The complete chloroplast genome sequence of Lilium tsingtauense Gilg (sect. Martagon, Liliaceae).

Lilium tsingtauense Gilg, known as Twilight Lily and distributed in Korea and eastern China, possessed a 151,983 bp chloroplast genome (cpDNA) sequence composed of a large single copy (81,424 bp), a small single copy (17,575 bp), and two inverted repeat regions (26,492 bp each). This cpDNA encoded 81 protein-coding genes, 30 tRNAs and four rRNAs. In contrast to the overlaps between ycf1 and ndhF at the IR/SSC junctions in other Lilium species, there was no record of this pattern in L. tsingtauense. Moreover, variable poly A sequences located downstream of start codon caused different annotations of cemA among Lilium species.

The Lilium is one of the largest genera in Liliaceae, comprising approximately 110 species distributed in temperate regions of the Northern hemisphere (Govaerts 2015). There have been many studies on its classification (Baker 1874; Wilson 1925; Comber 1949; Lee & Kim 1993; Dubouzet & Shinoda 1999; Hayashi & Kawano 2000; Sultana et al. 2010; Gao et al. 2013; Kim et al. 2015). However, only four records of Lilium’s complete cpDNA sequences were recorded in NCBI Organelle Genome Resources (http://www.ncbi.nlm.nih.gov/genome/browse/?report =5). Therefore, it is crucial to enlarge the data of complete cpDNA sequences which can be useful for understanding the evolution of Lilium and related species in Liliaceae.

Lilium tsingtauense Gilg was sampled at Mt. Namdeogyu (Geospatial coordinates: N: 35° 45′39.96″, E: 127° 40′40.87″) in Korea, and voucher specimens were deposited in the herbarium of Gachon University (Herbarium number: GCU1407796). Complete chloroplast genome of L. tsingtauense (GenBank accession no. KU230438) was sequenced using Hiseq sequencing system and long range PCR method. It was 151,983 bp and comprised a large single copy region (LSC) of 81,424 bp, a small single copy region (SSC) of 17,575 bp and two invert repeat regions (IR) of 26,492 bp. The AT and GC contents are 62.9% and 37.1%, respectively. The cpDNA portions of coding regions are 55.74%, and non-coding regions including intron and intergeneric spacer region are 44.26%. A total of 135 coding regions were identified in the L. tsingtauense cpDNA genome. Among them, there were 81 protein-coding genes, 30 tRNAs and 4 rRNAs. Notably, the infA was pseudogene because of the presence of several internal stop codons within its coding region. In general, the cpDNA structure, gene content and gene order of L. tsingtauense are similar to previously reported species of L. longiflorum, L. superbum and L. hansonii. However, L. tsingtauense does not have the overlap between Ψycf1 and ndhF which was previously detected in cpDNA of Lilium. In monocots, this overlap was found in IR/SSC boundary of Arecaceae (Yang et al. 2010; Huang et al. 2013). The magnoliids species (i.e. Drimys granatensis, Chloroanthus spicatus) also showed such overlap event (Cai et al. 2006; Hansen et al. 2007). Based on the observation of IR/SSC junctions in orchid’s cpDNA, Luo et al. (2014) suggested the route of IR/SSC junction in which during the expansion of ycf1, the overlap between ndhF and ycf1 occurred from three independent events. Comparative analysis (except one IR region) among four Lilium cpDNAs revealed a total of 96 insertion and deletion (indels) regions. Among them, the majority of indels (75 regions) was found in LSC, whereas only 10 and 11 regions were detected in IR and SSC regions, respectively. Moreover, most of these indels were located in non coding regions (90.6%). Phylogenetic analysis inferred from 78 protein-coding genes of chloroplast genome showed a close relationship of L. tsingtauense and L. hansonii with high bootstrap percentage and posterior probability (Figure 1).

Figure 1.

Phylogenetic tree inferred by maximum-likelihood and Bayesian method using 78 protein-coding gene sequences of 18 species including L. hansonii (NC_027674), L. tsingtauense (KU230438), L. longiflorum (KC968977), F. cirrhosa (NC_024278), S. china (HM536959), H. tubiflora (KM078036), C. japonica (KF951065), P. verticillata (KJ433485), T. maculatum (NC_027738), X. tenax (KM0078035), V. patulum (KF437397), G. superba (KP125338), C. autumnale (KP125337), B. edulis (KM233641), A. aurea (KC968976), L. radicans (KM233640), C. lineare (NC_026785) and A. trichopoda (NC_005086) . Bootstrap percentage and posterior probability are shown above and below branches. The phylogenetic analysis was constructed using web-based RAxML tool (http://embnet.vital-it.ch/raxml-bb/; Stamatakis et al. 2008) and MrBayes ver. 3.2.6 program (Huelsenbeck & Ronquist 2001). The model of GTR + I + R resulted from jModelTest ver 0.1.1 (Posada 2008) was used in both Maximum likelihood and Bayesian analysis with 100 replications and 1,000,000 generations, respectively. The phylogenetic tree was manually modified using Figtree v1.4 (http://tree.bio.ed.ac.uk/software/figtree/).