Targeted enrichment of novel chloroplast-based probes reveals a large-scale phylogeny of 412 bamboos

Abstract

Background: The subfamily Bambusoideae belongs to the grass family Poaceae and has significant roles in culture, economy, and ecology. However, the phylogenetic relationships based on large-scale chloroplast genomes (CpGenomes) were elusive. Moreover, most of the chloroplast DNA sequencing methods cannot meet the requirements of large-scale CpGenome sequencing, which greatly limits and impedes the in-depth research of plant genetics and evolution.

Results: To develop a set of bamboo probes, we used 99 high-quality CpGenomes with 6 bamboo CpGenomes as representative species for the probe design, and assembled 15 M unique sequences as the final pan-chloroplast genome. A total of 180,519 probes for chloroplast DNA fragments were designed and synthesized by a novel hybridization-based targeted enrichment approach. Another 468 CpGenomes were selected as test data to verify the quality of the newly synthesized probes and the efficiency of the probes for chloroplast capture. We then successfully applied the probes to synthesize, enrich, and assemble 358 non-redundant CpGenomes of woody bamboo in China. Evaluation analysis showed the probes may be applicable to chloroplasts in Magnoliales, Pinales, Poales et al. Moreover, we reconstructed a phylogenetic tree of 412 bamboos (358 in-house and 54 published), supporting a non-monophyletic lineage of the genus Phyllostachys. Additionally, we shared our data by uploading a dataset of bamboo CpGenome into CNGB ( https://db.cngb.org/search/project/CNP0000502/ ) to enrich resources and promote the development of bamboo phylogenetics.

Conclusions: The development of the CpGenome enrichment pipeline and its performance on bamboos recommended an inexpensive, high-throughput, time-saving and efficient CpGenome sequencing strategy, which can be applied to facilitate the phylogenetics analysis of most green plants.

Keywords: Bamboo phylogeny; Bambusoideae; Chloroplast; Probe; Targeted enrichment.

Fig. 1

The circle of alignment and depth sketch of a core CpGenome by BRIG. The CpGenome of Arabidopsis thaliana with a length of 154,478 bp was used as the core sequence of pan-genome. Please see the details for Methods. The inner circles show the alignment of 7 clade CpGenomes to A. thaliana using BLAST+. The black circle indicates gene positions, and adjacent colorful circles manifest the genome structure of A. thaliana. Based on DOGMA, the CpGenome was divided into four sections: Inverted Repeat A (IRA), Small Single Copy (SSC), Inverted Repeat B (IRB), and Large Single Copy (LSC). The outer circle shows the depth of the probes mapping to A. thaliana

Fig. 2

Evaluation of the pipeline performance in woody bamboos. a A dot plot provides the average depth (×) and coverage ratio of the 99 plant CpGenomes used to design the probes. The red and blue dots represent bamboos and other plant species, respectively. The black lines represent the average depth (×) and a coverage ratio of the bamboo species. b A dot plot provides log10(cover length) and the coverage ratio of the 468 plant CpGenomes used to evaluate the probes. The red and blue dots represent bamboos and other plant species, respectively. The black lines represent log10(cover length) and the coverage ratio of the bamboo species, respectively. c A box plot of gene number, genome size, and raw bases (bp) of the sequenced bamboos CpGenomes in this study. d Evaluation of mapping and coverage of the probes compared to the in-house and released bamboo CpGenomes. The mapping ratio represents the proportion of reads obtained by the probes aligned with the released bamboo CpGenomes. Mapping coverage represents the proportion of the assembled CpGenomes based on the probes aligned with the released bamboo CpGenomes

Fig. 3

A species tree of Phyllostachys clade based on 76 chloroplast genes. The species tree divided into 2 parts, which labeled with different background colors. Numbers at the node indicated the bootstrap values and bootstrap values lower than 80 were concealed. The red, purple, grey, and blue blocks in the tree represented P. sect. Heteroclada species, P. sect Phyllostachys species, unlabeled and non-Phyllostachys species, respectively. Name with ‘LOC’ stands represented newly sequenced sequences in this study