Introduction

The Phaeophyceae (brown algae) are multicellular photosynthetic marine organisms and display great morphological and physiological diversity (Yang et al. 2012). There are approximately 2,000 species of brown algae all over the world (Charrier et al. 2012). However, the data on their mitochondrial genomes are limited until now. The first complete mitochondrial DNA (mtDNA) sequence from Pylaiella littoralis (Ectocarpales) was reported in 2001 (Oudot-Le Secq et al. 2001). To date, the mitochondrial genomes of brown algae have been completed in 16 species/subspecies belonging to seven genera (Table 1). The information of complete mitochondrial genomes provides important insights into our understanding of the evolution of the Phaeophyceae and their organellar genomes (Oudot-Le Secq et al. 2006; Yotsukura et al. 2009; Zhang et al. 2013).

Table 1 Completely sequenced mitochondrial genomes from brown algae
Full size table

Mitochondria arose from a single endosymbiotic event of a α-proteobacterial ancestor and prokaryotic nature across the eukaryotic domain (Gray et al. 2001). The well-conserved genetic function of mtDNA involved several biological processes, e.g., electron-transport and oxidative phosphorylation, a distinctive protein-synthesizing system, RNA maturation, and protein import (Lang et al. 1999). More evidence highlights the divergent trends in mtDNA structure and gene expression mechanisms of many eukaryotic lineages (Burger et al. 2003). In the past decades, specific, short mitochondrial fragments were used in investigation of molecular evolution and reconstruction of phylogeny among Phaeophyceae (e.g., Engel et al. 2008; Draisma et al. 2010; Silberfeld et al. 2010). However, short fragments can only be used to a limited extent to resolve phylogenetic relationships at higher taxonomic levels. In recent years, the phylogeny of higher-level taxa have often been reconstructed based on the complete mtDNA sequences, which provide adequate information especially for the reliable resolution of the phylogenetic relationships of closely related species (Yotsukura et al. 2009; Zhang et al. 2013).

Sargassum is a widely distributed genus on rocky intertidal shores worldwide and represents one of the most species-rich genera of the marine macrophytes (Mattio and Payri 2011). Sargassum horneri (Turner) C. Agardh is a dioecious macroalga specially distributed in the Pacific Northwest coasts (Hu et al. 2011). This alga forms forests or meadows in sublittoral regions, serving as nursery habitats and spawning grounds for marine invertebrates and fish (Yatsuya 2008). Because of its ecological and environmental importance, S. horneri has received much attention over the past 30 years thus much of its basic physiological processes involving growth, development, reproduction, and taxonomy (e.g., Choi et al. 2008; Komatsu et al. 2008; Uwai et al. 2009). However, basic understanding of its genome is unknown.

In this investigation, the complete sequence of mitochondria genome was determined for S. horneri. Comparative genomic analyses and its structure and sequence were completed. The new information should provide a better understanding of mitochondrial genome diversity in the Phaeophyceae and insights into molecular evolution of Sargassaceae.

Materials and methods

Sample collection and DNA extraction

Mature plants of Sargassum horneri (Turner) C. Agardh were collected from the rocky shore at Xiaohuyu, Nanji Islands, Wenzhou, Zhejiang Province, China (27°27′N, 121°04′E) in April 2007 (Pang et al. 2009). Plants were transported to the laboratory in coolers (5–8 °C) within 24 h after collection. Algal culture was performed in 80-L polypropylene (PP) tanks under solar irradiance. Fresh algal tissue was selected and stored in the ultra-low temperature freezer (−80 °C). Frozen tissue was used for DNA extraction. Algal tissue was ground to fine powder in liquid nitrogen. Total DNA was extracted using a Plant Genomic DNA Kit (Tiangen Biotech, Beijing, China) according to the manufacturer’s instructions. The concentration and the quality of isolated DNA were assessed by electrophoresis on 1.0 % agarose gel (Liu et al. 2013).

PCR amplification and sequencing

The whole mitochondrial genome of S. horneri was amplified using the long PCR technique (Cheng et al. 1994) and a primer walking method. Primer sets for the long PCR were as follows: mt23S-FB (Draisma et al. 2010)/trnW-trnI-R (Voisin et al. 2005), trnW-trnI-F (Voisin et al. 2005)/CAR4A (Kogame et al. 2005), CAF4A (Kogame et al. 2005)/cox1-1378R (Silberfeld et al. 2010), cox1-117 F (Bittner et al. 2008)/nad5-R (5′-AGCATAACGACAGTTAAACT-3′, this study), cox2-F (5′-TTGGWAAAAGATTATGCTCTTAA-3′, this study)/trnS-R (5′-TTGATTTAGCAAACCAAGGCTT-3′, this study), and trnS-F (5′-AAGCCTTGGTTTGCTAAATCAA-3′, this study)/mt23S-RB (Draisma et al. 2010). Primer sets were used to amplify the entire S. horneri mitochondrial genome in six large fragments. PCR reactions were carried out in 50 μL reaction mixtures containing 32 μL of sterile distilled H2O, 10 μL of 5 × PrimeSTAR GXL buffer (5 mM Mg2+ plus, Takara, Japan), 4 μL of dNTP mixture (2.5 mM each), 1 μL of each primer (10 μM), 1 μL of PrimeSTAR GXL DNA polymerase (1.25 units μL−1, Takara, Japan), and 1 μL of DNA template (approximate 50 ng). PCR amplification was performed on a T-Gradient Thermoblock Thermal Cycler (Whatman Biometra, Goettingen, Germany) with an initial denaturation at 94 °C for 3 min, followed by 30 cycles of denaturation at 94 °C for 20 s, annealing at 50–52 °C for 50 s, extension at 68 °C for 1 min kb−1, and a final extension at 68 °C for 10 min. Long PCR products were purified using a Qiaquick Gel Extraction Kit (Qiagen, Germany). Sequencing reactions were performed using ABI 3730 XL automated sequencers (Applied Biosystems, USA).

Sequence alignment and genome analysis

The DNA sequences were manually edited and assembled using the BioEdit program (http://www.mbio.ncsu.edu/BioEdit/bioedit.html). DNA sequences of the complete mitochondrial genome of S. horneri were determined by comparison with published sequences for several brown algae (Oudot-Le Secq et al. 2001, 2002, 2006; Yotsukura et al. 2009; Zhang et al. 2013). Protein-coding genes were annotated by Open Reading Frame Finder (http://www.ncbi.nlm.nih.gov/gorf/orfig.cgi) and DOGMA (Wyman et al. 2004). Ribosomal RNA genes were identified by BLAST searches (Altschul et al. 1997) of the nonredundant databases at the National Center for Biotechnology Information. Transfer RNA genes were searched for by reconstructing their cloverleaf structures using the tRNAscan-SE 1.21 software with default parameters (http://lowelab.ucsc.edu/tRNAscan-SE/. Schattner et al. 2005). Gene map of mitochondrial genome was generated using OGDRAW (Lohse et al. 2013).

Phylogenetic analysis

Seven brown algal complete mitochondrial sequences were used along with S. horneri in phylogenetic analysis. Dictyota dichotoma (AY500368) was used as outgroup based on Oudot-Le Secq et al. (2006). Phylogenetic relationships within Phaeophyceae were evaluated based on two datasets from mtDNA. One dataset included three ribosomal RNA genes (rnl, rns, and rrn5), and another contained 35 functionally known protein-coding genes as well as their concatenated protein sequences (rps2-4, rps7, rps8, rps10-14, rps19; rpl2, rpl5, rpl6, rpl14, rpl16, rpl31; nad1-7, nad9, nad11; cob; cox1-3; atp6, atp8, atp9; and tatC). Both datasets were subjected to concatenated alignments using ClustalX 1.83 with the default settings (Thompson et al. 1997). Base composition and pairwise comparison were examined by the MEGA 5.2 software (Tamura et al. 2011). Neighbor-joining (NJ) and maximum likelihood (ML) trees were conducted with 1,000 bootstrap replicates using MEGA5.2. The evolutionary distances were computed based on the Kimura two-parameter model for DNA sequences (Kimura 1980) and on the Whelan and Goldman + Freq. model for protein sequences (Whelan and Goldman 2001). Final results are presented using gaps as fifth character states for bases in DNA sequences.

Results and discussion

Genome organization

The complete mtDNA of S. horneri is a circular molecule of 34,606 bp in length (Fig. 1). It is shorter than other available Phaeophyceae mt genomes except for D. dichotoma (31,617 bp. Table 1). The S. horneri mitochondrium is gene dense, with 95.71 % of the sequence specifying genes and open reading frames (ORFs), and only 4.29 % noncoding (Table 2). The percentage of the intergenic spacer (4.29 %) in S. horneri is higher than that in D. dichotoma (3.21 %) but lower than those in other brown algae (5.62–6.81 %), which are closely related to the mt genome size and gene content. The overall AT content of S. horneri mtDNA is 63.84 %, similar to 62.01 % in P. littoralis to 66.49 % in Ectocarpus siliculosus. Intergenic spacers of S. horneri mtDNA are significantly rich in AT content (74.28 %). This pattern is similar to other brown algal mtDNA with the exception of P. littoralis, in which the AT content of protein-encoding region was 70.00 %, higher than that of intergenic spacers (67.41 %).

Fig. 1
figure 1

Gene map of Sargassum horneri mitochondrial DNA. Genes (filled boxes) shown on the outside of the map are transcribed in a clockwise direction, whereas those on the inside of the map are transcribed counterclockwise

Full size image
Table 2 General features of mitochondrial genomes in eight species of brown algae
Full size table

The S. horneri mt genome contains three ribosomal RNA genes (rRNA), 25 transfer RNA genes (tRNA), 35 protein-coding genes identified by sequence homology, as well as two ORFs (Table 3). Its total gene number is 65 and lower than other brown algae (66−79 genes). Genes in S. horneri mtDNA are encoded on both strands and are not interrupted by intron. The heavy strand (H) and the light strand (L) encode 59 and 6 genes, respectively.

Table 3 Genes identified in mitochondrial genomes of eight species of brown algae
Full size table

In S. horneri mtDNA, intergenic spacers average only 28.0 nucleotides with a range of 0 to 172 nucleotides, which is significantly shorter than those of other Phaeophyceae mtDNA except for D. dichotoma (18.8 nucleotides with a range of 0 to 74 nucleotides). Twelve pairs of genes overlap by 1 to 66 nucleotides, with the average size of 13.7 nucleotides, which are in the same range as those of other brown algae. Two conserved overlapping regions, ATGA overlapped by rps8 and rpl6, and A overlapped by rpl6 and rps2 reported by Oudot-Le Secq et al. (2006), are found in S. horneri mtDNA.

All protein-coding genes encoded by S. horneri mtDNA have a methionine (ATG) start codon. Other start codons have been found in other brown algal mtDNA, e.g., a GTG codon found at the beginning of rps14 genes in D. dichotoma, ORF379 in Fucus vesiculosus, and ORF157 gene in Laminaria digitata; a TTG codon at the beginning of ORF211 gene in Desmarestia viridis, ORF37 gene in D. dichotoma, and nad11 gene in P. littoralis, a CTG codon at the beginning of nad11 gene in E. siliculosus. Three stop codons are used, with a preference of 64.86 % for TAA (16.22 % for TAG and 18.92 % for TGA). This feature was shared in all of reported brown algal mtDNA, with proportional changing to a certain extent.

The mitochondrial genome of S. horneri shows remarkable similarity to other brown algae in terms of gene content, gene organization, AT composition, and gene sequences. In terms of all this attributes, S. horneri mtDNA is more similar to F. vesiculosus mtDNA as expected. This suggests that S. horneri has a close evolutionary relationship with F. vesiculosus as compared to other brown algae supporting current taxonomic systems. This was confirmed by recent morphological, phylogenic, and classification data (Guiry and Guiry 2014).

The mtDNA genomes from different brown algae display a highly conservative architecture (Oudot-Le Secq et al. 2006; Yotsukura et al. 2009; Zhang et al. 2013), but differ in terms of different genome size. Size of mtDNA genomes are reduced by several mechanisms including but not exclusive of gene transfer to the nucleus, intron loss, elimination of intergenic spacers, and overlap of adjacent coding genes (Burger et al. 2003). Comparisons among eight brown algae have indicated that the reduction of mtDNA size in S. horneri are closely associated with the above aspects, e.g., the loss of one putative tRNATyr gene and ORF379 found in F. vesiculosus, no intron found in genes, the average spacer size of 28.0 bp, and the average overlap size of 13.7 bp.

Ribosomal and transfer RNA genes

The predicted sizes of rnl and rns genes in S. horneri are 2,667 and 1,536 bp, respectively, consistent with currently reported Phaeophyceae rRNAs (Oudot-Le Secq et al. 2006; Yotsukura et al. 2009). Both can fold in accordance with conserved, eubacteria-like model. The rrn5 gene of the S. horneri mtDNA is separated from the rns by six nucleotides, TTAAGA and overlaps the tRNAMet-3 gene with three nucleotides, TGC.

The S. horneri mt genome contains 25 tRNA genes ranging from 72 to 88 nucleotides in size (Table 4). All of the tRNA genes could fold into a typical cloverleaf secondary structure as determined by tRNAscan-SE. The anticodon usage is identical to that found in other brown algae. Among them, six tRNA genes (Leu-1, Leu-2, Leu-3, Tyr, Ser-1, and Ser-2) contain the extra arms. The tRNAMet-2 gene located between tRNAGln and ORF41genes was labeled as tRNAIle in F. vesiculosus, L. digitata, Saccharina japonica, D. dichotoma, D. viridis, and P. littoralis, but also tRNAMet-2 in E. siliculosus. The counterparts of the tRNAIle-2 gene were different in other brown algae, e.g., X (uncertain amino acid) in F. vesiculosus, L. digitata and S. japonica, Lys in D. dichotoma and D. viridis, Ser in E. siliculosus, and Arg in P. littloralis. However, one putative tRNA gene (the tRNATyr gene) shared by F. vesiculosus and D. viridis mt genomes were not detected in the S. horneri mt genome.

Table 4 The aligned sequences of the 25 tRNA genes detected in the Sargassum horneri mtDNA
Full size table

Protein-encoding genes and ORFs

The identified mitochondrial protein-encoding gene set is the same as those in currently reported Phaeophyceae mt genomes (Oudot-Le Secq et al. 2006). These 35 genes encoded 17 ribosomal proteins (rps2-4, 7, 8, 10–14, 19; rpl2, 5, 6, 14, 16, 31), 10 subunits of the NADH dehydrogenase (nad1-7, 4 L, 9 and 11), apocytochrome b (cob), 3 first subunits of the cytochrome oxidase (cox1-3), 3 subunits of the ATPase (atp6, 8, and 9), and a protein transporter component of the secY-independent pathway (tatC). No notable reduction or extension of gene length was observed compared to other Phaeophyceae except the cox2 gene. The cox2 gene in S. horneri contains an in-frame insertion of 2,277 bp. Such insertion also exists in the cox2 genes of F. vesiculosus (2,283 bp), E. siliculosus (2,946 bp), P. littoralis (2,973 bp), L. digitata (2,979 bp), S. japonica (2,982 bp), and D. viridis (2,994 bp), but is absent in the D. dichotoma cox2 gene. The insertion with a phage-type RNA polymerase gene found in P. littoralis mtDNA is not detected in the S. horneri mt genome.

Two conserved ORFs (ORF129 and ORF41) are found in S. horneri mtDNA, whereas another conserved ORF, ORF379 in F. vesiculosus, is missing from S. horneri mtDNA. All currently reported Phaeophyceae mt genomes contain only one specific ORF, ORF129 (S. horneri), ORF131 (F. vesiculosus), ORF129 (L. digitata), ORF130 (S. japonica), ORF111 (D. dichotoma), ORF143 (D. viridis), ORF128 (E. siliculosus), and ORF127 (P. littoralis). The homologous genes of ORF41 were lost in P. littoralis, but are present in other brown algae, ORF43 (F. vesiculosus), ORF42 (E. siliculosus), ORF40 (L. digitata), ORF41 (S. japonica), ORF39 (D. viridis), and ORF37 (D. dichotoma). No unique ORF was found in the S. horneri mtDNA.

Phylogenetic relationships within Phaeophyceae

In order to carry out a basic phylogenetic analysis, 3 ribosomal RNA genes and 35 protein-coding genes and the derived amino acid sequences identified in S. horneri were aligned with other brown algal species. Both maximum likelihood and neighbor-joining analyses were run, with the analyses generating slightly different topologies between two datasets and high bootstrap support values. Phylogenetic trees showed that eight species of brown algae well fell into five clades standing for five different orders: Laminariales (two species), Desmarestiales (one), Ectocarpales (two), Fucales (two), and Dictyotales (one). Sargassum horneri was tightly combined with F. vesiculosus, forming the Fucales clade in all the phylogenetic trees with high bootstrap support values (99–100 %) (Figs. 2 and 3).

Fig. 2
figure 2

Phylogenetic tree of brown algal relationships derived from three rRNA genes (rnl, rns, and rrn5) in the mt genome. Dictyota dichotoma served as outgroups. Numbers in each branch indicated maximum likelihood (ML, above) and neighbor-joining (NJ, below) bootstrap values

Full size image
Fig. 3
figure 3

Phylogenetic tree of brown algal relationships derived from the nucleotide (a) and amino acid (b) datasets of 35 functionally known protein-coding genes (rps2-4, rps7, rps8, rps10-14, rps19; rpl2, rpl5, rpl6, rpl14, rpl16, rpl31; nad1-7, nad9, nad11; cob; cox1-3; atp6, atp8, atp9; and tatC) in the mt genome. Dictyota dichotoma served as outgroups. Numbers in each branch indicated maximum likelihood (ML, above) and Neighbor-joining (NJ, below) bootstrap values

Full size image

The topology of the trees obtained in this study was similar to that from the previous results based on few mitochondrial, plastid, and nuclear markers (e.g., Silberfeld et al. 2010; Charrier et al. 2012). However, based on the rRNA dataset, Laminariales firstly combined with Desmarestiales with bootstrap values of 95 % and then Ectocarpales with bootstrap values of 100 %, while, in the latter dataset, Laminariales firstly combined with Ectocarpales with low bootstrap values (ML/NJ = 60/51 %) based on DNA sequences but high bootstrap values (ML/NJ = 89/91 %) based on protein sequences. These different topologies revealed that the extent of sequence variation or evolution was different between rRNA region and the total protein-coding gene region.

The completely sequenced and annotated mtDNA of S. horneri represents an important source of information for understanding not only evolution of the mitochondrial genome in the Phaeophyceae but also provides insights into the evolutionary history of the class. Mitochondrial DNA markers designed for phylogenetics were scarce due to the limited data but very useful to understand phaeophycean phylogenies. Lane et al. (2007) and Uwai et al. (2007) used cox1 and cox3 to define species limits and genetic relationships in the orders Laminariales, respectively. The mtDNA 23S sequence (rnl) and the variable mt23S-tRNAVal intergenic spacer were employed to infer interspecific and intergeneric relationships and delineate genera within the order Fucales (Coyer et al. 2006; Draisma et al. 2010). Based on ten mitochondrial (cox1, cox3, nad1, nad4, and atp9), plastid, and nuclear loci, Silberfeld et al. (2010) hypothesized that the brown algal crown radiation (BACR) likely represents a gradual diversification during the Lower Cretaceous instead of a sudden radiation as proposed earlier. Based on the complete mtDNA sequences of S. horneri, new mtDNA markers could be selected and employed to study the diversity of S. horneri populations and Sargassaceae phylogenetics in the future.