Renanthera Lour. (Aeridinae, Vandeae, Orchidaceae) comprises about 19 species of epiphytic and/or lithophytic perennial herbs, mainly distributed in China, India, Indonesia, Malaysia, New Guinea, Philippines Vietnam, and the Solomon Islands (Chen and Jeffrey, 2009; Chase et al., 2015). Members of Renanthera produce a spectacular branched inflorescence containing numerous flowers with large, yellow, orange and red lateral sepals (Luo et al., 2012). Due to the magnificent floral features, these orchids have received great attention by orchid breeders and collectors as houseplants, floral design, and landscape gardening. However, their over-exploitation in the wild has disturbed their original habitat severely and threatened their survival in the wild. As a result, all members of Renanthera, except for R. imschootiana, which is listed in of the Convention on International Trade in Endangered Species of Wild Fauna and Flora (CITES) since 1975 (UNEP-WCMC, 2020). In China, three species are recorded, including R. citrina, R. coccinea, and R. imschootiana (Chen and Jeffrey, 2009); however, two of them, R. coccinea, and R. imschootiana are categorized as endangered and critically endangered, respectively, in the Red List of China Higher Plants (Qin et al., 2017).

In general, the morphological features, such as the number of pollinia, presence or absence of a nectar spur, shape and size of labellum, and shape of the column-foot, have been used as key characters in the taxonomic classification of Aeridinae (Pridgeon et al., 2014). However, the true phylogenetic relationships at subtribe level were unclear due to multiple evolutionary stages in the evolution of character states over a short period of time (Topik et al., 2005). Genetic studies on Renanthera are limited. Previous molecular phylogenetic studies of Aeridinae using a combined dataset of four plastid regions (atpI-atpH, matK, psbA-trnH, trnL-trnF) and the nuclear ribosomal internal transcribed spacer (ITS) region reportedly resolved the phylogenetic relationship within Renanthera (Zou et al., 2015). However, the phylogenetic tree was not supported with a strong backbone. The lack of informative sites resulted in a number of uncertainties in the molecular placements of many genera in the subtribe.

With the development of next-generation sequencing (NGS) technology, sequencing of the complete organellar genome has become simpler and affordable. The use of genome-scale data to resolve relationships in complicated plant groups appears promising. Coupled with advanced bioinformatic analytic tools, the phylogenetic relationships of many members of Orchidaceae, which has been uncertain for many years, were resolved (Li et al., 2019; Kim et al., 2020a). The complete plastid genome (plastome) sequences of 10 taxa of Vandeae, Kim et al. (2020b) revealed the molecular placement of Vandeae in the Epidendroideae. The phylogenetic relationship among members in Vandeae, with a focus on Aeridinae, was further identified in a phylogenetic reconstruction based on complete plastome sequences (Liu et al., 2020). Despite the well-supported tree reported by Liu et al. (2020), members of Renanthera were not included.

As an understudied yet threatened species in China, three species that are native to China, including R. citrina, R. coccinea, and R. imschootiana, as well as a domesticated species, R. philippinensis, were included in this study. To provide genetic data at genome level, we sequenced and characterized the complete plastomes of four species of Renanthera. At the same time, we analyzed the sequence repeats, the contraction and expansion of the inverted repeat (IR) region, identified the variable sites, and revealed the phylogenetic relationships between the four species of Renanthera to further elucidate evolutionary patterns in Renanthera at the plastome level. The findings obtained through this study yielded useful genomic information on these threatened species that may aid in conservation strategies.

In this study, we obtained novel plastome sequences of four species of Renanthera using next-generation sequencing techniques. Although there is a published report on the plastome sequence of R. imschootiana (Xin et al., 2020), information on its plastome sequence was not publicly available. Thus, it was not incorporated into this study. In this study, it was observed that the structure, gene content, and gene order in the four species of Renanthera were highly conserved. The plastome is generally maternal inherited in most angiosperms, including orchids, and usually comes with little recombination and a highly conserved structure across closely-related species (Wicke et al., 2011). Based on the study of Liu et al. (2020), the plastomes for members of Aeridinae were reported to range from 146,000 to 150,000 bp, except in two species, Schoenorchis seidenfadenii (142,859 bp) and Uncifera acuminata (145,191 bp). The species of Renanthera with the smallest plastome size, R. imschootiana, had a relatively smaller plastome size when compared to most of the members of Aeridinae; the three other species of Renanthera had a plastome size similar to most of the members of Aeridinae. The highly conserved number of genes present in the plastome of Renanthera suggested that the difference in plastome size could be due to the presence of indels in the intergenic spacer regions. However, similar to the rest of the members of Aeridinae, the overall GC content of Renanthera was not more than 37%.

The plastid ndh genes experienced independent loss in species of Renanthera (Table 2). Absence or pseudogenization of ndh genes in the plastomes of members of Aeridinae or Vandeae is considered a common phenomenon (Kim et al., 2020b; Jiang et al., 2022). The literature (Kim et al., 2015; Kim et al., 2020a) has reported that epiphytic orchids lose or pseudogenize their ndh genes when compared to terrestrial orchids, which is consistent with our findings for the four species of Renanthera in this study, which are epiphytes. Members of Renanthera rarely grows on the ground, they are usually epiphytic on tree trunks and/or lithophytic on rocks along valleys; thus, it is speculated that other species of Renanthera would undergo ndh gene loss, similar to those reported in this study. While it was proposed that the loss of ndh genes could be random in photosynthetic lineages, ndh gene loss is not only confined to Orchidaceae, but also reported in other autotrophic plants (Sabater, 2021). In general, ndh genes mediate cyclic electron transfer, which can protect plants from photoinhibition caused by various environmental stresses, and play an important role in maintaining efficient photosynthesis (Mi, 2016). The loss of ndh genes in the orchid plastome may be due to gene substitution or gene translocation from the plastome to the mitochondrial or nuclear genome (Lin et al., 2015). Usually, the loss of ndh genes in non-angiosperms, as in Pinus and Welwitschia (McCoy et al., 2008; Zeb et al., 2020), is not believed to be life-threatening. However, it could provide inverse effect to angiosperms as it will prevent plants from evolving and diversifying (Sabater, 2021), and will reduce the ecological adaptations of the species (Sun et al., 2020).

SSR are widely found in various species and are useful for studies of molecular evolution and genetic diversity (Doorduin et al., 2011). Of the 62–74 SSRs detected in the four species of Renanthera, the majority of them were identified from the LSC (73.2%) region, follow by SSC (19.4%) and IR (7.4%) region. Therefore, the distribution of SSRs in the chloroplast genome of Renanthera is uneven. The SSRs numbers in the LSC region of plastome of R. coccinea and R. philippinensis are 53 and 47, respectively. The distribution of SSR in the four species of Renanthera was consistent with that of most angiosperms (Xu et al., 2021; Sun et al., 2022; Xia et al., 2022).

The codon usage among the four plastomes of Renanthera was consistent across the four species; the CDSs of the four species of Renanthera had the same preferences in codon usage for all 20 amino acids, indicating that the nucleotide sequences in the CDS of the four species were identical, which is deemed accurate. This is because plastome sequences of closely related species have greater conservatism as they were inherited from their most recent common ancestor (Khan et al., 2021). Among these amino acids, leucine (n = 19,700) is the most frequently used, cysteine (n = 4,422) is the least universal amino acid in the cp genomes of these species. Alternatively, the high leucine frequency in the four plastomes can be attributed to the fact that leucine is largely needed in the chloroplasts due to its important function in photosynthesis-related metabolism (Knill et al., 2009). Cysteine, which was the least abundant amino acid among all the CDS analyzed, is a reactive element that can be considered toxic when it is accumulated above a certain level allowed by the host (Hildebrandt et al., 2015). This is consistent with the codon preference of other genera in the Orchidaceae, such as Thuniopsis and Polystachya. (Jiang et al., 2022; Li et al., 2022).

Contraction and expansion of the IR region is usually observed in most angiosperms (Li et al., 2019), including Renanthera. This was true when the 3’ end of the rpl22 gene of all four species of Renanthera that should be intact in the LSC region extended into the IRb region for 31–32 bp. The ycf1 gene in the SSC region of R. citrina was observed crossing over JSA, extending into the IRa region; while ycf1 of the remaining three species of Renanthera were intact in the SSC region. Generally, angiosperms retain the structure and size of the cp genome (Ebert and Peakall, 2010). However, due to evolutionary events such as contraction and expansion in the genome, variation can occur in the boundaries of inverted repeats (IRs) and single copy regions (LSC and SSC) (Raubeson et al., 2007) allowing certain genes to enter the IR or SC regions. Successive IR expansions have shown the importance of the JLA and JLB junctions for the analysis of evolutionary processes, providing clues about the origin and evolution of species (Zhang et al., 2013). Yet, in the case of Renanthera, although with the limited sampling size reported in this study, the finding suggested that the IR boundary shifts in Renanthera can be considered to be relatively minor, as this can be related to the small difference observed in the genome size across different species.

Sequence divergence analysis conducted on the four plastomes revealed that the IR region was highly conserved when compared to the single-copy regions. To our knowledge, there are no published reports on the molecular identification of the species of Renanthera. We speculated that this is either due to difficulty in species sampling, or the genus has not received attention. As molecular genetic tools could aid in the identification of the species of orchids and forensic detection at border controls (De and Pathak, 2018), the identification of useful gene regions in Renanthera will definitely contribute to safeguarding these orchids by security agencies. Molecular genetic tools such as DNA barcoding are among the promising techniques in the identification of species of land plants (Hollingsworth, 2011). The gene loci used in DNA barcoding usually contains sufficient informative sites that can delimit closely related species efficiently. The application of highly variable regions derived from the plastome sequence as potential DNA barcoding loci has been proven to be promising, as demonstrated in several findings (Vaughn et al., 2014; Lee et al., 2021). Three highly variable regions were identified in this study to facilitate marker development of Renanthera. It is noteworthy that an increased sampling size is needed to obtain reliable results (Li et al., 2015). The findings from this study may contribute to the development of DNA barcoding markers from Renanthera.

Despite being a valuable and threatened orchid, there are no reports on the phylogenetic relationships in Renanthera. According to the phylogenetic analysis using short gene sequences, Renanthera is monophyletic (Zou et al., 2015). However, due to small sampling size, we were not able to provide supporting evidence on this matter based on our phylogenetic analysis. Interestingly, the Aerides clade, where Renanthera is placed, was sister to Vanda clade when using the complete plastome sequence dataset. This finding was incongruent with the results by Zou et al. (2015) that used short gene sequences. Given that there are no known records on natural hybridization among species of Vandeae, the phylogenetic tree based on the plastome dataset is considered reliable (Liston et al., 2021). However, increased taxon samplings could reduce computational errors in phylogenetic tree reconstructions (Zwickl and Hillis, 2002); Zou et al. (2015) included 74 genera of Aeridinae in their analysis. In this study, only 24 genera of Aeridinae were included and the Aerides clade was represented by only one genus, Renanthera. Despite the fact that the phylogenetic analysis was well-resolved based on the plastome dataset, the finding by Zou et al. (2015) on the molecular placement of the Aerides clade in Aeridinae could be more reliable due to greater taxon sampling size. Thus, to ascertain the molecular placement of the Aerides clade in Aeridinae at the plastome level, it is suggested to include plastome data for several taxa of the Aerides clade such as species of Aerides, Arachnis, and Rhynchostylis.

Our study revealed that the plastome of Renanthera is well-conserved across four species, R. citrina, R. coccinea, R. imschootiana, and R. philippinensis. Phylogenetic analysis based on the CDS sequences revealed that Renanthera is closely related to Holcoglossum, Neofinetia, Pendulorchis, and Vanda and the relationship of the four species of Renanthera was monophyletic. The genomic data obtained from this study will provide a useful resource for conservation, as well as to contributing in studies on the phylogeny and evolution of Renanthera and Orchidaceae.