1 Introduction

From 2007 to 2017, large-scale green tides occurred every year in the Yellow Sea of China. The phenomenon was caused by eutrophication, which occurred due to excessive dumping of fertilizers combined with increased temperatures (Valiela et al., 1997; Raffaelli et al., 1998). These green tides had negative impact on the local economy and environment. Worsening the impact, winds and currents can drive the tides into other coastal areas (Liu et al., 2009). To the best of our knowledge, the world's largest green tide caused mainly by Ulva prolifera O. F. Müller (Chlorophyta, Ulvophyceae) occurred in 2008 in Qingdao on the Yellow Sea coast in northeastern China. There are different assumptions about the origin of this green tide. Based on satellite remote sensing images, the origin of this green tide was postulated as being from the southern Yellow Sea and closely related to the expansion of Pyropia aquaculture regions along the coastline of Jiangsu Province (Liu et al., 2009; Hu et al., 2010; Keesing et al., 2011). Using a combined method of morphology and an analysis of internal transcribed spacer (ITS) and rbcL sequences, Pang et al. (2010) found that the Ulva blooms originated from land-based aquaculture ponds; however, the result was not confirmed when using ISSR molecular markers (Liu et al., 2011). Because of similar morphologies in Ulva species and close relatives, it is necessary to find a suitable molecular marker to distinguish the species and determine the origin of the major species of green tides.

In eukaryotes, nuclear ribosomal DNA is a cluster structure that contains multiple tandem transcription units, and each rDNA transcriptional unit consists of the three rRNA coding sequences (18S rDNA gene, 5.8S rDNA gene and 28S rDNA gene), two internal transcribed spacers (ITSs) and one intergenic spacer (IGS; Volkov et al., 2004; Poczai and Hyvonen, 2010; Shaw, 2013). Due to different evolutionary rates, different regions of ribosomal transcription units can be regarded as molecular markers for identifications and phylogenetic studies of species at different levels. In marine algae research, Hoham et al. (2002) combined 18S rDNA and rbcL to analyze the phylogeny of Chloromonas and Chlamydomonas (Chlorophyceae, Volvocales), emphasizing snow and other cold-temperature habitats. Riethmuller et al. (1999) demonstrated that Phytophthora de Bary and the Peronosporales were a common natural group by phylogenetic studies based on nuclear large subunit ribosomal DNA sequences. Luo et al. (2012) found that the bloom-forming algae in the Yellow Sea in 2009 and 2010 were the same species through combining ITS including 5.8S rDNA sequence and morphological data. The rRNA sequence such as ITS region, 18S rDNA, 28S rDNA, 5S and partial rbcL gene sequences were applied to demonstrate sequence variation within a species; in addition, the complete chloroplast genome of U. flexuosa and U. linza and the mitochondrial genomes of U. pertusa have been published with the popularization of high-throughput sequencing (Cai et al., 2017; Wang et al., 2017; Liu and Bi, 2017). However, the IGS sequences between the LSU and SSU have not attracted enough attention.

The aim of this study was to amplify the IGS sequences of U. prolifera and apply them to study the connection of the green tides in the Qingdao coastal area, the Jiangsu coastal area and the Yellow Sea.

2 Materials and Methods 2.1 Sample Collection

All algal samples used in this experiment were stored in the Algae Laboratory of the School of Basic Medical and Biological Sciences, Soochow University, and information on the samples used in this study is presented in Table 1. Species identification was based on morphological characteristics and ITS sequences. Once the samples were collected, they were washed with seawater repeatedly, and the samples were dried in the shade to a moisture content of 30%–40%. The samples were transported to the laboratory in an insulated specimen box at 4℃.

2.2 Morphological Observation

When the fresh algae were obtained, the main axis, diverge branches, colors and basal part (holdfasts) were recorded and photographed. The samples were sectioned into thin slices with a knife for observation under a microscope. Then, the cell shapes, sizes and arrangements were observed horizontally and vertically; the cell interior structures, such as chromatophores and pyrenoids, were also measured. The microscopy (Nikon N90i) and the stereomicroscopy (NiKon SMZ 1500) with an image collection system were applied in this research to obtain the main characteristics of the samples.

2.3 Total RNA Extraction

Fresh algae samples of Blidingia sp. were ground into powder using an RNase free mortar and pestle with liquid nitrogen, followed by adding 1 mL of total RNA extractor (Sangon Biotech, Shanghai) and covering the powder completely according to the manufacturer's instructions. A 1% agarose gel electrophoresis was used to measure the total RNA quality, and then, the samples were aliquoted and stored at -80℃ until used.

2.4 DNA Extraction

The algae were taken out from the refrigerator and placed in sterile double distilled water for several hours while the surface of the algae was constantly brushed to remove debris from the surface of the thallus, soaked in 0.7% KI for 10 min, and then rinsed several times with sterile seawater. Total genomic DNA was extracted by the Plant Genomic DNA Kit (TIANGEN. BIOTECH (BEIJING) Co., Ltd.), following the procedure in the instruction manual.

2.5 PCR Amplification and Sequencing

The primers (Table 2) used in this study were designed with primer 5.0 software. PCR amplification was carried on in a 50-μL volume containing 32.7 μL ddH₂O, 5 μL dNTP Mix (2.5 mmol L^-1), 5 μL of the template DNA, 5 μL 10× LA Taq buffer (Mg²⁺), 1μL of each primer (20 n mol L^-1), and 0.3 μL LA Taq DNA polymerase (Takara Biotechnology (Dalian) Co., Ltd.). The colony PCR method was performed on a final volume of 20-μL mixtures containing 13.8 μL ddH₂O, 2 μL dNTP Mix (2.5 m mol L^-1), 2 μL 10× Ex Taq Buffer (Mg²⁺), 1 μL LB medium containing positive colony, 0.5 μL of each primer (20 nmol L^-1), and 0.2 μL Ex Taq DNA polymerase (Takara Biotechnology (Dalian) Co., Ltd.).

The PCR cycle employed for the amplification of ITS genes was referenced in Lin et al. (2013), and the other cycles are presented in Table 3. The colony PCR method was as follows: 3 min initial denaturation at 95℃ followed by 30 cycles of denaturation at 95℃ for 30 s, primer annealing at 55℃ for 30 s, and extension at 72℃ for 1–5 min, with the final extension step at 72℃ for 5 min. A 1% agarose gel electrophoresis was used to analyze PCR products.

PCR products were purified using the TaKaRa Agarose Gel DNA Purification Kit Ver.4.0 (TaKaRa Biotechnology (Dalian) Co., Ltd.) and the TA-cloned into pEASY-T3 vector and then transformed into Trans1-T1 phage resistant chemically competent cells (Beijing TransGen Biotech. Co., Ltd.). The transformed cells were spread onto an LB agar plate containing X-gal, Amp and IPTG; then, the plates were cultivated at 37℃ for at least 12 h. Positive colonies were selected and cultivated, and the transformed colonies were confirmed by colony PCR. Three positive recombinant colonies of each amplification product were sequenced using the method of Sanger dideoxy sequencing by GENEWIZ biotechnology Co., Ltd., Suzhou, China.

2.6 Sequence Alignments and Analysis

The nucleotide BLAST was used for sequence alignment with the database to confirm the amplification results. Tandem repeats within IGS were identified using the tandem repeat finder (Benson, 1999). The sequence alignments were constructed using Clustal W (Higgins et al., 1994), and phylogenetic analyses used the neighbor joining method with 1000 replications in MEGA 5.0.

3 Results and Discussion 3.1 Morphological Observation

The samples of U. prolifera in this research were divided into 2 categories, based on distinct morphology differences: Type 1. Frond with main branch clearly distinguished with many branches was found frequently in the Yellow Sea and Qingdao coastal waters (Fig. 1). Type 2. Algae with enlarged vesicular twist with air chambers and denseness branch, and with irregular cell shapes in the section, grown in the way of floating, mainly found in the Jiangsu coastal waters (Fig. 2).

3.2 Amplifications of 28S rDNA and IGS Sequences

The total genomic DNA or cDNA were used to amplify 28S rDNA and IGS sequences with corresponding primers. The amplified bands of 28S rDNA in U. prolifera are shown in Fig. 3A using the primers (28F1/R1). Since it was difficult to amplify the full length of 28S rDNA in Blidingia sp. with a pair of primers, the primer (28F2/R2) was used to amply the partial 28S rDNA in Blidingia sp., the primer (28F3/R3) was used to amply the other part of 28S rDNA in Blidingia sp. and the amplified bands are shown in Figs. 3B and 3C respectively. After the analysis, the lengths of 28S rDNA in U. prolifera and Blidingia sp. were 3259 bp and 3282 bp, respectively. The IGS sequence of U. prolifera was longer than that of Blidingia sp. (Fig. 3D), and the primers for amplifying IGS were designed based on the 18S rDNA and 28S rDNA sequences; therefore, the sequencing of the IGS sequence contained part of the 18S rDNA and 28S rDNA sequences. After removing the 18S rDNA and 28S rDNA partial sequences, the complete lengths of IGS sequences in U. prolifera and Blidingia sp. were 3388 bp and 3059 bp, respectively. Each regional sequence of the ribosomal was submitted to GenBank, and the accession numbers are presented in Table 4.

Using the primers (F1/R1), the corresponding IGS sequences were amplified from the genomic DNA of samples used in this study; the 1% agarose gel electrophoresis is presented in Fig. 3E. Removing the 5' end with 28S rDNA overlap and the 3' end overlap with 18S rDNA, the length of the IGS sequence was changed from 2473 bp to 3388 bp. The partial sequences of IGS in U. prolifera were amplified by IGSF1/IGSR1, and the 1% agarose gel electrophoresis is presented in Fig. 3F.

3.3 Structural Analysis of the IGS Sequence in U. prolifera and Blidingia sp.

The IGS sequence was analyzed by the tandem repeat finder, and a two tandem repeat sequences were found, and there were also a short dyad and palindromic sequence both in U. prolifera and Blidingia sp. In addition, the location of these special structures in the IGS sequence of the two species were not the same (Fig. 4), and the G + C contents of the IGS sequence in U. prolifera and Blidingia sp. were 52.42% and 53.09%, respectively. The detailed characteristics of these special structures are presented in Table 5.

3.4 Phylogenetic Analyses

A total of 28 Ulva and 3 Blidingia individuals collected from the Hainan coastal area, Qingdao coastal area and Jiangsu coastal area from 2016 to 2017 were analyzed using IGS sequences. With the IGS sequence analyses, these samples were grouped into three major clades (Fig. 5). From the NJ tree constructed by the IGS sequences, the three samples of U. intestinalis formed one clade with 99% bootstrap values. The individuals of U. flexuosa and U. compressa were grouped into a major branch with 98% bootstrap values, but the two samples of U. compressa were in an independent small branch with a bootstrap support of 87%. The 19 individuals of U. prolifera and 3 individuals of Blidingia sp. formed a major clade. The individuals of D1, D3, M, S, QD and JS1 sites in the Qingdao coastal area; the samples of SE2, SE3, SE4 and SE5 sites in the Jiangsu coastal area; and the samples of E2, E6, F2, F3, F5, F6, F7, F8 and F10 sites in the Yellow Sea were in an independent small branch with 99% bootstrap values.

4 Discussion

According to the morphological characteristics of the collected algae of U. prolifera, we divided the samples into two types. In type 1, the main branch was clearly distinguished with many branches; the cell surface view was without distinct longitudinal rows and 1 pyrenoid; two-layer cells had square or sub-circular shapes on the view of cross section. In type 2, algae had an enlarged vesicular twist with air chambers and dense branches, without distinct longitudinal rows and 1 pyrenoid cells in the surface view with irregular cell shapes in the cross section. Zhang et al. (2013) also found four morphological forms of U. prolifera, i.e., filamentous, tubular, cystic and folded blades, in the Yellow Sea during green tides. However, the analysis of ITS sequences of these samples showed that there were no significant differences among them. Therefore, why did they present different morphological features? U. prolifera had more but shorter branches when they were cultured under lower temperature and salinity conditions, as higher temperatures did not enhance the net photosynthetic rate, while lower salinity conditions did enhance the rate (Gao et al., 2016). The close relationship between morphology when grown under axenic conditions has been demonstrated (Matsuo et al., 2005; Wichard, 2015). Therefore, the same species of U. prolifera have various morphologies in different environments, which means that U. prolifera has the characteristic of morphological plasticity.

This study is the first time that nrDNA-LSU sequences and intergenic spacer (IGS) sequences between the LSU and SSU in U. prolifera have been amplified. In the Blidingia sp., the length of the LSU is 3282 bp, and the IGS is 3059 bp; the LSU is 3259 bp with an IGS of 3388 bp in U. prolifera. We know that the genes in the non-coding region are prone to mutation and deletion. Considering the length of the IGS sequence of five green algae in this study, the results showed a large difference, and the length of the IGS sequence varies from 2473 bp to 3388 bp. Moreover, the IGS sequence may bear a functional sequence, such as promoter, enhancer, transcription stop signals and reproduction start signals (Bhatia et al., 1996). In this study, tandem repeat sequences, short dyads and a palindromic sequence were found in the IGS sequences. Burton et al. (2005) has demonstrated that the tandem repeat sequences function as transcriptional enhancers. The functions of short dyads and the palindromic sequence are currently unclear, but we can speculate that they may have the important functions because the IGS sequences of these algae samples in current study all have these same structures, yet their positions are inconsistent.

In each ribosome transcription unit, non-coding sequences (ITS and IGS) have a higher evolutionary rate than the coding sequences of 18S rDNA, 5.8S rDNA and 28S rDNA, and the IGS region is more variable than the ITS region and is often used for phylogenetic studies in interspecies or intraspecies (Schmidt et al., 2008; Dai et al., 2008; Li et al., 2010; Bertoldo et al., 2011; Onkendi and Moleleki, 2013). There are some differences between the IGS marker and the other published markers (e.g., ITS, 5S etc), the IGS marker is a more sensitive molecular marker than these markers, but it is difficult to amply the IGS marker due to its complex structure. From the NJ tree constructed by partial IGS sequences, the three samples of U. intestinalis formed one clade with 99% bootstrap values. The individuals of U. flexuosa and U. compressa were grouped into a major branch with 98% bootstrap values, but the two samples of U. compressa were in an independent small branch with a bootstrap value of 99%. The 19 individuals of U. prolifera and 3 individuals Blidingia sp. formed a major clade, and the individuals of U. prolifera from the Qingdao coastal area and Jiangsu coastal area were in an independent small branch with 99% bootstrap values. It is strange that U. prolifera belonging to the Ulva genera and Blidingia sp. belonging to the Blidingia genera were clustered into the same branch, as this result is inconsistent when using the other molecular tags; therefore, further study is needed to validate this result. The individuals of U. prolifera from the Qingdao coastal area, Jiangsu coastal area and the Yellow Sea were clustered into one clade with 99% bootstrap values. In addition, the IGS sequence is almost identical. Thus, the U. prolifera that caused the green tide in the Qingdao coastal area, Jiangsu coastal area and the Yellow Sea are the same species and have the same origin.

In this study, the nrDNA-LSU sequences and intergenic spacer (IGS) sequences were amplified for the first time, which enriches the information on the nuclear ribosomal DNA cluster sequences of U. prolifera. Furthermore, the classification of green algae was carried out by IGS sequence for the first time.

Acknowledgements

This work was supported by the National Key R & D Program of China (Nos. 2016YFC1402102 and 2016YFC 1402104) and the National Natural Science Foundation of China (No. 41276134). Thank Captain Lin Wei in the collection of experimental samples.