Introduction

The majority of photosynthetic dinoflagellates possess plastids containing chlorophylls (Chl) a+c and a unique carotenoid, peridinin (Taylor 1980). The peridinin-containing plastids are most likely derived from a red algal endosymbiont. However, some dinoflagellate lineages are known to bear non-canonical plastids lacking peridinin which are classified into seven types according to Moestrup and Daugbjerg (2007). Traditionally, pigment compositions have been used to infer the origin of the non-canonical plastids; for example, the non-canonical plastids containing 19′-hexanoyloxyfucoxanthin, fucoxanthin, and alloxanthin advocated the evolutionary connections to the plastids in haptophytes, diatoms, and cryptophytes, respectively (Withers and Haxo 1975; Withers et al. 1977; Suzuki and Ishimaru 1992; Meyer-Harms and Pollehne 1998; De Salas et al. 2003; Bergholtz et al. 2005; Tamura et al. 2005). The origins of these non-canonical plastids inferred from pigment compositions were confirmed by phylogenetic analyses of plastid-encoded genes (e.g., Takishita et al. 1999, 2002; Tengs et al. 2000; Horiguchi and Takano 2006). Importantly, the remote relationship amongst the dinoflagellate (host) lineages bearing non-canonical plastids indicates that the plastid replacement events took place in separate branches in dinoflagellate phylogeny (Saldarriaga et al. 2001; Shalchian-Tabrizi et al. 2006).

The genus Lepidodinium bears non-canonical plastids containing Chl-a+b (Watanabe et al. 1987, 1990), but their pigment composition has not been fully settled. Lepidodinium plastids have been stated to contain prasinoxanthin, a carotenoid exclusively found in a few members of the green algal class Prasinophyceae, implying that this dinoflagellate lineage acquired its current non-canonical plastids through the endosymbiosis of a prasinophyte (Elbrächter and Schnepf 1996). Nevertheless, the presence of prasinoxanthin in L. chlorophorum was quoted as a personal communication in Elbrächter and Schnepf (1996), and therefore, needs to be re-examined. To elucidate the precise origin of the Lepidodinium plastids, it would be ideal if the pigment composition, cell morphology, and gene sequences are available for the same strain maintained in the laboratory. Recently, Laza-Martinez et al. (2007) called the popular hypothesis for the origin of the Lepidodinium plastids into question, as no sign of prasinoxanthin was detected in the L. chlorophorum culture. Laza-Martinez et al. (2007) admitted that the L. chlorophorum culture examined was unfortunately not mono-algal, and they did not provide morphological or sequence data for the particular strain. Furthermore, to our knowledge, the prasinoxanthin-lacking ‘L. chlorophorum’ strain has not been deposited in any culture collection.

Some of us have studied the evolutionary characteristics of the L. chlorophorum plastid by sequencing 11 plastid-encoded genes of L. chlorophorum strains (Takishita et al. 2008; Matsumoto et al. 2011a, b). The results showed that the plastids of the L. chlorophorum strains originated from one of “core chlorophytes,” which did not contain prasinoxanthin. These strains have been maintained as mono-algal cultures in the laboratory, and their morphologies have been well characterized. Thus, these strains can be used to address the pigment composition of L. chlorophorum. In this study, we investigated the pigment composition of the L. chlorophorum strain NIES-1868, and revealed that this strain apparently lacked prasinoxanthin. In addition, we newly determined the nucleus-encoded 18S ribosomal RNA (rRNA) genes of strains NIES-1867 and NIES-1868, as well as those of two another Lepidodinium strains. The rRNA sequences from the four Lepidodinium strains determined in this study and four of those deposited under the name of L. chlorophorum or L. viride in GenBank database shared nucleotide identity of ≥99 %, and formed a tight clade in our phylogenetic analyses. Thus, the Lepidodinium species, of which 18S rRNA sequences bear high identities to that of strain NIES-1868, most likely lack prasinoxanthin.

Materials and methods

Algal cells

Two strains of L. chlorophorum (NIES-1867 and NIES-1868) were purchased from the Microbial Culture Collection at the National Institute for Environmental Studies (NIES; 16-2, Onogawa, Tsukuba, Ibaraki 305-8506, Japan). We considered the two strains to belong to identical species (or extremely closely related species) because (1) they originated from the exact same water sample collected in the East China Sea (see Takishita et al. 2008), (2) their cell morphologies were identical under the light microscope (data not shown), and (3) most gene sequences available for both strains were identical (see below). Additionally we examined two original clones of Lepidodinium strains: strain ‘MH 360,’ which was confirmed as Lepidodinium sp. by light microscopy, and strain ‘T03-1,’ which was identified as L. viride based on the presence of the body scales on the cell surface. Strain MH 360 was isolated from sea water collected offshore of the Sanriku region, Japan (Pacific Ocean), on September 9, 2009, while strain T03-1 was isolated from sea water collected offshore southeast of Phuket Island, Thailand (Andaman Sea), on September 18, 2003. The four strains were cultured in MNK medium (Noël et al. 2004) at 20 °C under a 14-h light/10-h dark cycle.

The prasinophycean green alga Micromonas pusilla (NIES-1411) was purchased from the Microbial Culture Collection at NIES and cultured in f/2 medium (Guillard and Ryther 1962) at 20 °C under a 14/10 h LD cycle.

Pigment analysis

Lepidodinium chlorophorum strain NIES-1868 and M. pusilla cells were harvested by centrifugation. After removal of a transparent viscosity layer on the cell pellet, pigments were extracted with 100 μl of 100 % methanol, and the pigment extract was collected into a 1.5 ml centrifuge tube after centrifugation. We repeated the extraction until the extract was no longer colored. The extracted pigments were subjected to reverse-phase high-performance liquid chromatography (HPLC) with a Waters Symmetry C8 column (150 mm × 4.6 mm; particle size 3.5 μm; pore size 100 Å). The HPLC was performed as described in Zapata et al. (2000) without any modifications. The eluted pigments were detected by the absorbance at 440 nm and identified by their elution patterns compared to those reported in Zapata et al. (2000).

Preparation of nucleic acids, PCR, cloning, and sequencing

From the Lepidodinium cells, total DNA samples were extracted using the DNA Isolation Reagent (TaKaRa), and then, served to amplify nucleus-encoded 18S rRNA genes. We amplified 18S rRNA genes from total DNA by a pair of primers: 5′-GAAACTGCGAATGGCTCT-3′ and 5′-CCGCAGGTTCACCTAC-3′. The amplification conditions consisted of 30 cycles of 96 °C for 20 s, 55 °C for 20 s, and 72 °C for 3 min. Amplicons were cloned into pGEM T-Easy vector (Promega), and both strands were completely sequenced. The rRNA gene sequences determined in this study were deposited in DDBJ/EMBL/GenBank under the accession numbers AB686253–AB686256.

Phylogenetic analysis

We retrieved nucleus-encoded 18S rRNA sequences of phylogenetically diverged dinoflagellates, including five sequences deposited as Lepidodinium, as well as closely related alveolates, from NCBI GenBank database. These retrieved sequences were then aligned with the four Lepidodinium sequences determined in this study. The details of Lepidodinium 18S rRNA sequences used in this study are listed in Table 1. The initial alignment was prepared by CLUSTALW ver. 1.83 (Thompson et al. 1994), and then manually refined. After exclusion of all ambiguously aligned positions, the final 18S rRNA alignment included 105 taxa with 1,496 nucleotide positions.

Table 1 Nucleus-encoded 18S rRNA sequences of Lepidodinium species used in this study
Full size table

The rRNA data set was subjected to the maximum-likelihood (ML) analysis under the GTR model incorporating among-site rate variation approximated by a discrete Γ distribution with four equally probable categories (GTR + Γ model) by using RAxML ver. 7.2.1 (Stamatakis 2006). The ML tree was heuristically searched from 10 distinct parsimony trees. A ML bootstrap analysis (100 replicates) was conducted under the GTR + Γ model with heuristic tree search from a single parsimony tree per replicate.

The same alignment was also subjected to Bayesian analysis with the GTR + Γ model by using MrBayes v.3.1.2 (Ronquist and Huelsenbeck 2003). Two parallel Markov chain Monte Carlo runs, each consisting of four incrementally heated chains, were run for 106 generations, sampling log-likelihoods and trees at 100-generation intervals. The first 20,000 generations were discarded as “burn-in,” and the majority consensus tree and Bayesian posterior probabilities were calculated from the remaining trees.

Results

Pigment composition

Micromonas pusilla is one of the standard algae containing prasinoxanthin (Zapata et al. 2000). As anticipated, our HPLC analysis of M. pusilla successfully identified prasinoxanthin, as well as Mg-3,8-divinyl-pheoporphyrin a5 monomethyl ester (MgDVP), uriolide, 9′-cis-neoxanthin, violaxanthin, micromonal, chlorophyll b (Chl b), chlorophyll a (Chl a), β-carotene, and an unknown carotenoid (Fig. 1, top). In contrast, in the same analysis of L. chlorophorum strain NIES-1868, 9′-cis-neoxanthin, violaxanthin, antheraxanthin, zeaxanthin, lutein, Chl b, Chl a, and β-carotene were detected, while prasinoxanthin was not found (Fig. 1, bottom). The overall pigment composition in strain NIES-1868 is of typical green algae, providing no clue for the origin of the L. chlorophorum plastid.

Fig. 1
figure 1

HPLC chromatograms of pigments extracted from Micromonas pusilla NIES-1413 (top) and Lepidodinium chlorophorum NIES-1868 (bottom). Both algae were cultivated under uni-algal conditions. MgDVP, Mg-3,8-divinyl-pheoporphyrin a5 monomethyl ester

Full size image

18S rRNA analysis

We newly determined 18S rRNA sequences from four Lepidodinium strains NIES-1867, NIES-1868, MH 360, and T03-1. The rRNA sequences from our strains determined here and the four sequences deposited in GenBank database—L. viride (AF022199), L. viride (DQ499645), L. chlorophorum (AY331681), and L. chlorophorum (AM184122)—appeared to be identical or almost identical (Table 2). In the 18S rRNA phylogeny, the eight Lepidodinium sequences tightly grouped together, and then, this Lepidodinium clade was placed within the dinoflagellate clade (Fig. 2). Interestingly, the rRNA sequence deposited as ‘L. chlorophorum’ (JF791033) shared lower nucleotide identities with those of other Lepidodinium species (~96 %; Table 2), and it was excluded from the Lepidodinium clade (Fig. 2). Of note, Gymnodinium catenatum, instead of ‘L. chlorophorum’ (JF791033), branched at the base of the Lepidodinium clade, although the relative position of JF791033 to other Lepidodinium sequences was inconclusive (Fig. 2).

Table 2 Pair-wise nucleotide identities among the nucleus-encoded 18S rRNA sequences from eight Lepidodinium species and Gymnodinium catenatum
Full size table
Fig. 2
figure 2

Maximum-likelihood phylogeny based on nucleus-encoded 18S rRNA sequences. Only bootstrap values ≥50 % are presented. The bipartitions supported by ≥0.95 Bayesian posterior probabilities are highlighted by thick lines. The Lepidodinium sequences are shaded. The rRNA sequences determined in this study are highlighted by arrowheads

Full size image

Discussion

We here investigated the pigment composition of L. chlorophorum strain NIES-1868, which is mono-algal culture maintained in the laboratory. Our results presented here indicate that L. chlorophorum strain NIES-1868 lacks prasinoxanthin, and the current Chl-a+b-containing plastid in this strain might be derived from a prasinoxanthin-lacking green alga. These data are consistent with the data presented in Laza-Martinez et al. (2007), suggesting no prasinoxanthin in their L. chlorophorum culture. Importantly, the data on the pigment from strain NIES-1868 can be directly combined with the plastid-encoded gene sequences from the same strain determined in Matsumoto et al. (2011b). Indeed, the absence of prasinoxanthin is apparently consistent with a multigene phylogenetic analysis, which successfully excluded the evolutionary affinity between the plastid in strain NIES-1868 and any of those containing prasinoxanthin (Matsumoto et al. 2011b). Thus, we firmly conclude that (1) the L. chlorophorum strain NIES-1868 lacks prasinoxanthin and (2) the current Chl-a+b-containing plastid in this strain is derived from a prasinoxanthin-lacking green alga, as suggested in Matsumoto et al. (2011b).

The analyses of nucleus-encoded 18S rRNA sequences clearly indicated that the sequence of strain NIES-1868 was very closely related to those of the Lepidodinium cells sampled at various geographical locations. Thus, both pigment composition and phylogenetic data based on plastid-encoded genes from strain NIES-1868 discussed in the previous paragraph are most likely applicable to the cells corresponding to the rRNA sequences, which robustly grouped with the NIES-1868 sequence in the ML analysis. The rRNA sequence deposited as JF791033 in GenBank database, which was excluded from the Lepidodinium clade, may not have originated from a member of the genus Lepidodinium.