De novo and comparative transcriptome analysis of cultivated and wild spinach

Spinach (Spinacia oleracea L.) is an annual or biennial plant which belongs to the family Amaranthaceae. It is widely cultivated as an economically important green leafy vegetable crop for fresh consumption and processing1. The annual worldwide gross production of spinach in 2013 was approximately 23 million tonnes, of which around 91% was produced in China (FAOSTAT; http://faostat3.fao.org). Spinach is a rich source of iron, lutein, folate, vitamins, minerals, and antioxidants (USDA Nutrient Database; http://ndb.nal.usda.gov/ndb/search/list). Currently the major aims of spinach breeding programs are to develop varieties with traits including increased disease resistance (particularly against Peronospora farinosa downy mildew) and abiotic stress tolerance, late bolting, and improved yield and quality such as decreased levels of nitrate and oxalate, and increased levels of folate in spinach leaves2. Several markers linked to downy mildew resistance3 and sex determination4,5,6 have been developed. Efforts, although limited, have also been taken toward cloning genes of interest7,8,9 and functions of a few genes involved in stress responses have been characterized using transgenic approaches10. Despite considerable progress in the genetic improvement of spinach, it is still difficult to develop varieties with desirable traits, mainly due to the very limited genomic and genetic resources currently available for spinach. Spinach is a diploid species (2n = 2x = 12)4, with an estimated genome size of 989 Mb11. Currently, there are only 225 spinach expressed sequenced tags (ESTs) and 1,053 nucleotide sequences, among which the vast majority are chloroplast genome sequences, that are publicly available in GenBank. This leads to very limited molecular markers in spinach that are tightly linked with interesting traits. Recently the genome of sugar beet (Beta vulgaris ssp. vulgaris), another species in the Amaranthaceae family, has been reported12. For the purpose of comparative genomics and evolutionary analysis, Dohm et al.12 also generated a draft genome of a cultivated spinach, which was recently annotated13. Although the assembly represents only half of the spinach genome and contains many short assembled fragments, it contains the majority of the transcribed region13 and provides a valuable resource for spinach research and breeding. Furthermore, a more comprehensive spinach genome assembly is being generated (https://pag.confex.com/pag/xxiii/webprogram/Paper16426.html), providing additional valuable resource for spinach. In spinach, two known wild species S. turkestanica Iljin and S. tetrandra Stev. have been documented. The two wild species are found to be distributed over western parts of Asia, S. turkistanica in Turkmenistan, Uzbekistan, and Kazakhstan, and S. tetrandra in the Caucasus area, in Armenia and Kurdistan between Iran, Iraq, and Turkey13. The exact origin of the cultivated spinach is still unknown. The geographical distribution of these wild species and the generally high sexual compatibility with cultivated S. oleracea suggest that cultivated spinach may have originated through the domestication of one or both of the wild species14. The wild S. tetrandra and S. turkestanica have been used as parents to construct genetically broad segregating offspring populations which have been further used to construct genetic maps and to map genetic factors determining dioecious sex expression in spinach4,5,6. In addition, the two wild species have already proven to be valuable sources of different kinds of disease resistances15,16,17. However, so far, exploring the wild relatives for spinach improvement has been limited and the genetic structure of spinach germplasm remains largely unknown. Thus, developing genomic resources of spinach and further research on the genetic diversity and phylogenetic relationship of the spinach germplasm will provide valuable information that can be used for better germplasm utilization and for facilitating breeding of new spinach varieties. In this study, we report the transcriptome characterization of cultivated and wild spinach using the high-throughput Illumina sequencing technology. Strand-specific RNA-Seq libraries were constructed and sequenced for a total of nine spinach accessions including three from cultivated S. oleracea, three from wild S. tetrandra and three from wild S. turkestanica. The high-quality Illumina reads were de novo assembled into unique transcripts, which were then extensively evaluated and annotated. Single nucleotide polymorphisms (SNPs) and differentially expressed genes among the nine spinach accessions were identified and phylogenetic relationship and genetic diversity of cultivated and wild spinach were inferred. Our transcriptome data provide a valuable resource for future functional studies and marker assisted breeding in spinach.