Complete chloroplast genome sequence of Barleria prionitis, comparative chloroplast genomics and phylogenetic relationships among Acanthoideae

Background The plastome of medicinal and endangered species in Kingdom of Saudi Arabia, Barleria prionitis was sequenced. The plastome was compared with that of seven Acanthoideae species in order to describe the plastome, spot the microsatellite, assess the dissimilarities within the sampled plastomes and to infer their phylogenetic relationships. Results The plastome of B. prionitis was 152,217 bp in length with Guanine-Cytosine and Adenine-Thymine content of 38.3 and 61.7% respectively. It is circular and quadripartite in structure and constitute of a large single copy (LSC, 83, 772 bp), small single copy (SSC, 17, 803 bp) and a pair of inverted repeat (IRa and IRb 25, 321 bp each). 131 genes were identified in the plastome out of which 113 are unique and 18 were repeated in IR region. The genome consists of 4 rRNA, 30 tRNA and 80 protein-coding genes. The analysis of long repeat showed all types of repeats were present in the plastome and palindromic has the highest frequency. A total number of 98 SSR were also identified of which mostly were mononucleotide Adenine-Thymine and are located at the non coding regions. Comparative genomic analysis among the plastomes revealed that the pair of the inverted repeat is more conserved than the single copy region. In addition high variation is observed in the intergenic spacer region than the coding region. The genes, ycf1and ndhF and are located at the border junction of the small single copy region and IRb region of all the plastome. The analysis of sequence divergence in the protein coding genes indicates that the following genes undergo positive selection (atpF, petD, psbZ, rpl20, petB, rpl16, rps16, rpoC, rps7, rpl32 and ycf3). Phylogenetic analysis indicated sister relationship between Ruellieae and Justcieae. In addition, Barleria, Justicia and Ruellia are paraphyletic, suggesting that Justiceae, Ruellieae, Andrographideae and Barlerieae should be treated as tribes. Conclusions This study sequenced and assembled the first plastome of the taxon Barleria and reported the basics resources for evolutionary studies of B. prionitis and tools for phylogenetic relationship studies within the core Acanthaceae.


Background
The Acanthaceae Juss. Ex Bercht.& J. Presl is among the largest family in the order Lamiales with ca. 3800 recognized species accommodated in ca. 200 genera [1], the members of the family are mainly diversified in the sub tropics and tropics, with few species in the temperate zones [2]. The family is close to Bignoniaceae family in the Lamiales order [3]. The main centres of distribution of the species in the family are Africa, Central America and Asian continent particularly Malaysia, Indonesia and Brazil [4]. They are characterized by having decussate phyllotaxis, while some species have congest whorled phyllotaxis, the leaves are usually simple with toothed margin, opposite, existipulate and contained calcium oxalate crystals or hypodermal calcium carbonate cystolith [5,6].
In an effort to resolve taxonomic issues of the family and its species, researchers for the past decades works extensively in delimiting the family [7][8][9][10], identifying major clades in the family [11][12][13][14]. Scotland and his colleagues carried out infrafamilial studies using floral parts [15][16][17], their findings gives more insight on the infra familial classification of the family and gives morphological synapomorphies of the major lineages. Recently, phylogenetic approach was used to reveal the relationships between the lineages [18][19][20]. Despite these researches, the classifications of the species within the Acanthoideae are still not clear.
The chloroplast organelle is one the most distinguishing featured that differentiates plant cell and other type of cells; therefore it is the most noticeable feature in plants. The organelle which is semi-autonomous is believed to have evolved decade of millions years ago from cynobacterium [21,22]. The plastome of flowering plant is conserved than the other genomes (i.e mitochondrial and nuclear genomes), in addition the genome is small compared with the others and it is used frequently in phylogeny studies due to its low rate of nucleotide substitution [23]. The chloroplast genome is typically quadripartite in structure, containing large single copy (LSC) and small single copy (SSC) separated by pair of inverted repeat (IR) [24]. The genome organization, its content and gene structure are highly conserved [25]. Due to its conserved nature, the cp genome contents are widely used by researchers as a tool to investigate phylogenetic relationship and in genomic studies [26]. Single nucleotide polymorphisms as well as insertion/deletions which are among the evolutionary hotspot of the organelle are believed to be use as a tool to solve taxonomic issues among taxa that their phylogenetic relationships are unresolved. Phylogenetic relationship generated from single or combination of few genes are being replaced by the ones constructed from the whole genome as a result of new DNA sequencing methods such as next generation sequencing (NGS). The introduction of next generation sequencing has increased the availability of data for solving phylogenetic relationship issues. However, in spite of its importance, the approach is not fully and well utilize by researchers in plant systematic studies [27][28][29]. One of the most important benefits of next generation sequencing technique is that it generate very high amount of sequences compared with sanger sequencing technique. Additionally, the platform used in next generation sequencing like Illumina is very cheap process [30]. This approach has been used to generate huge number of data for inferring phylogenetic relationship in different taxonomic levels inference [31][32][33][34].
With the advent of next generation sequencing, importance of plastome sequence in resolving phylogenetic relationships and the great number of genera in Acanthaceae, only plastome of few genera have been sequenced and no phylogenomic studies have been conducted for the family.
In this research, we sequenced and characterized the plastome of Barleria prionitis and compared the genome with cp genomes from Acanthoideae species. We used data from the whole chloroplast genome of 8 genera belonging to the Acanthoideae to reveal their tribal positions. This is as a result of incongruent of previous studies in placing the genera in their respective tribes [35]. placed Barlerieae and Andrographideae as sub tribes under the tribe Justicieae, this classification has been reported by other student of Acanthaceae [27]. classify the sub family Acanthoideae into two tribes, placing Ruelliinae, Justiciinae, Andrograpiinae and Barleriinae under the tribe Ruellieae. Findings of recent studies by McDade and her colleagues using molecular data contradict with previous classifications. Therefore, there is need to use complete chloroplast genome to address the correct placement of the genera into their respective tribes. The result of this study will be useful for developing makers, provide resources for evolutionary studies and authentication of B. prionitis and the inference of phylogenetic relationships within Acanthoideae.

Characteristics of B. prionitis chloroplast genome
The complete plastome sequence of B. prionitis was reported to be 152,217 bp in size and has a structural organization of quadripartite containing a large single copy (LSC, 83, 772 bp), a pair of inverted repeat (IRa and IRb 25, 321 bp each) and small single copy (SSC, 17, 803 bp) ( Fig. 1  The complete chloroplast genome of B. prionitis contained 113 different genes out of which 18 are duplicated in the IRA and IRB region, totaling 131 genes. The number of rRNA genes, tRNA genes and protein-coding genes in the genome are 4, 30 and 80, respectively ( Fig. 1 and Table 2). Four rRNA, seven protein coding and tRNA genes are located in the pair of the inverted repeat region of the plastome whereas the large single copy region harbored 62 protein-coding sequence and 22 tRNA genes, the remaining one tRNA and 12 protein coding genes are located in the single copy region. Among the genes coding for protein, many of them started with the codon ATG while few starts with other codon such as ACG and GTG, this is also reported in other chloroplast genome of angiosperms.

Ribosomal proteins
Small subunit of ribosome rps2, rps3, rps4, rps7 c , rps8, rps11, rps12 c , rps14, rps15, rps,16 a , rps18, rps19 Transcription Large subunit of ribosome rpl2 a,c , rpl14, rpl16, rpl20, rpl22, rpl23 a , rpl32, rpl33, rpl36  intron viz.: ndhB, trnA-UGC, trnI-GAU and rpl2 are situated in the inverted repeat region and the other 12 in the large single copy region. clpP and ycf3 are the only genes with two intron, while the other 12 genes have one intron, this is consistent with that of S. cusia [36]. trnK-UUU is the gene with longest intron with 2460 bp because of the situation of matK in the gene. The frequency of the codon usage present in the plastome of B. prionitis was computed using the nucleotide sequence of protein-coding genes and tRNA genes 100,319 bp, the result is presented in Table 4, the results showed the genes in the plastome are encoded by 33, 436 codons. The codons that codes for the amino acids Leucine appears more frequently in the genome 3286 (9.83%) (Fig. 2), comparable to that of Ailanthus altisssima and the ones coding for Trp have the lowest 622 (1.86%) in the plastid sequence. Guanine-Cytosine ending are more common than the Adenine-Thymine ending, this is incongruent with other cp genome sequence [38][39][40]. The result of the analysis show that there is low codon usage bias in the plastome sequence of B.prionitis (Table 4). 29 codons have RSCU values greater than 1 and all of them are characterized with Adenine-Thymine ending while for 30 codons, were less than 1 and are all of Guanine-Cytosine ending. The amino acids The prediction of RNA editing sites present in the plastome sequence of B. priniotis was done by means of PREP suite. The first codon of the first nucleotide was used in all the analysis. The results as shown in (Table 5) showed that most of the conversions in the codon positions are from Serine to Leucine. Generally, the editing sites observed in the plastome were 61 which are distributed between the 19 protein-coding genes. psaB is found to have the highest number of editing site (13 sites) followed by ndhB (9 sites), rpoB (6 site) and rpl20, accD, rps, atpI, rpl2, rpoA have the lowest number of editing site with 1 editing site each. Nine (9) RNA editing site in ndhB has been confirmed in the plastome of other species [41][42][43]. Conversions of proline to serine were observed, which involves the changing of the amino acids in the RNA editing site from apolar to polar group. Genes such as petD, ndhC, atpB, clpP, ndhE, petL, ndhG, petG and ccsA among others do not possess RNA editing site in their first codon of the nucleotide.

Long repeats
Repeat sequence in the chloroplast genome of B. prionitis were screen using REPuter programme with default settings, the programme revealed that only three types of repeats were present in the genome viz. Palindromic, forward and reverse, the complement repeat is not detected within the plastome ( Table 6). The result revealed 18 palindromic repeats, 25 forward repeats and 6 reverse repeats ( Table 6). Most of the repeats size are between 20 and 29 bp (78.6%), followed by 10-19 bp (10.20%) whereas 40-49 bp are the least (4.08%). In all, there are 49 number repeats in B. priniotis plastome. In the first location, 65.30% of the repeats are contained in the non coding region; this is comparable to the cp genome of Fagopyrum dibotrys [44]. Eight repeats were located in the tRNA (16.32%), the other 9 repeats (18.36%) are situated in the protein coding genes in particular rpl2, ndhA, ycf1, ndhC, and ycf2. Among the protein coding genes ycf2 contained 2 forward palindromic and repeats.
The rate of repeats among eight Acanthoideae plastomes was compared, the results indicates that complement, palindromic, reverse and forward type of repeats occurred in the plastome of J. flava, A. paniculata, S. cusia, B. ciliaris and R. breedlovei, whereas no complement repeats detected in the cp genomes of B. prionitis, E. attenuatus and A. knappiae (Fig. 3). S. cusia, B. ciliaris and A. paniculata are found to have high frequency of palindromic repeats (23) and J. flava is found to have the least (16). R. breedlovei, S. cusia and A. paniculata have15 forward repeats in their plastome and the frequency of reverse repeats is identical in the plastome of A. paniculata, S. cusia and J. flava. Complement repeat is absent in B. prionitis, E. attenuates, A. knappiae and is the least repeat in the plastome of J. flava, A. paniculata, B. ciliaris, R. breedlovei and S. cusia.

Microsatellite analysis
Microsatellites (SSRs) are short repeat of nucleotide sequences (1-6 bp) that are distributed throughout genome. This short repeats are used as important makers for evolutionary studies of plants [45]. In this research, a total number of 98 microsatellites were identified in the chloroplast genome of B. priniotis ( Table 7). Most of the microsatellites in the plastome are mononucleotide (83.67%) and majority of them are polythymine 58.53% followed by poly A (polyadenine) 40.24%, only one Poly G (polyguanine (1.21%) is present where as no poly C detected in the genome. Among dinucleotide only 5 repeats were detected, TA repeated four times and AT only once. Considering sequence complimentary, two trinucleotide AAG/CTT and AAT/ATT, four tetra AAAC/GTTT, AAAG/CTTT, AAAT/ATTT, AATC/ ATTG and only one penta AAATGG/ATTTCC were detected in the genome (Fig. 4a) whereas no   The plastome sequences of eight Acanthaceae species namely (B. prionitis, J. flava, B. ciliaris, A. paniculata, E. attenuatus, R. breedlovei, A. knappiae and S. cusia were compared. To check the level of nucleotide sequence variation between the sampled plastomes of Acanthoideae species, the programe mVISTA was used to aligned the sequences with the annotation of B. prionitis as reference. Result of the alignment indicates that the plastomes are extremely conserved, however some level of variations were detected. The pair of the iverted repeat is highly conserved than the small single copy region and large single copy region. Additionally, the proteincoding genes are highly conserved than the non coding region, mostly the integernenic spacer regions. The intergenic spacer regions with high level of variation within the gemone are trnL -trnA, trnH-GUG -psbA, trnC -petN, trnL -trnF, accD -psaI, rps12-trnV, rps15 -ycf1, rps16 -trnQ (Fig. 5). The protein coding genes that showed sequence divergence are ycf2, psbL, atpE, rbcL, petB, petA, and atpF.
The plastome sequence of flowering plant is reported to have generally been conserved [46], although there is a little variations in size and boundries of the single copy and inverted repeats as a results of the evolutionary happenings such as contraction and expansion in the plastome architecture [47,48]. The comparison between the invterted repats and single copy regions boundries in the eight plastome of Acanthaceae (B. prionitis, B. ciliaris, A. paniculata, E. attenuatus, R. breedlovei, J. flava, S.cusia and A. knappiae are presented in (Fig. 6). There is a little variation in the boundaries of the IR-SSC and IR-LSC of the plastomes (Fig. 6),the rps19 is located in LSC region of B. prionitis, B. ciliaris, A. paniculata and A. knappiae.
The following genes trnH, rps19, ycf1 and ndhF are located at the junction of IR-SSC and IR-LSC of J. flava and E. attenuatus plastomes slightly variation in number of nucleotides (Fig. 6). In the SSC/IRb border of the eight plastomes, ycf1 and ndhF genes are found.    (Fig. 7). This shows that the majority of the genes undergo negative selection only few of them were under positive selection. The values of synonymous (dS) rate ranged from 0.01 to 0.38 in all the genes (Fig. 7). Some of the genes including psaJ, atpH, ndhC, psaI, psbE, rpl2, psbH, psbI, psbL and psbF showed no nonsynonymous changes.

Phylogenetic analysis
To determine the phylogenetic relationship and tribal positions of the nine species of Acanthaceae, we used the plastome of the eight species to reconstruct phylogenetic tree. The phylogenetic analyses were performed using Maximum likelihood and Bayesian inference (BI) with Erythranthe lutea, Scrophularia dentate, Lysionatus pauciflorus and Tanaecium tetragonolobum as outgroup. The resulting tree from Bayesian inference (BI) and Maximum likelihood analyses were congruent with high support PP, 1.0 and MP, 100 in all relationships (Fig. 8).
All the nine species clustered in one clade with strong support and are divided into two major sub clades. Sub clade 1 which is monophyletic includes A. knappiae and B. ciliaris (Acantheae) is sister to large clade 2 containing Ruellieae, Barlerieae, Justicieae, Andrographideae. Within the second clade Justicieae and Ruellieae are sister taxa as well as Barlerieae and Andrographideae.

Discussion
In this study, we sequenced the plastome sequence of B. prionitis using Illumina sequencing technology. This is a new approach of obtaining cp genome without prior isolation of the cpDNA and it has been used in several studies. The analysis of the cp genome revealed that the genome has a quadripartite structure; with a pair of inverted repeats regions (IRa and IRb) separated by small single copy region (SSC) and large single copy region (LSC). The organization and structure of the B.  prionitis cp genome is similar to other sequenced Acanthaceae cp genomes [49,50]. Notably, there is high variation in terms of genome size and organization between B.prionitis and S. cusia, this is as a result of IR contraction. The size of the genome 152,217 bp is comparable to other sequenced cp genome of Acanthaceae species, longer than A. paniculata [51], R. breedlovei [50] and S. cusia [52] shorter than E. attenuatus [49]. The size of the genome in all the studied species is relevant to variation in the LSC region. The cp genome of B. prionitis was found to posses 38.3% GC content, as in S.cusia [52]. Additionally, rps12 was recognized as transspliced gene, this was reported in other species [52][53][54].
The arrangement and gene contents of the B. prionitis cp genome is similar to other sequence cp genome of Acanthaceae [50,51] but is different with that of S. cusia which has trnH-GUG in the inverted repeat regions and ycf2 in the large single copy [52]. Some of the genes in the cp genome of B. prionitis start with ACG, GTG and ATC codon, this phenomenon have been reported in angiosperm chloroplast genome [36,37,55]. Repeat elements present in cp genome are correlated with the genome recombination and rearrangements [56,57]. The cp genome of B. prionits is found have low number of repeats compared to sequenced Acanthaceae plastome [47,51,52]. Acanthaceae plastomes contained low repeats compared with other angiosperm cp genome. Most of the repeats were located in the non coding region and ycf genes (ycf1 and ycf2), this has been commonly observed in plastome of angiosperms [58]. Chloroplast microsatellites (cpSSRs) are short repeat in chloroplast genome inherited from a single parent, hence are often used as molecular makers in evolutionary studies such as genetic diversity, they also play role in identification of species [59][60][61]. cp microsatellites analysis, reveal total number of 98 SSRs in the cp genome of B. prionits of which most are mononucleotides, A and T. Poly A and T are reported to be the most abundant repeat in cp genome of plants [62][63][64]. Most of the cpSSRs are located in the non coding region whereas few are located in the protein coding genes region. The microsatellite detected in this study will be useful in evolutionary studies of the genus Barleria as well as identification and conservation of the genus. Variation in size among cp genome is as a result of contraction and expansion of the inverted repeats (IRs) [65]. Contraction and expansion in IRs region were observed in the cp genome B. prionitis and other sequenced Acanthaceae. The size of the inverted repeats ranges from 16, 328 bp in S. cusia to 25, 761 bp in E. attenuatus. Despite the similar lengths of the IR regions of B. prionitis and the other Acanthaceae species with the exception of S. cusia some level of expansion and contraction were observed. There are variation in the border of IR-SC region among the eight species compared, we identified six type of junctions based on the position of rps19, rpl2 and trnH, which occur as a result of contraction and expansion in the inverted repeat region. Type I occurs in three species B. prionitis, A. knappiae and A. paniculata, one of the duplicated rpl2 is located in the LSC region while the other is in the IRb region whereas only 1 rps19 is present in the LSC region. Type II was found in E. attenuatus, here the two rps19 are located in the inverted repeat regions (IRa and IRb) and the rpl22 gene is located in the LSC region. Type III pattern occurs in S. cusia and is characterized by having trnH-GUG duplicated in the inverted region. Type IV has no genes in the IRb/LSC border and was only found in R. breedlovei. In type V which is observed in the genome J. flava, some part of the rps19 gene are located in the inverted repeat region while some are located in single copy region, another remarkable observation is that the two rps19 are of unequal length. The last pattern, type VI occurs in B. ciliaris and is characterized by having rps19 in the LSC region and rpl2 in the IRb region. All the genomes have ndhF in the IRb/SSC border as well as ycf1 in the SSC/IRa border. It is observed that there is extension of inverted repeat into the single copy region in genome of S. cusia which made the LSC region to have length of 93, 666 bp. Despite the conserve nature of the cp genome, some variation could be detected [65]. The positioning of ycf1 gene in IRb, is considered a pseudogene in many flowering plant plastomes. In addition, the stop codon is absent in the ycf1 gene sequence and this result to the differences in the distribution of genes in single copy and inverted repeat borders. The result of the comparative genome analysis using mVISTA revealed that the genome is relatively conserved with some degree of variation, which mostly occurs in the non coding region as a result of insertion and deletion. The results of the alignment showed no considerable structural rearrangements, like gene relocation or inversion were detected in the plastomes. The structural rearrangement was detected in the cp genome of S. cusia. DNA barcodes are sequences in the genome unique to particular taxa and are used as reliable tools for identification of plants and resolving phylogenetic relationship [65,66]. The alignment of the eight cp genome reveals variable regions which includes trnH-GUG -psbA, rps16 -trnQ, trnC -petN, accD -psaI, clpP intron, trnL -trnF, rps15 -ycf1, rps12-trnV, trnL -trnA, atpE, atpF, rbcL. These regions will be used as makers for identification of the sampled Acanthaceae species as well as resolving phylogetic relationships in the family. Most of the variable regions are located in the single copy region particularly the large single copy, this is consistent in most angiosperms.
Synonymous (dS) and non synonymous (dN) substitution rate as well as dN/dS ration were calculated to evaluate sequence divergence and purifying selection in the protein coding genes. The result indicates low sequence divergence in most of the genes (dS < 0.1). The dN/dS analyses show that most of the protein coding genes were under negative selection, only few genes (atpF, petD, psbZZ, rpl20, petB, rpl16, rps16, rpoC, rps7, rpl32 and ycf3) were under positive selection (dN/dS > 1), comparable findings were reported for other plastomes [66][67][68].
Complete chloroplast genome is a good resource for inferring evolutionary and phylogenetic relationships [69][70][71]. Many researchers have used the plastome sequence to resolve phylogenetic relationships at various taxonomic levels [72,73]. Until this study, the phylogenetic relationships and tribal classification of Acanthaceae The top arrow shows transcription direction, blue colour indicatesprotein coding, pink colour shows conserved non coding sequence CNS and light green indicates tRNAs and rRNAs. The x-axis represents the coordinates in the cp genome while y-axis represents percentage identity within 50-100% was evaluated using only few genes and the tribal classification is still required to be clarified. In this study, we used the cp genome of nine species representing the four major tribes of the Acanthoideae and reconstructed phylogenetic relationships based on maximum parsimony and Bayesian inference methods. The resulting phylogenetic tree from the two methods showed the same topology with high resolution values at the clades. The result of this study based on nine Acanthaceae taxa confirm that Acanthoideae (the retinaculate clade) are monophyletic and also confirm the sister relationship between Acantheae (non cystolith clade) and the cystolith clade, this has been reported earlier [11][12][13]19] . The phylogenetic tree showed Justicieae and Ruelliae are sister taxa as reported previously [19] therefore should be regarded as separate tribes not as Justicieae or Ruelliae because the species within these two taxa are paraphyletic. The sister relationship between Andrographideae and Barlerieae is also confirm. Andrographideae and Barlerieae were placed in the tribe Justiceae as sub tribes [35,74]. Recently Scotland and Vollesen classified all species with cystolith under the tribe Ruellieae placing Andrographis, Barleria and Justicia under the sub tribes Andrographinae, Barleriinae and Justiciinae respectively. Our findings suggested that Andrographideae, Justicieae and Barlerieae should be treated as tribes not sub tribes.

Conclusion
In this study, we sequenced and reported the complete chloroplast genome of B. prionitis, providing valuable plastome genomic resources for the species. The plastome of B. prionitis has a typical gymonosperm cp genome structure and is comparable to other cp genome of Acanthaceae. Simple sequence repeats that will be used for evolutionary studies within Barleria were identified. The genome comparative analyses of 9 Acanthaceae reveal variable hotspot that could be used to develop DNA barcode for the identification of the species. These hotspots will also be useful in phylogenetic relationship   studies of the family Acanthaceae. The study also reveals that only few genes were under positive selection. The findings of the confirmed the tribal position of major genera within Acanthoideae and suggested that Andrographideae, Justicieae and Barlerieae should be treated as tribes not sub tribes.

Plant material and DNA extraction
Plant material was collected from Makkah Taif road, Saudi Arabia (39 0 20′ 0.30″E, 21 0 45′ 33.68″N) and identified by the curator of King Abdulaziz University Herbarium, Dr. Dhafer A. Alzahrani, the voucher specimen was deposited in the herbarium of King Abdulaziz University, Jeddah, Saudi Arabia, with voucher specimen number KAU22534. Total genomic DNA was extracted from leaves using Qiagen DNA extraction Kit according to manufacturer's protocol.

Gene annotation
Dual Organellar GenoMe Annotator (DOGMA) [77] was used to annotate the genes in plastome followed by manual adjustment of the positions of start and stop codons. TrNAscan-SE2.0 [78] was used to verify tRNA genes. Organellar Genome Draw (ORGDRAW) [79] was used to circular map of plastome. The complete chloroplast genome sequence of B. prionitis was submitted to GenBank (Accession number MK548575).

Sequence analysis
Relative synonymous codon usage values (RSCU), base composition and codon usage were analyzed using MEGA 6.0. PREP suite [80] with cutoff value of 8.0 was used to predict the RNA editing sites in the plastome.

Genome comparison
mVISTA [82] was used to compare the plastome using the annotation of B. prionitis as reference in the Shuffle-LAGAN mode [83].

Characterization of substitution rate
To detect the genes that were under selective pressure, DNAsp v5.10.01 [84] was used to analyze the synonymous (dS), nonsynonymous (dN) and dN/dS value of all the protein coding genes in sampled Acanthoideae species.

Phylogenetic analysis
For phylogenomic analysis, the cp genomes of Acanthoideae species deposited in the GenBank were recovered ( Table 8). The plastome of four species of the order lamiales were also downloaded and set as out groups ( Table 8). The downloaded sequences and cp genome of B. prionitis were aligned with MAFFT v.7 [85] and analyzed using Maximum parsimony with (PAUP version 4.0b10) [86] and Bayesian Inference with MrBayes version 3.2.6 [87].. To select the suitable model for Bayesian analysis jModelTest 3.7 [88] was used. Authors' contributions SSY, DAA and AA collected the data, designed and performed the experiment, SSY and EJA analyzed the data and drafted the manuscript, DAA supervised the project, all the authors edited and approved the manuscript.