Introduction

Saccharina japonica, which is indigenous to Japan but was accidentally introduced to China in the early twentieth Century, has been extensively cultivated in China due to its high biomass productivity and its economic value in the food, fertilizer, chemical, drug, and phycocolloid industries (Tseng 2001). Various classical breeding methods for terrestrial crops, including continuous selection, hybridization, and the development of interspecies hybrids, have been used as a part of its cultivation since the 1950s for the genetic improvement and development of a set of elite varieties and hybrids. Representatives of these varieties include Haiqin no.1, 90-1, and interspecies hybrids Dongfang no. 2 and Dongfang no. 6, which have all contributed significantly to its aquaculture performance, particularly its yield (Fang et al. 1962; Li et al. 2007, 2015; Zhang et al. 2007). Despite these achievements, the genetic study of S. japonica is far behind that of most model plants.

Microsatellites, also known as simple sequence repeats (SSRs), are hypervariable, co-dominant, and abundant in all organisms; therefore, they are frequently the target of genetic research. SSRs are divided into genomic SSRs (random genomic sequences) versus genic SSRs (coding sequences/expressed sequence tags (ESTs)) according to location in the genome. SSRs in coding sequences are more useful for genetic study compared with those in non-coding sequences, because they are potentially linked with particular regions that may contribute to agronomic phenotypes and have high transferability. Various genomic SSR markers have been developed for S. japonica and applied to genetic studies in previous studies, including diversity assessment (Shi et al. 2007; Liu et al. 2012; Zhang et al. 2015), linkage map construction, and quantitative trait loci mapping (Yang et al. 2009; Liu et al. 2010a). However, no genic SSR markers are available in this species except the EST-SSR markers from Liu et al. (2010b). Moreover, the relationship between these EST-SSR markers and other genes and markers is unclear due to limited flanking sequences and insufficient genome information. Thus, the EST-SSR markers’ applicability is no different than genomic SSR markers. We believe that it is necessary to develop genic SSRs with sufficiently informative flanking sequences to increase application depth and width.

The development of next-generation sequencing technologies, such as the Roche/454 and Illumina Solexa platforms, has enabled the highly efficient identification of large numbers of SSRs, including genic SSRs, for an increasing number of species (Mikheyev et al. 2010; Dutta et al. 2011; Frenkel et al. 2012; Zheng et al. 2013). The Illumina platform operates with a much higher throughput and lower costs than the 454 sequencing platform. Although the Illumina platform generates shorter reads, those reads can be assembled into a draft genome and then used to discover SSRs with sufficient flanking sequences. Thus, the use of Illumina sequencing data to identify SSRs has evolved into a highly successful approach (Cai et al. 2013; Silva et al. 2013; Shi et al. 2014). Genomic SSRs have previously been identified based on a 500 bp insert Illumina paired-end shotgun survey for S. japonica (Zhang et al. 2014). In this study, we report the first identification of genic SSRs from the assembled and annotated S. japonica genome, from which we developed and characterized novel polymorphic trinucleotide genic SSR markers, providing a useful tool for studying breeding programs and germplasm conservation.

Materials and methods

Genomic DNA was extracted from inbred lineage using a previously described protocol (Cock et al. 2010). Three paired-end shotgun libraries (180, 500, and 800 bp) were constructed following the standard Illumina TrueSeqDNA sample preparation kit protocol, and then sequenced on an Illumina HiSeq 2000 (USA) following standard protocol. The raw reads were cleaned up by trimming the adapter and removing low-quality bases. The genome was assembled using SOAPdenovo (Li et al. 2010). Next, Augustus software was used to predict genes to divide the assembled genome into non-coding versus coding sequences (Stanke et al. 2008). Genic SSRs were then identified using the MISA perl script (Thiel et al. 2003), with the default minimum number of repeats set to 12, 6, 5, 5, 5, and 5 for the mono-through hexanucleotide simple sequence repeats, respectively. Primers were designed from the trinucleotide SSRs (repeat number more than 10) with sufficient flanking regions using the online software Primer3 (Koressaar and Remm 2007; Untergrasser et al. 2012), and synthesizedby Sangon Biotech (China).

Fifty-seven trinucleotide repeat microsatellite primer pairs were synthesized and evaluated by DNA amplification and agarose gel electrophoresis. Primer pairs that successfully amplified fragments were picked for subsequent polymorphism analysis. The polymorphisms of these loci were tested in 55 Saccharina japonica individuals, which were collected from an aquafarm located in Chudao Island in Shandong Peninsula, China (37°01′N, 122°33′E) in March, 2015. A 2.0-cm-diameter blade tissue disc was scrubbed with absorbent cotton, washed with sterile seawater, and frozen at −70 °C for DNA extraction. Genomic DNA was extracted using a plant genomic DNA kit (Tiangen Biotech, China) according to the manufacturer’s protocol. Polymerase chain reactions were performed in a total volume of 25 μL containing 1× buffer, 1.5 mM MgCl2, 0.2 mM dNTPs (each), 0.2 μM primers (each direction), 1U Taq polymerase (Tiangen Biotech), and about 100 ng DNA as the template. Thermocycling conditions are as follows: an initial step at 94 °C for 3 min; followed by 32 cycles of denaturation at 94 °C for 45 s, annealing at the appropriate temperature for every pair of primers (Table 1) for 45 s, and extension at 72 °C for 1 min; and a final extension at 72 °C for 10 min. PCR products were separated on a 6 % denaturing polyacrylamide gel, and visualized by silver staining. To assess genetic diversity, the number of alleles and the observed/expected heterozygosity were calculated using POPGENE 32 (Yeh et al. 2000). Linkage disequilibriumand Hardy–Weinberg equilibrium were tested using Arlequin version 3.5 (Excoffier and Lischer 2010).

Table 1 Characteristics of 26 polymorphic trinucleotide genic SSRs loci for Saccharina japonica
Full size table

Results and discussion

In total, 11,973 SSRs were identified in 58.3 Mb of assembled genomic coding sequence, a frequency of 205.5 genic SSRs per Mb of coding sequence. Trinucleotide SSRs were dominant (>90 %) among the nearly 12 thousand genic SSRs, followed in abundance by the mono-, hexa-, and dinucleotide SSRs; whereas tetra- and pentanucleotide SSRs were not commonly identified (<2 %) (Fig. 1). Therefore, we used trinucleotide SSRs to develop primer pairs for potential application as SSR markers. Microsatellite abundance generally decreased significantly with increasing motif repeat numbers (Fig. 1). Schlttöerer demonstrated that the mutation rate of SSRs increases with repeat number in previous mutational dynamics studies (2000). Although no study exists showing a robust relationship between microsatellite marker polymorphism level and the motif length, we randomly selected trinucleotide genic SSRs with repeat numbers of more than 10 to design primers.

Fig. 1
figure 1

Relative frequency (%) of SSR type in the Saccharina japonica assembly genome by number of repeats. The graph is based on a total of N = 11,973 genic SSRs detected in 58 Mbp genomic coding sequence

Full size image

Among the 57 primer pairs we designed, 32 easily amplified DNA, producing discrete, clear agarose gel electrophoresis bands (Fig. 2). Of these, 26 loci displayed polymorphism among the 55 S. japonica individuals tested (Fig. 3). The number of alleles per locus ranged from 2 to 7, with an average of 3.46, which is lower than the 5.7 average number of alleles per locus Liu et al. (2010b) reported for S. japonica EST-SSR loci. Saccharina japonica breeding strategies have included selecting only desirable individuals with excellent economic traits as seedlings, whereas others are discarded. Therefore, we propose that the lower allele count average that we found in this farmed population may be caused by the continuous artificial selection of desirable economic traits during the domestication process, further increasing allelic losses, reducing genetic diversity, and losing genetic resources (Liu et al. 2012). Alternatively, continuous inbreeding may also play a similar role after a few generations of repeated use (Shan et al. 2011). Either way, this highlights the need for protecting, managing, and utilizing the limited natural S. japonica resources still available during its continued cultivation. Additionally, the lower polymorphism level seen in this study may be due to the influence of different motif repeat lengths or repeat types. In Brassica, the polymorphism level of microsatellite markers for penta- and hexanucleotide repeats was significantly higher than that for mono- to tetranucleotide repeats (Shi et al. 2014). However, no study has yet shown a strong relationship between microsatellite marker polymorphism level and motif length. The 26 polymorphic loci we found had observed and expected heterozygosities varying between 0.128–0.652 and 0.130–0.676, respectively (Table 1). After adjustment by Bonferroni correction, significant deviation from the Hardy–Weinberg equilibrium was detected at two of our SSR loci (ZpsjE-20 and ZpsjE-53), probably also the result of the samples being from a cultivated population. There was no evidence of significant linkage disequilibrium except for markers ZpsjE-1 and ZpsjE-16, which were strongly linked (P < 0.0002).

Fig. 2
figure 2

Agarose gel electrophoresis of PCR products by 10 novel Saccharina japonica trinucleotide genic SSR markers. Lanes1–10:ZspjE37, 38, 40, 41, 44, 45, 49, 53, 54, 57; each primer pair was used to amplify DNA from two sporophyte individuals

Full size image
Fig. 3
figure 3

SSR pattern of 55 Saccharina japonica individuals amplified by SSR primer ZpsjE-10

Full size image

A preliminarily assembled and annotated S. japonica genome enabled us to identify a large numbers of genic SSRs with sufficiently informative flanking sequences in a highly efficient manner. The novel genic SSR markers we developed in this study are sufficiently variable for population diversity analyses. These markers will be valuable tools for the conservation and management of S. japonica germplasm resources, and for further genetic studies, including parentage analysis, linkage mapping, and marker-assisted breeding. Furthermore, the S. japonica draft genome was reported recently (Ye et al. 2015). This will enable us to map our genic SSRs on the genome and determine positional relationships between the genic SSR markers and functional genes. The existence of genic SSR markers linked to quantitative traits that can be easily scanned and identified will greatly facilitate genetic breeding research in Saccharina.