Insights into the phylogenetic relationship of the lamiids orders based on whole chloroplast genome sequencing

1 Introduction

Asterids is the largest group in eudicots and comprising two clades, campanulids and lamiids (Zhang et al., 2020; Li et al., 2021). The lamiids are the largest, most species-rich and most diverse clade of asterids, with estimates ranging between 40,000 to 50,000 species, representing 15% of eudicots (Refulio-Rodriguez and Olmstead, 2014; Yang et al., 2020). According to the Angiosperm Phylogeny Group (APG IV, 2016), the clade is composed of eight orders, Icacinales (Include Icacinaceae and Oncothecaceae), Garryales and Metteniusales, referred to informally as “basal lamiids”. The rest of the lamiids include Boraginales, Gentianales, Lamiales, Solanales, and Vahliales, known informally as “core lamiids”. Different phylogenetic studies strongly favoured the monophyly of the clade (Albach et al., 2001; Soltis et al., 2011; Refulio-Rodriguez and Olmstead, 2014). The lamiids has been referred to by several names including asterids I, euasterids I or lamiidae (Takhtajan, 1987; Chase et al., 1993, APG I, 1998. Olmstead et al., 2000; Soltis et al., 2000, APG II, 2003; Soltis et al., 2011). Multiple characters of lamiids have been inferred as ancestral states such as woody habit, superior ovaries, unitegmic ovules, trilacunar nodes, scalariform perforation plates, presence of iridoids cellular endosperm and opposite leaves (Stull et al., 2018; Wang et al., 2022). However, considerable morphological differences within the lamiids clade, the non-molecular synapomorphies are uncertain (Stull et al., 2015).

The formation of lamiids has been clarified by numerous phylogenetic studies during the last few decades. However, the relationships among the orders of lamiids remain unclear or inconsistent. Firstly, the evolutionary relationships amongst the core lamiids group, Boraginales, Gentianales, Lamiales, Solanales, and Vahliales are contradictory (Fig. 1 A-G). Based on one nucleus gene (18S) and three chloroplast genes (atpB, ndhF and rbcL) Albach et al. (2001) (Fig. 1-A) have found support for the relationships (((Lamiales–Boraginales), Vahliales), Gentianales–Solanales) with strong Maximum Parsimony (MP) support. Bremer et al. (2002) (Fig. 1-B) recovered a different topology (((Lamiales–Solanales), Boraginales–Vahliales), Gentianales) using six plastid genes (rbcL, ndhF, matK, trnL, trnV and rps16), however, the MP support was low. Qiu et al. (2010) (Fig. 1-C) based on four mitochondrial marks (atp1, matR, nad5, and rps3) obtained different relationships ((((Lamiales–Vahliales), Gentianales), Solanales), Boraginales) but with low Maximum Likelihood (ML) support. Soltis et al. (2011) (Fig. 1-D) recovered the pattern (((Boraginales–Lamiales), Solanales–Gentianales), Vahliales) using 17 genes including two nucleus genes (18S and 26S rDNA), eight plastid genes (rbcL, rpoC2, rps16, rps4, psbBTNH, ndhF, matK, and atpB), and four mitochondrial marks (rps3, nad5, matR, and atp1) with strong ML support. Refulio-Rodriguez and Olmstead (2014) (Fig. 1-E) using one mitochondrial gene (rps3) and nine chloroplast genes (trnV-atpE, trnL-F, rps16, rps4, rbcL, psbBTNH, ndhF, matK and atpB) have recovered (((Boraginales–Lamiales), Solanales–Vahliales), Gentianales) with moderate ML support but strong Bayesian Inference (BI) support. Stull et al. (2015) (Fig. 1-F) have recognized a different topology (((Lamiales – Vahliales), Solanales), Gentianales–Boraginales) based on 73 protein-coding genes, with strong BI support, however, ML support was low. Zhang et al. (2020) (Fig. 1-G) recovered (((Boraginales–Gentianales), Solanales), Vahliales–Lamiales) based on 213 transcriptomes/genomes, with strong ML support.

Close

Fig. 1

Seven different topologies from previous phylogenetic studies show different results in identifying the evolutionary relationship among the core lamiids.

Export to PPT

Secondly, the relationships between the other three lamiids orders, Icacinales, Garryales, and Metteniusales are controversial. Soltis et al. (2011) (Fig. 2-H) recovered the pattern ((Icacinaceae– Garryales), Oncothecaceae) with moderate ML support, however, Metteniusales order was not included. Refulio-Rodriguez and Olmstead (2014) (Fig. 2-I) recovered a different relationship (((Icacinaceae), Metteniusales–Oncothecaceae), Garryales) with weak BI, ML and MP support. Stull et al. (2015) (Fig. 2-J) have found support for the relationships ((Icacinales–Metteniusales), Garryales) with relatively strong ML support. Moreover, Metteniusales was found to be the immediate sister to the core lamiids in Gonzalez et al. (2007), Icacinales by Soltis et al. (2011), and it was Garryales in Refulio-Rodriguez and Olmstead (2014) and Stull et al. (2015).

Close

Fig. 2

Three different topologies from previous phylogenetic studies show different results in identifying the evolutionary relationship among the basal lamiids.

Export to PPT

Generally, as reviewed above, the relationships among the major lineages of lamiids have suffered from instability. Most of the phylogenetic studies that covered the evolutionary relationships between the orders of the lamiids were based on a few markers or genes (including chloroplast, nucleus and mitochondrial DNA). Using a single or a few genes may lead to different results in identifying the evolutionary relationship among species in comparison to chloroplast (cp) genome sequencing, which is based on the whole genome (Yao et al., 2020). Since the plastid genome approach has the capacity to resolve evolutionary relationships among some complex taxa (Huang et al., 2020). The plastid genome has been extensively used in plant phylogenetic studies (Adachi et al., 2000). The chloroplast genome of angiosperms species highly conserved in term of structural, gene content and arrangement (Fonseca et al., 2022). In this work, we sought to cover the relationships among the major lineages of lamiids reported in previous studies and determine the closest relatives of each order using chloroplast genome sequences data. Systematics is discussed in light of the phylogenetic analyses.

2 Materials and methods

3 Results

The topologies resulting from the Maximum Parsimony and Bayesian analyses were virtually identical. The phylogenetic tree is presented in Fig. 3 with bootstrap (BS) and posterior probability (PP) support values. The basal lamiids together with core lamiids formed a strongly supported clade (BS = 100, PP = 1).

Close

Fig. 3

Phylogenetic tree reconstruction of the 49 species based on the whole plastid genomes using Maximum parsimony (MP) and Bayesian inference (BI) analyses; the tree illustrate the relationships among the major lineages of lamiids and the figures in the branch nodes represent the total of bootstrap (BS)/posterior probability (PP).

Export to PPT

4 Discussion

5 Conclusion

Larger-scale phylogenetic studies of angiosperms have not definitively determined the relationships among the major lineages of lamiids. Most of the molecular analyses that covered the evolutionary relationships within the lamiids were based on a few markers or genes. Using whole chloroplast genome sequencing gives a more reliable result in identifying the evolutionary relationship among species in comparison to the use of a few genes. In this study, we used 49 whole chloroplast genome sequences, 47 taxa that represent all major lineages identified to date within lamiids except Vahliales; the cp genome of this order was not available, while two taxa represent campanulids clade as outgroup. In basal lamiids clade, our results suggest that Garryales as sister to Metteniusales, while Icacinales is the immediate sister to the other orders within the core lamiids. In core lamiids, Lamiales was found to be sister to Solanales, and Boraginales as sister to Gentianales. This finding increases our understanding of relationships among lineages of lamiids. However, these relationships need more fully investigated with additional sequence data and taxa, especially the chloroplast genome sequence of Vahliales taxa, which was not sampled in the present study.