to be responsible for the length variation in ITS1. The presence of STR in fungi is well documented (Platas, 2001). In addition, these STR motifs have been found in various arthropods (Kumar, 1999; Harris and Crandall, 2000), and flowering plants (Liston et al., 1996). Previously, Levinson and Gutman (1987) reviewed the repetitive DNA sequences in light of STRs. However, to the best of our knowledge, no studies have identified STRs in ITS1 of dinoflagellates. Several mechanisms have been proposed to generate length variation. These include intra- and inter-strand recombinational effects, such as unequal crossing over; other mechanisms involve failures in the replication of DNA such as slipped-strand mispairing (Levinson and Gutman, 1987) or replication slippage (Pinder et al., 1998). In contrast to STR, long internal repeats are not frequent in ITS1 regions. Long internal repeats include repetitive elements with comparatively long repeat units, and they have only been reported in few taxonomic groups such as the trematode Paragonimus westermani (van Herwerden et al., 1999) and the dipteran Simulium damnosum (Tang et al., 1996). Previously, von der Schulenburg et al. (2001) found long tandemly repeated sequence motifs of greater than 60 bp in several insects Adalia, Harmonia and Psyllobora. More recently, Warberg et al. (2005) reported that the repetitive sequences (130 bp) in the ITS1 region of rDNA was consistently present in congeneric microphallid species (Trematoda: Digenea). In these cases, the generation of the repetitive sequences has been proposed to occur through the same mechanisms as STR: replication slippage, unequal crossing over, and biased gene conversion (e.g. Levinson and Gutman, 1987; Elder and Turner, 1995). However, these studies did not find any relationship between the long repeats and taxonomical groups. Furthermore, it is not known whether the long repeat units within the ITS1 region have any functional importance, nor is it known how they might have arisen and how they are maintained. Additionally, we found that C. polykrikoides has comparatively long internal repeat sequences in the ITS1, which is the first such repeats to be discovered in alveolates. At the present stage, the function and biogenesis of repeat sequences are not clear, and later more detailed studies are needed to improve understanding of both evolutionary variation and the molecular processes that lead to the generation of repetitive sequences.