Status of genetic studies and breeding of Saccharina japonica in China*

WANG Xiuliang , YAO Jianting , ZHANG Jie , DUAN Delin ,

1 Key Lab of Experimental Marine Biology, Institute of Oceanology, Center for Ocean Mega-Science, Chinese Academy of Sciences, Qingdao 266071, China

2 Lab for Marine Biology and Biotechnology, Qingdao National Laboratory for Marine Science and Technology, Qingdao 266071, China

Abstract Saccharina japonica is one of the most important economic brown seaweeds. It is intensively cultivated on large scales in a number of Asian countries. The current annual, global production is about 8 million tons valued as about 4 million US dollars. Considerable efforts have been made to S. japonica in China since the 1950s on its cultivation. To further advance the cultivation of this species, detailed research of genetics and breeding studies are required. Recently, with the advancement of sequencing techniques, the genomics and comparative transcriptomics data were yielded, and quantitative trait locus (QTL) mapping has been conducted, along with genetic linkage maps constructed to this species. New strains have been bred and selected, with better characteristics, e.g. higher seawater temperature resistances and higher yields. In this review, we present the current status of genetic and breeding studies that have been performed to S. japonica in China, and provide guidelines for future developments in the areas of genetic selection and breeding for this species.

Keyword: Saccaharina japonica; next generation sequencing; genomics; transcriptomics; quantitative trait locus (QTL) mapping; DNA markers; linkage map; breeding; hybridization; cytogenetics

1 INTRODUCTION

Saccharina japonica is regarded as one of the most important, cultivated brown seaweeds in the world. In China, its large-scale industrialization started from the late 1950s. The current annual, global production is approximately 8 million tons (wet weight) with an approximate value of 4 millions US dollars. In the past, strain selection and breeding, classical quantitative genetic studies and limited cytological studies were summarized to S. japonica by Patwary and van der Meer (1992). At the beginning of the 21stcentury, based on the advancement of sequencing techniques (Shendure and Ji, 2008), genomics, comparative transcriptomics, proteomics, and QTL mapping were applied to study S. japonica due to its economic value (Hafting et al., 2015). In addition, new varieties of S. japonica have been developed, which provided enhanced characteristics of higher yields or temperature resistance, etc. This review presents the status of genetic and breeding research as applied to S. japonica in China, and the key future prospects in this fi led are outlined.

2 CULTIVATION OF S. JAPONICA IN CHINA

Fig.1 S. japonica life history and cultivation practice in China (rewritten from Duan et al., 2015)

Saccharina japonica is naturally distributed around the sub-littoral areas of the Northwestern Pacifi c, such as far-eastern Russia, Kamchatka Island, Hokkaido, the Kurile islands, and northern coastal area of Korean in the Japan Sea (Zhang et al., 2015c, 2019b). Previously it was named as Laminaria japonica (Lane et al., 2006). Lane et al. (2006) renamed L. japonica as S. japonica, by restoring it to the genus Saccharina, according to molecular phylogenies using multi-marker systems (Yoon et al., 2001; Lane et al., 2006; Bartsch et al., 2008).

In China, S. japonica is regarded as an alien species. Early in 1927, it was discovered attached to rocks on the coast of Dalian. Later, a Japanese technician, Mr. Yoshiro Otuki conducted the pioneer research on kelp cultivation trials using this species in Yantai in early 1943. However, his initial efforts failed (Fang and Zhang, 1982). Since then, Chinese phycologists and technicians have tackled two major bottlenecks in the cultivation of S. japonica: fi rstly, the ability to culture sporelings during summer with natural light and cool seawater in the nursery greenhouse, and secondly, the need for long-lines, or floating raft culture systems in the open sea. Once these issues were addressed, the industrial cultivation of S. japonica was eventually implemented (Tseng et al., 1955, 1957, 1962; Tseng, 1981; Su et al., 2017) (Fig.1).

With the mix of fertile male and female gametophytes, it was regarded as a new way for seedling nursery (Pang and Wu, 1996; Li et al., 1999; Zhang et al., 2008a). If under the suitable cultural conditions, the mixed female and male gametophytes could be induced to fertilize and yield into young sporophytes after 20 days. Although this method was comparatively efficient of time and cost, only limited company adopt this method for cultivars yielding on small scale in China, because of the weak adherence to the substratum to the young seedling at the initial stage, and the seedling lost in the production.

Presently, S. japonica commercialization was mainly conducted in Asia countries, especially in Korea, North Korea, Japan, and China. In 2016, the total wet weight production for this species in China was about 7.3 million tons, valued as 3.89 millions US dollars (FAO, 2018); still there are small scaleproduction in the far east of Russia (mainly in Primorie) (Selivanova et al., 2006).

Table 1 New cultivars developed for S. japonica

3 SELECTION AND BREEDING OF NEW VARIATIES FOR S. JAPONICA

At the present time, the breeding for new kelp cultivars was mainly performed in China, Korea, and Japan. Hwang et al. (2019) reviewed the breeding progress in the eastern Asian countries mainly for the fi ve species, S. japonica, Pyropia spp., Undaria spp., Cladosiphon okamuranus, and Nemacystus decipiens. To S. japonica, the fi rst cultivar “Haiqing No.1” was bred in 1962, which characterized of faster growth and late sorus matures and resistant to higher illumination (Fang et al., 1963). Recently, about 10 new varieties were selected or hybridized (Table 1), and were applied in the production in China.

For the breeding strategies, one way is directional selection during the continuously inbreeding (Fang et al., 1965), another way is hybridization of the gametophytes according to genetic distance, “Danhai No.1” and “Danza No.10” are the fi rst varieties hybridized (Fang et al., 1983, 1985). With the integration of DNA markers techniques, the molecular assistance in selection and breeding were applied in the kelp breeding, such as parental analysis to Dongfang No.2 (Shi et al., 2008), prediction of heterosis (Li et al., 2008a) and genetic variation to the new cultivars (Li et al., 2012; Zhang et al., 2016a, 2018a). Furthermore, it was adopted for the identifi cation of genotype for the phenotype (Liu et al., 2011; Wang et al., 2018), and we believed that this techniques will play more important role in the kelp selection and breeding.

4 DNA MARKERS AND THEIR APPLICATIONS IN GENETIC ANALYSIS

Usually, the DNA markers are classifi ed into three categories, hybridization-based, PCR-based and sequence-based (Table 2.1 in Xu, 2010). Previously, without knowing the genomic data of S. japonica, RAPD, AFLP, ISSR and SCAR were adopted for many genetic analysis, such as analyzing different morphological sub-species in Japan (Yotsukura et al., 2001), identifying different gametophytic germplasm varieties (Wang et al., 2004), discriminating female gametophytes (Liu et al., 2009b; Gu et al., 2014; Zhang et al., 2018c), and verifying the parent of one new developed hybrid variety (Shi et al., 2008).

Codominant SSR markers was preferred for population genetic analysis and genetic linkage map construction due to its highly polymorphism. So far, there are three resolutions for developing SSR marker for S. japonica. One way was developing SSR markers directly from SSR-enriched genomic library (Billot et al., 1998; Shi et al., 2007; Zhang et al., 2015a). Another way was developing microsatellites DNA markers through EST databases mining, and generated markers for genetic diversity analysis (Liu et al., 2010b; Wang et al., 2011). Liu et al. (2012b) ever developed 13 SSRs to S. japonica from the 2 668 non-redundant EST sequences of L. digitata. The last method was developing SSRs via the genomic and transcriptomic data (Zhang et al., 2014, 2016b, 2018b, 2019a; Li et al., 2016a; Peng et al., 2016). Li et al. (2016a) identifi ed 181 595 SSR loci using the MISA program, and applied the marker to 24 S. japonica individuals analysis.

Recently, huge SNP markers were developed for genetic linkage map construction and QTL mapping with the next generation sequencing platform (Zhang et al., 2015f; Wang et al., 2018). Generally, it is expected that the high throughout available SNP markers, will be the fi rst candidate for S. japonica genetics analysis in the future.

5 CYTOGENETICS STUDY

With the improvements of squash technique to chromosomes analysis, the chromosome number was characterized in Saccharina related species (Lewis, 1996). Its haploid number estimated were from 22 (Yabu, 1973; Tai and Fang, 1977), 28–35 (Nakahara, 1984), 16–24 (Lewis et al., 1993) and 32 (Yabu and Yasui, 1991) respectively. Zhou et al. (2004) reported the chromosome numbers of haploid male gametophyte was 31. Liu et al. (2012d) either observed the haploid male and female gametophytes was 31, and 62 for diploid sporophytes, and the size for each of 31 chromosomes were from 0.57–2.17 μm. Besides, many DNA markers were adopted for FISH (fluorescence in situ hybridization) to S. japonica, such as sex-specifi c marker (Liu et al., 2012c; Gu et al., 2014), nuclear ribosomal RNA genes (Liu et al., 2017b), Arabidopsis-type telomere sequence (TTTAGGG)n (Yang et al., 2017). Recently, a total of 1 576 scaffolds, accounting for about 65% of the assembled S. japonica genome sequences, was roughly mapped to the 31 genetic linkage groups (Fan et al., 2020), and the genome of E. siliculosus was assembled into 28 linkage groups, and it was near the chromosome level, referred to the 25-chromosome number (Cormier et al., 2017). So far, using singlecolor FISH, only several DNA markers have been successfully anchored to the chromosomes of S. japonica, respectively. It is expected that, in the near future, Multi-color FISH, which was fi rstly invented to study human chromosomes (Speicher et al., 1996), could be modifi ed and used to distinguish each of 31 tiny chromosomes with similar size.

6 MOLECULAR GENETICS FOR S. JAPONICA

6.1 Nuclear genomic studies

With the fast development of sequencing technologies, many seaweed nuclear genomes were sequenced (Table 2). These provide the data for illustrating the structure and function of seaweed genomes and functional genes, and understanding the evolutions of multicellular algae, and unique genes related to phycocolloids and abiotic stress (Cock et al., 2010; Prochnik et al., 2010; Collén et al., 2013; Nakamura et al., 2013; Zhou et al., 2013; Ye et al., 2015; Nishitsuji et al., 2016, 2019; Brawley et al., 2017; De Clerk et al., 2018; Lee et al., 2018; Sun et al., 2018; Nishiyama et al., 2018; Arimoto et al., 2019; Liu et al., 2019). Ye et al. (2015) fi rst reported one draft genome sequence to S. japonica, it characterized with 537 Mb from the female gametophytes of one S. japonica strain Ja, and contained 18 733 protein-coding genes annotated, it was larger than that of E. siliculosus with 16 256 genes. Compared to E. siliculosus, S. japonica process more signifi cant gene expansion in the 58 families, which contained mainly for iodine concentration, vanadium-dependent haloperoxidases, cellulosesynthase, mannuronan C-5-epimerases and alpha-(1,6)-fructosyltransferase. Although S. japonica genome has been sequenced, assembled only at the contig and scaffold level, the gold standard reference genome without gaps still should be developed with the new generation sequencing technology at the next step. At present, many protein-coding genes have been annotated, but few were studied experimentally at the biochemistrical level (see more in part 8). Only recently small RNAs (microRNA, lncRNA) and DNA methylation were analyzed to S. japonica genome (see more in part 6.3). Noncoding functional elements such as gene-specifi c enhancer in the genome of S. japonica have not been explored yet.

Table 2 Sequenced seaweed nuclear genomes

6.2 Transcriptomic studies

Crépineau et al. (2000) fi rstly reported transcriptional analysis to L. digitata with carbonconcentrating, cell wall biosynthesis and halogen metabolism related genes. To S. japonica, it was fi rst analyzed using EST sequencing data (Xuan et al., 2012), and more recently, RNA-seq techniques was adopted for analysis the gene transcriptional patterns related with light, temperature and growth stages (Deng et al., 2009; Wang et al., 2013a; Liu et al., 2014b; Ding et al., 2019; Shao et al., 2019a, 2019b). Under blue light conditions, 11 660 (16.5%) differentially expressed unigenes were detected compared with that in the dark treatments (Deng et al., 2009), while under 20°C, there are total of 947 up- or down-regulated genes (Liu et al., 2014b). Tanscriptomic studies for S. japonica have facilitated its genome annotation, and produced many differentially expressed genes under different conditions; however, it is still in its infancy. Integrated with the other data, such as small RNAs, proteome, to decipher the regulatory networks of S. japonica developmental biology, transcriptomic studies will be meaningful and helpful for its cultivation and breeding.

6.3 MicroRNA and DNA methylation

Fig.2 Chloroplast genome of S. japonica (rewritten from Wang et al., 2013b by OGDRAW version 1.3.1)

Recently, microRNA and DNA methylation were reported in S. japonica (Liu et al., 2015; Fan et al., 2020), Cock et al. (2017) indicated that none of S. japonica miRNAs sequences are similar with that of Ectocarpus sp., which implied that miRNAs evolve rapidly in the brown algae. Fan et al. (2020) reported that the cytosine methylations are important to the formation of diploid sporophyte and haploid gametophyte life-cycle stages, tissue differentiation and metabolism. Yang et al. (2020) proved that the non-coding RNAs participate in the regulation of CRY-DASH in the growth and early development of S. japonica. These primary data showed that, compared with the land plants and animals, S. japonica has lower levels of DNA methylation, and its microRNAs play a less central role in gene regulation, which is consistent with its simple morphology and far phylogenetic distance from higher plants and animals.

6.4 Chloroplast and mitochondria genomes

At present, over 10 chloroplast genome sequences from the brown seaweeds were reported (NCBI, https://www.ncbi.nlm.nih.gov/genome/organelle/), such as E. siliculosus and Fucus vesiculosus (Le Corguillé et al., 2009), Sargassum horneri (Liu and Pang, 2016), U. pinnatifi da and Costaria costata (Zhang et al., 2015d, 2015e), Dictyopteris divaricata (Liu et al., 2017a), S. confusum (Liu et al., 2018). Overall, the sizes of sequenced brown seaweed chloroplast genomes ranged from 124 kb to 140 kb, with the largest 139 954 bp for E. siliculosus (Le Corguillé et al., 2009). The G+C contents were from 28.94% to 31.19%.

Wang et al. (2013b) fi rstly reported the chloroplast genome of S. japonica with 130 584 bp in size, it consisted of 139 protein-coding genes, 29 tRNA genes, and 3 ribosomal RNA genes (Fig.2). The structure analysis showed that the S. japonica chloroplast genome has a large and a small singlecopy region (LSC and SSC), separated by two copies of inverted repeats (IR1 and IR2), which is the typical structure character for the known brown seaweed chloroplast genome.

The organelles’ Rubisco spacer sequences and others were applied for maternal inheritance pattern analysis in Fucus, Alaria and S. japonica (Kraan and Guiry, 2000; Coyer et al., 2002; Li et al., 2016b).

Since the report of mitochondrial genome of Pylaiella littoralis (Oudot-Le Secq et al., 2001), many other brown algal mitochondrial genomes were registered in NCBI. Generally, the size of brown algal mitochondrial genomes were from 31.6–58.5 kb, and it consisted of 3 ribosomal subunit genes (rRNAs), 24–26 transfer RNA genes (tRNAs), 35–36 proteincoding genes, and 2–16 uncharacterized open reading frames (ORFs), and some noncoding intergenic spacer regions. Yotsukura et al. (2010) analyzed the mitochondrial genomes of S. japonica, S. longipedalis, S. angustata and S. coriacea. The total size of S. japonica mitochondrial genome were 37 657 bp, which included 3 ribosomal RNA genes, 25 transfer RNA genes, 11 ribosomal small subunit protein genes and 6 large subunit genes, 10 NADH dehydrogenase genes, 3 cytochrome oxidase genes, 1 apocytochrome b gene, 3 f0-ATPase genes, 1 tatC gene and 3 ORFs.

The mitochondrial genome data were adopted for evolutional and phylogenetic studies. Liu and Pang (2015) revealed that P. fascia has a close evolutionary relationship with E. siliculosus. Balakirev et al. (2012) applied COI sequence, nuclear (ITS) and plastid (rbcLS) sequences for the morphological “longipes” (LON) and “shallow-water” (SHA) forms analysis, and indicated that the “typical” (TYP) and LON forms are genetically similar even with different phenotype and ecological variations. Moreover, the COI and trnW- trnL sequences in mitochondrial genome were either applied for phylogeographic patterns analysis in S. japonica populations (Zhang et al., 2015c), and rpl6- rps2 and trnW- trnL of mitochondrial DNA markers were adopted for maternal inheritance analysis to S. japonica (Li et al., 2016b).

7 POPULATION GENETICS AND QUANTITATIVE GENETIC

Neefus et al. (1993) used isozyme techniques to study Laminaria populations, and detected the lower genetic diversity within and between their populations, since then many seaweeds population genetic studies were conducted in algal biogeographical and phylogeographical analysis (Wattier and Maggs, 2001; Hu and Fraser, 2015; Zhang et al., 2015c, 2019b; Luttikhuizen et al., 2018). To the S. japonica, its cultivated population genetic diversities were lower than those in Russian and Japanese natural populations (Bi et al., 2011; Shan et al., 2011, 2017; Liu et al., 2012a; Yotsukura et al., 2016; Li et al., 2017; Zhang et al., 2017). Recently, Shan et al. (2019) proved that there are large genetic divergence in the farmed populations and the subtidal “wild” populations, and suggested that there is a need to introduce the wild S. japonica germplasm for enhancing the cultivars genetic diversity in China.

There were many traditional quantitative studies on the blade length, width and thickness, the stipe length, and the iodine content (Fang, 1983; Wu and Lin, 1987). With the integrations of feasible marker systems, QTL analysis to blade length and width were available, and usually F2populations were used for the QTL mapping (Li et al., 2007b; Yang et al., 2009; Liu et al., 2009a, 2010a; Zhang et al., 2015b, 2015f). Wang et al. (2018) used SLAF-seq (selective length amplifi ed fragment) techniques to generate 31 linkage groups, and localizing 12 QTLs to its blade length and 10 to width, either verifi ed that 14 Tic20 (translocon at the inner envelope membrane of chloroplasts) and three peptidase genes are candidates related with the blade length and width. One T ic20 gene from S. japonica was verifi ed in the diatom only (Chen et al., 2019). Compared with that in rice, where about 2 000 genes were reported to be cloned and partially studied (Li et al., 2018), although dozens of QTLs were successfully mapped especially to the blade length and width of S. japonica, few candidate genes were found to date.

8 FUNCTIONAL GENE IDENTIFICATION

Before the 21stcentury, for brown seaweeds, only chlorophyll a/ c-binding protein fucoxanthin gene in L. saccharina (Caron et al., 1996) and one metallothionein gene in F. vesiculosus (Morris et al., 1999; Lee and Lee, 2001) were analyzed. While in S. japonica, Fu et al. (2009) fi rstly cloned one cDNA hsp70 full sequences and quantitatively detected its expression under the different temperature, other genes such as Rubisco gene, trehalose-6-phosphate synthase gene, AUREOCHROME gene and candidate gametophyte sex determination gene ( SJHMG) were reported afterwards (Deng et al., 2014a, b; Shao et al., 2014b; Zhang et al., 2019c). Moreover, carbonic anhydrase gene, cbbX gene, mannuronan C5-epimerase gene and alginate lyase gene were either reported (Shi et al., 2010; Ye et al., 2014; Inoue et al., 2016; Inoue and Ojima, 2019). Besides, many genes were annotated after transcriptomic analysis, which related to blue and red light respectively (Deng et al., 2009; Wang et al., 2013a), under the temperature stress (Liu et al., 2014), and different developmental stages (Ding et al., 2019; Shao et al., 2019b). Biochemical characterization of mannitol-2-dehydrogenase (Shao et al., 2014a), GDP-mannose dehydrogenase genes (Zhang et al., 2016c), PMM/PGM (phosphomannomutase/phosphoglucomutase) (Zhang et al., 2018d), CRY-DASH gene (Yang et al., 2020) were either documented. Recently, Luan et al. (2019) verifi ed 40S ribosomal protein S6 was closely related to SjAUREO protein when detected with yeast two-hybrid library and interaction systems.

9 GENE MODIFIED TECHNIQUE

The technological platforms for gene transformation were established to crops including rice etc., especially with the Agrobacterium-mediated technology (Dunwell and Wetten, 2012). However, to develop feasible platforms for gene transformation in seaweeds is still progressing. Recently, one stable transformation was reported in U. mutabilis (Oertel et al., 2015). Mikami (2014) testifi ed the transformation methods in seaweed. The bombardment techniques were proved to be successful in many seaweeds, such as P. yezoensis (Hirata et al., 2011, 2014; Shin et al., 2016; Kong et al., 2017), K. albariza (Wang et al., 2010), U. prolifera (Wu et al., 2018), S. horneri (Pang et al., 2019) and S. japonica (Jiang et al., 2002; Zhang et al., 2008b; Li et al., 2009), even with the disadvantages of transient expressions in the seaweeds. In general, we believed that transgenetic system could be modifi ed with other techniques, such as RNAi, CRISPR-Cas9 manipulations, to generate one stable platform for gene transformation in S. japonica.

10 CONCLUSION AND PERSPECTIVE

With the new S. japonica genome information, we can annotate more genes with the actual functions. Nevertheless, there are many problems existed and need to be solved. Firstly, it is required to explore the function of those genes involved with alginate production and the mannitol pathway in vitro. Secondly, no genotype has been yet mapped to the phenotype of S. japonica, and it required verifi cation of those loci related to blade length and width. Thirdly, there is an urgent need to establish one stable platform for the functional gene verifi cation in the future. Lastly, the selection and breeding of S. japonica should be strengthened for sustainable aquaculture of commercial seaweeds in China, and the genetic and breeding strategies will be shifted from roughly to precisely in the future.

11 DATA AVAILABILITY STATEMENT

T he chloroplast genome analyzed in this study has been submitted to the NCBI database with the accession number NC_018523.1. All seaweed genomic sequences were retrieved from NCBI genome database.

12 ACKNOWLEDGMENT

We thank anonymous reviewers for their critical comments and suggestions for this paper, and Dr. Alan T Critchley for the help in English revision.