Comparative chloroplast genomics of six Bupleurum (Apiaceae) accessions: candidate barcodes, phylogeny based on available plastomes, and candidate RNA-editing sites

Abstract

Bupleurum L . (Apiaceae), a taxonomically intricate genus of about 190 species and a source of Radix Bupleuri (Chai Hu), is difficult to discriminate because of convergent morphology, infraspecific variation, and limited genomic sampling. This study aimed to characterize plastome variation, identify and validate candidate molecular markers, reconstruct plastid phylogenetic relationships, and assess candidate plastid RNA-editing sites in Bupleurum . We assembled six plastomes from subgenus Bupleurum , screened 51 Bupleurum plastomes for diagnostic loci, reconstructed whole-plastome and partitioned protein-coding-sequence phylogenies, and predicted plastid C-to-U RNA-editing candidates across the six newly assembled plastomes using a PREP-Cp-compatible workflow. Candidate barcode performance was evaluated against the reference plastome phylogenies, and codon-based models were used to test for positive selection. The plastomes were 154,496–155,778 bp with the canonical quadripartite structure and GC contents of 37.67–37.73%. Gene content was stable (131–132 genes; 86–87 protein-coding genes); B. falcatum subsp. cernuum lacked ycf15 but contained an additional inverted-repeat-associated ycf1 annotation. A/U-ending synonymous codons were favoured. Finite pairwise Ka/Ks estimates were below 1 for most genes, and site-specific codon models detected no positive selection. Each plastome contained 55–61 pure microsatellites, dominated by A/T mononucleotide motifs. MarkerSeek ranked 265 features and identified atpF–atpH , petA–psbJ , rpl32–trnL-UAG , and ycf1 as leading candidate barcodes. ycf1 recovered 38 of 41 nodes strongly supported by both reference trees, whereas a partitioned four-locus analysis recovered 40 of 41 and distinguished all 51 accession sequences. However, only one of seven multi-accession operational binomial groups was monophyletic, and only one showed a positive local barcode gap. The whole-plastome phylogeny recovered Bupleurum as monophyletic relative to Chamaesium . The two sampled Penninervia accessions occupied early-diverging positions without forming an exclusive clade. B. falcatum subsp. cernuum was sister to B. ranunculoides , with B. ranunculoides subsp. telonense sister to that pair. A partitioned 74-CDS analysis recovered the same key relationships and 45 of 50 internal bipartitions. Across the six newly assembled plastomes, 57–63 nonsynonymous C-to-U candidates were predicted per accession (367 total) in 21–22 genes; 269 affected the second codon position and 98 the first. Bupleurum plastomes are structurally conservative but retain localised divergence useful for marker development. Concordant whole-plastome and CDS genealogies support genus monophyly, whereas sparse Penninervia sampling and maternal plastid inheritance preclude rejecting traditional subgeneric classification. The predicted RNA-editing sites represent candidates for future experimental validation rather than an established Bupleurum editome. These genomic resources support authentication, conservation, and evolutionary research in Bupleurum .