Plastome variation and phylogeny of Taxillus (Loranthaceae)

Abstract

Several molecular phylogenetic studies of the mistletoe family Loranthaceae have been published such that now the general pattern of relationships among the genera and their biogeographic histories are understood. Less is known about species relationships in the larger (> 10 species) genera. This study examines the taxonomically difficult genus Taxillus composed of 35-40 Asian species. The goal was to explore the genetic diversity present in Taxillus plastomes, locate genetically variable hotspots, and test these for their utility as potential DNA barcodes. Using genome skimming, complete plastomes, as well as nuclear and mitochondrial rDNA sequences, were newly generated for eight species. The plastome sequences were used in conjunction with seven publicly available Taxillus sequences and three sequences of Scurrula, a close generic relative. The Taxillus plastomes ranged from 121 to 123 kbp and encoded 90-93 plastid genes. In addition to all of the NADH dehydrogenase complex genes, four ribosomal genes, infA and four intron-containing tRNA genes were lost or pseudogenized in all of the Taxillus and Scurrula plastomes. The topologies of the plastome, mitochondrial rDNA and nuclear rDNA trees were generally congruent, though with discordance at the position of T. chinensis. Several variable regions in the plastomes were identified that have sufficient numbers of parsimony informative sites as to recover the major clades seen in the complete plastome tree. Instead of generating complete plastome sequences, our study showed that accD alone or the concatenation of accD and rbcL can be used in future studies to facilitate identification of Taxillus samples and to generate a molecular phylogeny with robust sampling within the genus.

Fig 1. Phylogenetic tree of Taxillus and other Loranthaceae.

This ML tree derives from analysis of the 67 gene sequence dataset (63 PCGs and four rRNAs). Only bootstrap values < 100% are labeled and Schoepfia jasminodora was used as the outgroup. Triangles at the branch tips represent multiple accessions of the same species and the number of accessions used is shown in parentheses. Four morphological characters and their states are shown to the right of the phylogenetic tree. The grey shading marks the five major clades of Taxillus (see text).

Fig 2. Comparison of plastome and nuclear rDNA ML gene trees for Taxillus species.

A. ML tree derived from an analysis of the whole plastome dataset. This tree is generally congruent with the one obtained from mitochondrial rDNA sequences. The Maximum Likelihood Bootstrap values (MLBS) and Maximum Parsimony Bootstrap (MPBS) values are indicated above and below the branches, respectively. B. Tree inferred from the ML analysis of the nuclear rDNA sequences. MLBS and MPBS values are indicated above and below the branches, respectively.

Fig 3. Nucleotide variability in 29 Taxillus plastomes.

Sequence diversity was calculated using a sliding window analysis (window size = 500 bp, step size = 100 bp). The six most variable regions are labeled.

Fig 4. Proportion of parsimony informative sites and gene lengths for 29 Taxillus plastomes.

Intergenic spacer regions (IGS), protein coding genes (PCG) and rrn genes (rRNA) are plotted by length (empty vertical bars, length from the sequence alignment) and proportion of parsimony informative sites (vertical filled bars).

Fig 5. A heatmap of pairwise genetic distance values of the plastome and nuclear rDNA sequences.

The upper right portion of the matrix contains genetic distances calculated from the nuclear rDNA sequence while the lower left portion contains genetic distance computed from whole plastome sequences. All the analyses were performed using the Kimura 2-parameter (K2P) model.