The Chloroplast Genome of the Endemic Species Abrus bottae Deflers: Comparative and Phylogenetic Analysis with Closely Related Species of Abreae Hutch

Abstract

Abrus bottae belongs to the subfamily Papilionoideae DC. and the family Fabaceae Lind., endemic to the Arabian Peninsula. This genus encounters numerous taxonomic issues concerning both the quantity of species within the genus and the systematic relationships among its species. Notably, there is a complete absence of sequence data in the GenBank database for this species. A molecular and phylogenetic study of the chloroplast genome of the species A. bottae was performed in this work. The chloroplast genome is 152,540 bp in size and exhibits a typical quadripartite structure, consisting of a substantial single-copy region of 83,507 bp, a small single-copy region of 17,681 bp, and a pair of inverted repeat regions of 25,676 bp each. The chloroplast genome of Abrus bottae encompasses 130 genes. An analysis of nucleoside diversity revealed 26 nucleotide polymorphism sites with Pi values (a measure of genetic variation within species) ≥ 0.04, serving as hotspots of variation. This work represents the first molecular phylogenetic study on the endemic species Abrus bottae and presents a comparative and phylogenetic analysis of the cp genomes of related taxa within the tribe Abreae. These outcomes can be used to develop DNA barcodes to study variation among the Abrus species.

1. Introduction

Fabaceae is the third-largest angiosperm family, comprising 770 genera with more than 19,500 species [1]. Members of this family are used as food and forage crops, and they are also a model for biodiversity and genetic evolution. This family is classified into six subfamilies on the basis of the plastid gene matK: Cercidoideae LPWG., Detarioideae Burmeist., Duparquetioideae LPWG., Dialioideae LPWG., Caesalpinioideae DC., and Faboideae Rudd (syn. Papilionoideae DC.) [2]. Papilionoideae represents the most extensive subfamily of Fabaceae, with 28 tribes, 478 genera, and approximately 13,800 species [3] characterized by high species diversity. Furthermore, Papilionoideae consists of seven principal clades: Cladrastis, Genistoids, Dalbergioids, Mirbelioids Benth, Millettioids, Robinioids, and the inverted-repeat-lacking clade (IRLC).

Abrus Adanson is classified under the Millettioid/Phaseoloid clade. The genus belongs to the tribe Abreae of the subfamily Papilionaceae [4]. The term Abrus is a species derived from “habro-” (Greek), meaning delicacy, elegance, beauty, and softness. The taxonomic meaning of Abrus Adanson is linked to a finely textured leaflet and its remarkable flower. Plants of this genus exhibit various growth forms, such as climbers, lianas, or herbs, and are characterized by laciniate pinnately compound leaves. Nine stamens of this herb are connate [4]. Abrus Adans. is the only genus in the tribe Abreae. The genus has 17 species [5] distributed within tropical and subtropical regions on Asian and African continents [6].

The taxonomic status and tribe affinity of the genus Abrus have long been uncertain [4]. According to Hu et al. [7], Abrus has a close relationship with Millettieae s.l., being defined as the core Millettieae on the basis of a pseudoraceme inflorescence and the absence of canavanine (except in the Philonoptera clade) and a base chromosome number of x = 11 [8]. Recent phylogenetic investigations of the chloroplast genome have shown that the genus Abrus and tribe Abreae are members of the Millettioid clade [9,10].

There is an ongoing discussion over genus classification, as there is contradictory information on how many species it includes. The first detailed revisional work on this genus was by Breteler [11], who reported four species in the genus. In the later revision by Breteler [12], he concluded that the observed variation is not adequately documented and depicted, detracting from its overall credibility. He presumed that this complex, which includes a range of species, is morphologically challenging enough to keep it disputed [8].

Unlike Breteler, Verdout reported in great detail the variation observed and later suggested several subspecies to include the specimens he investigated. However, his definition of morphological characteristics, often foliar, is occasionally not cohesively and consistently coupled with geographical demarcation [8]. Given the broad diversity of species and interspecies relationships, the ability of these species to populate many types of ecosystems, from tropical rainforests to savannah and even semiarid environments, becomes evident. Some morphological characteristics are highly plastic and therefore hamper the identification of species on the basis of vegetative parts such as leaves and stems [13]. Thus, a wide difference in the rank of a particular plant group may present great difficulty for scientists studying this group of flora [8].

This finding points to some concerns regarding the status of these species and the fact that many of these species have been used in traditional medicine regimens. For example, A. pulchellus subsp. Cantoniensis and A. pulchellus subsp. Mollis (Hance) Verdc. have medicinal qualities and are often utilized as Abri Herba in different practical contexts. These two subspecies are morphologically identical. A recent comparative analysis of plastomes performed by Zhang et al. [14] revealed that A. pulchellus subsp. pistaceae shows large variations in plastid genomes.

Abrus schimperi Hochst. Ex Baker is commonly used traditionally as a root decoction for the treatment of pneumonia. This species harbours compounds with antiparasitic, antileishmanial, antiplasmodial, antimicrobial, and cytotoxic activities [15]. Abrus Cantoniensis Hance (ACH, also known as Abrus melanospermus subsp. Melanospermus) is known for its ability to relieve heat stress, detoxification effect, liver protection effect, and analgesic activity [16].

Abrus precatorius L. has been used for the topical treatment of abrasions, sores, and wounds caused by domestic animals such as dogs, cats, and rodents and is commonly combined with other drugs for the treatment of leukoderma, fever, cough and cold, bronchitis, and hepatitis. Seeds are employed in the treatment of diverse health ailments, such as headache, snakebite, blennorrhagia, carbuncle, cancer, cold, colic, conjunctivitis, convulsions, cough, diarrhoea, fever, gastritis, gonorrhoea, malaria, night blindness, ophthalmia, rheumatism, diabetes, and chronic nephritis [17].

Abrus bottae Deflers (Figure 1) is a grazed leafy shrublet that is 60 cm high. The leaf is a pinnately compound consisting of pairs of leaflets that are pale green. The flower is purple or lilac and is 1 cm long and has no scent. The fruit manifests as pods that are typically hairy. These pods are generally denser and more lignified, with an apex that may be acute or rounded and long brown seeds common in the Fayfa Mountains, Kingdom of Saudi Arabia, where the shrubs grow to 2 m tall and some of the stems are self-twining [18,19]. The native range of this species spans the Arabian Peninsula, particularly Saudi Arabia and Yemen [20]. Recently, Abrus bottae has been recorded on the IUCN Red List of endangered plants within the Al Dhale’ Governorate of Yemen [21]. Notably, there is a significant lack of morphological, chemical, and molecular phylogenetic information related to this species, and there are no sequences in the gene bank for this species.

On the other hand, the whole-chloroplast genome provides necessary data for the classification of plants, phylogenetic research, and historical biogeographic studies [22]. This study is the first to provide complete chloroplast genome data for Abrus bottae and evaluate its taxonomic status. In addition, it aims to contribute to the comparison of A. bottae species with other Abrean tribe members (Table S1, Figure S1) by assessing any differences in base composition, revealing the internal evolutionary genomic architecture and the levels of genetic diversity.

2. Materials and Methods

2.1. Sample Collection

The leaves of Abrus bottae were collected from Jabal Fayfa, which is located at approximately 17°14′45.1″ N and 43°05′27.2″ E in the Jazan Province of the southwest region of the Kingdom of Saudi Arabia (Figure S2). This part of the Kingdom forms an important centre for flowering plants, and the diverse floras harboured in different habitats make the SG a hotspot of biodiversity in the Arabian Peninsula [23]. This ecological peculiarity is associated mainly with high annual precipitation (>300 mm) and an elevated altitudinal gradient (ranging from sea level to 3100 m). The Abrus bottae material was collected in September 2024 and taxonomically identified with herbarium material and the relevant literature [18,19]. The selected samples were properly processed and stored in the botanical collection of Umm Al-Qura University located in the Makkah Region of Saudi Arabia. For DNA extraction and yield, fresh leaf samples were desiccated with silica gel.

2.2. Extraction of DNA

Deoxyribonucleic acid was extracted from the desiccated foliage of Abrus bottae using the CTAB DNA extraction method as described previously [24].

2.3. Genomic Library Development

One gram of DNA served as the initial substrate for the preparation of the samples. The DNA library was prepared according to the manufacturer’s protocol in accordance with the instructions provided in the Illumina TruSeq Nano DNA Kit (Illumina, San Diego, CA, USA). The template fragment size distribution was assessed using an Agilent Technologies 2100 Bioanalyzer (Agilent, Santa Clara, CA, USA), and library quantification was performed using qPCR.

2.4. Genome Sequencing

The constructed templates were sequenced on an Illumina sequencer using standard SBS technology. The raw sequenced data were analysed using Trimmomatic v0.38 [25] (http://www.usadellab.org/cms/?page=trimmomatic: accessed on 7 June 2025), comprising the excision of adapter sequences and the filtration of reads predicated on a quality cut-off (Q30 at 91%), and 12 Gb of filtered reads were obtained. Macrogen (https://dna.macrogen.com/: accessed on 20 April 2025, Seoul, Republic of Korea) was used to perform library preparation and sequencing.

2.5. Genome Sequence and Annotation

Initially, the quality of the raw reads was evaluated using the FastQC software tool v0.12.0. The subsequent filtering of these reads allowed NOVOPlasty version 4.3.1 [26] (https://github.com/ndierckx/NOVOPlasty: accessed on 7 June 2025), with a k-mer value of 33, to perform an assembly based on the whole-genome sequence of Abrus bottae to reconstruct the entirety of the chloroplast genome of this species. Published sequences of Abrus pulchellus subsp. Mollis (GenBank accession number: LC708259. 1), Abrus precatorius (MN709888. 1), and Abrus pulchellus subsp. Cantoniensis (MT328396. 1) were used as references in the assembly process. The resulting assemblies formed complete chloroplast genomes as single contigs. The chloroplast genome of A. bottae was predicted and annotated using the GeSeq pipeline [27] (https://chlorobox.mpimp-golm.mpg.de/geseq.html: accessed on 7 June 2025) (default, but with thresholds of 60 and 85 for protein-coding genes and RNAs, respectively). tRNA genes were identified using tRNAscan-SE version 2.0 [28] (https://bio.tools/trnascan-se: accessed on 7 June 2025). Circular gene maps of the chloroplast genomes were created with OGDRAW (Organellar Genome DRAW), version 1.3.1 [29] (https://chlorobox.mpimp-golm.mpg.de/OGDraw.html: accessed on 7 June 2025), using the gb format of the sequence files containing annotations. The genomic sequence of the chloroplast from Abrus bottae was archived in GenBank with the accession identifier PV582056.

2.6. Sequence Examination

Analysis of relative synonymous codon usage (RSCU) parameters, including nucleotide composition and codon employment, was performed with Graphical Codon Usage Analyser Version 1.2 (GCUA) [30] (https://gcua.schoedl.de/: accessed on 7 June 2025). Sites for RNA editing in protein-coding sequences were predicted using the PREPACT3 suite [31] (http://www.prepact.de/prepact-main.php: accessed on 7 June 2025), with a minimum requirement threshold of 0.8.

2.7. Examination of Repetitive Sequences in the Chloroplast Genome

MIcroSAtellite (MISA) version 2.0 [32] (https://webblast.ipk-gatersleben.de/misa/: accessed on 7 June 2025) was employed to detect simple-sequence repeats (SSRs) within the chloroplast genome of Abrus bottae as well as three additional species belonging to the tribe Abreae, i.e., A. pulchellus subsp. Mollis., Abrus precatorius, and Abrus pulchellus subsp. Cantoniensis. The numbers of copies in the following SSR motifs were as follows: 8 repeat units for mononucleotide repeats, 5 for dinucleotides, and 4 each for trinucleotides, tetranucleotides, pentanucleotides, and hexanucleotides. The program REPuter [33] (https://bio.tools/REPuter: accessed on 7 June 2025) was executed utilizing the standard parameters to search for palindromic, forward, reverse, and complement repeats in the plastid genome of Abrus bottae as well as three Abreae lineage species.

2.8. Comparative Chloroplast Genomics

The chloroplast genome of Abrus bottae was aligned and compared with those of three representative species of the Abreae tribe, including Abrus pulchellus subsp. Mollis, Abrus precatorius, and Abrus pulchellus subsp. Cantoniensis. All the genomic sequences, except for those of Abrus bottae, were retrieved from the NCBI database. This was noted in the annotation of Abrus bottae. For reference-guided comparative analysis, the annotated genomic sequence of Abrus bottae was used with the Shuffle-LAGAN in the mVISTA [34] web-based interface [35] (https://genome.lbl.gov/vista/mvista/submit.shtml: accessed on 7 June 2025) to obtain further information. Moreover, the IR boundaries and junctions of the SC regions were analysed with the IRSCOPE [36] (https://irscope.shinyapps.io/irapp/: accessed on 7 June 2025).

2.9. Nucleoside Diversity Statistics

Nucleotide diversity (Pi) was calculated from a consensus sequence alignment. Variability of nucleoside was estimated by sliding window analysis implemented in DnaSP v6.11.01, with a window size of 600 bp and a step size of 200 bp [37] (http://www.ub.edu/dnasp/: accessed on 7 June 2025). The boundaries of the inverted repeats (IRs) and the large single-copy (LSC) and small single-copy (SSC) regions were deduced from GBseq annotations, and the IRSCOPE [36] (https://irscope.shinyapps.io/irapp/: accessed on 7 June 2025) was used for chloroplast genome comparison within the Abreae taxa, with Abrus bottae as the reference species.

2.10. Characterization of the Substitution Rate

An estimation of nonsynonymous (Ka) and synonymous (Ks) substitutions as well as the Ka/Ks ratio was performed to determine differences in the mutation rates of chloroplast genomic sequences and to elucidate the phylogenetic history [38] of Abrus bottae compared with those of other members of Abreae. The Ka/Ks Calculator version 2.0 [38] (https://github.com/kullrich/kakscalculator2: accessed on 7 June 2025) with its default settings and the Nei and Gojobori substitution model were used.

2.11. Phylogenetic Analysis

To explore the phylogenetic status of Abrus bottae in Fabaceae, a MrBayes phylogenetic tree was generated on the basis of the whole-cp-genome sequences from 46 species, including 19 tribes from the subfamily Papilionoideae, and the sequence of Polygala tenuifolia was set as the outgroup. It identified the homologous genes shared by these selected taxa, and the number of single-copy orthologous genes was also calculated using OrthoFinder [39,40] and then concatenated with the full CDS. Sequence alignment was performed using MAFFT v7. 525 [41] (https://mafft.cbrc.jp/alignment/software/source.html: accessed on 7 June 2025). The alignment file created by MAFFT was then used to conduct Bayesian inference analysis (BI) in MrBayes v3.2.6 [42] (https://nbisweden.github.io/MrBayes/: accessed on 7 June 2025), and the resulting BI tree was displayed using the ITOL tool [43] (https://itol.embl.de/upload.cgi: accessed on 7 June 2025).

3. Results

3.1. Characteristics of the Chloroplast Genome of Abrus Bottae

The complete chloroplast genome of A. bottae, with a total length of 152540 bp (Figure 2), is circular. This genome is separated into four regions encompassing an LSC (large single-copy) region, an SSC (small single-copy) region, and two IRs (inverted repeats). The lengths of the LSC and SSC regions are 83,507 bp and 17,681 bp, respectively, whereas the IRa and IRb regions are both 25,676 bp in size (

Table 1. Genome features of the chloroplast genome of Abrus bottae.

Characteristics	Number
Genome size (bp)	152,540
IRA (bp)	25,676
IRB (bp)	25,676
LSC (bp)	83,507
SSC (bp)	17,681
Total genes	130
Total unique genes	111
rRNA	4
tRNA	28
Protein-coding genes	79
A%	30.73
T (U) %	31.54
G%	18.56
C%	19.17
GC%	37.72

).

The coding region is composed of 73,194 base pairs (bps), which cover 47.98% of the entire genome, and the noncoding region consists of 67,920 bp (4.53%). Adenine‒thymine (AT) pairings account for 62.64% of the genome, and guanine‒cytosine (GC) pairs account for 37.72% of the genome. The complete genomic composition is shown in Table 1 (A = 30.73%, T(U) = 31.54%, C = 19.17%, G = 18.56% for A. bottae). Moreover, functional annotation indicated that the gene classification of the A. bottae chloroplast genome is almost the same as that of other plant species (Supplementary Table S2).

A total of 130 different genes, consisting of 111 unique genes and 19 duplicated genes in the IR region, were recognized by an integrated study. The chloroplast genome contains 79 protein-coding genes, 28 tRNA genes, and 4 rRNA genes (Table 1). In particular, the LSC region consists of 62 protein-coding genes and 20 tRNA genes, and the SSC region contains 9 protein-coding genes and 1 tRNA gene. Moreover, 8 protein-coding and 10 tRNA genes are found within the IR region. The majority of protein-coding genes are initiated by the AUG (methionine) codon.

A total of 19 Abrus bottae genes contained introns, with 13 encoding protein-coding genes and 6 tRNA-encoding genes. The genes of ycf3 and clpP have two introns, and the remaining genes have only one (Table S3). The LSC regions include a total of 10 introns. Specifically, eight introns are located in the IR region, and one intron is located in the SSC region. In addition, the trnK-UUU gene is also distinguished by the longest intron of 2499 bp.

3.2. Relative Synonymous Codon Usage (RSCU)

The distribution of the 20 amino acid and termination codons among all confirmed protein-coding genes in the chloroplast of Abrus bottae is shown in Figure 3. The evaluation of codon usage bias in the plastome was performed by analysing nucleotide sequences in protein-coding and tRNA genes. A comprehensive analysis of protein-coding and tRNA genes (84,620 bp) of Abrus bottae revealed 24,398 different codons. Leucine was the most frequent amino acid, with a proportion of 10.59%, and cysteine was the least frequent, with a proportion of 1.11%. Arginine, leucine, and serine codons were over-represented (RSCU ≥ 6), whereas tryptophan and methionine codons presented low representation with no codon usage bias (RSCU value equal to 1), as shown in Figure 3. As shown in Table S4, on the basis of the RSCU values, 28 codons had an RSCU value greater than 1, 24 codons had A/T-ending codons, and 4 had C/G-ending codons. As shown in the table, tryptophan and methionine, both of which do not have codon usage bias, have RSCU values of 1 (Table S4).

3.3. RNA Editing Sites

The RNA editing sites in A. bottae are summarized in Table S5. In the cp genome, 67 putative RNA editing sites were predicted. Most changes in the position (22 sites) of the codon were associated with the amino acids from proline (P) to serine (S), and 19 sites changed from histidine (H) and tyrosine (Y) (from H to Y). In addition, other editing sites were related to the following changes in the amino acids: phenylalanine (F) at eight sites, isoleucine (I) at seven sites, leucine (L) at six sites, and valine (V) at three sites.

The results revealed the greatest number of editing sites in the rpoC2 gene, with six editing sites, followed by five for ycf2 and ndhF each. The rpoA gene has three editing sites and two editing sites for matK, rps2, rpoC, psaA, accD, rps18, rps8, rpl14, ndhB, ycf1, ccsA, ndhG ndhI, ndhH, and rps15. The remaining genes had one editing site. Analysis of RNA editing revealed that some genes, such as atpF, rps2, psbZ, psaA, petL, rps18, rps18, ndhG, ndhI, and rps15, do not have an anticipated locus at the initial codon of the foremost nucleotide.

3.4. Repeat Analysis

3.4.1. Prolonged Repetitions

Most of the long repeat sizes were under 30 bp in all the Abrus spps. (Figure S3a). In A. bottae (91.84%), the number of long repeats was <30, followed by 30–60 bp (8.16%).

The total number of long repeats reached 49 in A. bottae, A. pulchellus subsp. Mollis, and A. precatorius each, whereas 47 long repeats were noted in A. pulchellus subsp. Cantoni. The Abrus bottae chloroplast genome contains four different types of repeats, as outlined in Table S6, including palindromic (24), forward (16), reverse (7), and complement (2) repeats. Abrus pulchellus subsp. Mollis (Figure S3b) presented the highest frequency of palindromic repeats (28), whereas Abrus precatorius presented the highest frequency of forward and reverse repeats (19 and 8). Abrus bottae presented the greatest number of complement repeats (2). Notably, complement repeats were absent in Abrus pulchellus subsp. Mollis (Figure S3b). The protein-coding gene ycf2 is clearly represented, with six palindromic repeats and five forward repeats, resulting in the greatest number of repeat occurrences (Table S6).

3.4.2. Simple-Sequence Repeats (SSRs)

SSRs are extensively present within the Abrus spp. chloroplast genomes. Approximately 60% of microsatellite repeats occur in intergenic spacer regions (IGSs), and 25% occur in coding regions. The total numbers of SSRs detected in Abreae spp. are 335, 323, 341, and 344 in A. bottae, A. pulchellus subsp. Mollis, A. pulchellus subsp. Cantoniensis, and A. precatorius, respectively (

Table 2. cpSSRs in the chloroplast genomes of the Abreae tribe.

SSR Type	Repeat Unit	Abrus bottae	Abrus pulchellus subsp. Mollis	Abrus pulchellus subsp. Cantoniensis	Abrus precatorius
Mono	A	116	113	130	128
	C	12	11	8	7
	G	12	8	8	8
	T	146	130	131	131
Di	AG/CT	10	1	2	2
	AC/GT	2	1	1	1
	AT/AT	25	40	42	46
Tri	AAG/CTT	1	2	1	1
	AAT/ATT	0	4	5	7
Tetra	AAAG/CTTT	1	0	1	0
	AAAT/ATTT	4	4	2	3
	ACAG/CTGT	1	0	0	0
	AATT/AATT	2	0	1	1
	AATC/ATTG	1	0	0	0
	AAAC/GTTT	1	2	1	1
	AGAT/ATCT	1	3	4	4
	AATG/ATTC	0	2	2	2
Penta	AATAG/ATTCT	0	1	0	0
	AATAT/ATATT	0	1	0	0
Hexa	ACAGAT/ATCTGT	0	0	2	2

).

Single-nucleotide SSRs were the most common (286, 262, 277, and 274) in A. bottae, A. pulchellus subsp. Mollis, A. pulchellus subsp. Cantoniensis, and A. precatorius, respectively, accounting for 85.37%, 81.11%, 81.23%, and 79.65%, respectively. Of note, the mononucleotide SSR markers were also mostly A/T repeats. In the A repeat range of 113–130, T repeats the most in all the spp. of Abreus in the range of 130–146, in the G range of 8–12, and in the C range of 7–12 (Table 2).

Three types of dinucleotide SSRs were present in all the Abrus species (AG/CT, AC/GT, AT/AT). Trinucleotide SSRs (AAG/CTT) were present in all the species, whereas AAT/ATT was absent in Abrus bottae.

Tetranucleotide SSRs (AAAG/CTTT) are present in Abrus bottae and Abrus pulchellus subsp. Cantoniensis. In contrast, tetranucleotide SSRs (AAAT/ATTT, AAAC/GTTT, AGAT/ATCT) were present in all the Abrus species in this study. ACAG/CTGT (AATC/ATTG) repeats are observed only in Abrus bottae. However, AATG/ATTC SSRs are absent in Abrus bottae. The tetranucleotide SSR (AATT/AATT) is absent from Abrus pulchellus subsp. Mollis.

Pentanucleotide SSRs (AATAG/ATTCT, AATAT/ATATT) are only present in Abrus pulchellus subsp. Mollis. Hexanucleotide SSRs (ACAGAT/ATCTGT) are only present in A. pulchellus subsp. Cantoniensis and A. precatorius species (Table 2 and Figure S4).

The frequency of SSRs present in the dispersed SSR regions was determined for Abrus bottae and three other cp genomes of the Abreae tribe, as shown in Figure 4. The majority of simple-sequence repeats (SSRs) were predominantly found within the LSC region, followed by the SSC region, and relatively few SSRs were located in the IR regions. With respect to the cp genome of Abrus bottae, the majority of SSRs (65.5%) in the plastome were located within the LSC region, followed by 19.4% in the SSC region, and a rare number (8.1% each) were located in the IRA and IRB regions.

3.5. Comparative Genomes

To assess the degree of expansion of divergence among the genomic sequences, the mVISTA framework was employed to evaluate the congruity of the Abrus bottae sequences with three chloroplast genomes of the tribe Abreae currently available with GenBank, namely, Abrus precatorius, Abrus pulchellus subsp. Cantoniensis, and Abrus pulchellus subsp. Mollis. In summary, protein-coding genes presented greater conservation than did noncoding genes (Figure 5).

The noncoding regions presented high sequence divergence at certain sites: rps12-trnN-GUU, trnN-GUU-trnR-ACG, rrn16s-trnV-GAC, trnV-GAC-rps12, ndhB-trnLCAA, trnI-CAU-trnHGUG, psbA-trnK-UUU, trnK-UUU-rps16, trps16-trnQ-UUG, trnS-GCU-trnG-UCC, trnG-UCC-trnR-UCU, atpF-atpH, atpH-atpI, atpI-rps2, rpoC1-rpoB, rpoB-trnC-GCA, petN-psbM, psbM, trnD-GUC, trnD-GUC-trnY-GUA, trnY-GUC-trnT-GGU, trnT-GGU-psbD, psbZ-trnG-GCC, psaA-ycf3, trnT-UGU-trnL-UAA, trnM-CAU, atpB-rbcL, petA-pspL, psbL-petL, rps12-clpP, clpP-psbB, pspH-petB, petB-petD, rpI16-rps19, rpI2-trnI-CAU, Ycf15-trnL-CAA, trnL-CAA-ndhB, ndhB-rps7, rps12-trnv-gac, trnR-ACG-trnN-GUU, trnN-GUU-ycf1, ndhF-trnL-UAG, ndhD-ndhE, ndhE-ndhI, and ndhA-ndhH.

Only the divergence of protein-coding genes occurred in a lesser number of loci: rps12, matK, psbK, psbI, atpH, atpI-rps2, rpoC2, rpoC1, petN, rps4, ndhJ, ndhK, ndhC, accD, psaI, ycf4, cemA, petA, psbL, psaJ, rpI33, psbH, rpoA, rps11, rpI36, intA, rps8, rpI14, rpI22, Ycf2, Ycf1, ccsA, ndhD, ndhE, ndhI, ndhH, and rps15.

3.6. LSC/SSC and IR Boundaries

The comparison of the LSC, SSC, and IR boundaries involves comparing the chloroplast genomes of Abrus pulchellus subsp. Mollis, Abrus pulchellus subsp. Cantoniensis, and Abrus precatorius of the tribe Abreae, with Abrus bottae as a reference (Figure 6). Differences were observed in the length of the single-copy regions of the species, leading to differences in the loci of genes near the delimitations of the LSC, SSC, and IR regions, such as the rps19, ndhF, ycf1, and trnH genes. The LSC region was 83,507–86,633 bp long in Abrus pulchellus subsp. Mollis, and this region was the shortest in Abrus bottae. The length of the SSC region ranged from 17,681 to 18,425, with the smallest SSC region identified in Abrus bottae and the longest in Abrus pulchellus subsp. Cantoniensis. The IR region remained relatively constant, with a length of 25,676–25,796 bp.

The rps19 gene is located on the LSC-IRB border in both A. pulchellus subsp. Mollis and A. precatorius, while it extends inside the LSC region in Abrus bottae (188 bp) and A. pulchellus subsp. Cantoniensis (2276 bp). The ndhF gene is located at the IRB–SSC junction in both Abrus bottae and Abrus pulchellus subsp. Mollis. The gene is located in the SSC region (30 bp) in Abrus pulchellus subsp. Cantoniensis and Abrus precatorius.

The ycf1 gene is located on the SSC–IRA border in the following species: A. bottae, A. pulchellus subsp. Mollis, and A. precatorius. However, the ycf1 gene was more expansion in the SSC region (1066 bp) of A. bottae. In contrast, the ycf1 gene extends inside the SSC region (463 bp) in both A. pulchellus subsp. Cantoniensis and A. precatorius. Of note, the ycf1 gene is not present in the SSC-IRA junction of Abrus pulchellus subsp. Molli.

The trnH gene is located much closer to the IRA–LSC borders in all Abrus species, with distances of 11 and 53 bp that flank the border in Abrus bottae and Abrus pulchellus subsp. Mollis. This gene is located 60 bp away from the IRA–LSC border in Abrus pulchellus subsp. Cantoniensis and 58 bp away from the border in Abrus precatorius (Figure 6). Taken together, regarding the IR shrinkage/expansion events among these four Abrus species, the general pattern was consistent in terms of gene content at the junctions (rps19, ndhF, ycf1, and trnH), but some differences in gene positions at the junctions were detected.

3.7. Nucleoside Variability

The nucleotide diversity results showed that inverted repeats are more conserved compared to single-copy regions (Figure 7). The graph represents approximately 26 points of polymorphism that are mainly concentrated in the LSC regions and have Pi values ≥ 0.04. Therefore, these regions appear to represent polymorphism hotspots, particularly at the “atpE, atpF, atpH, atpI, ndhC, ndhK, psaB, psbC, psbM, psbZ, rpoB, rpoC1, rpoC2, rps14, rps16, rps2, rps4, ycf3, trnD-GUC, trnF-GAA, trnG-GCC, trnG-UCC, trnL-UAA, trnS-GGA, trnT-GGU, and trnT-UGU” regions. In total, the level of polymorphic DNA varies with Pi ranging from 0.41 to 0.53 (average Pi = 0.46, Table S7 for the genes with Pi ≥ 0.04).

3.8. Analysis of the Rate of Substitution

The genetic distances for the synonymous (Ks) and nonsynonymous (Ka) substitutions and the Ka/Ks ratios were estimated for 79 protein-coding genes representing the genes common to the studied Abrus species (Abrus bottae, Abrus precatorius, Abrus pulchellus subsp. Cantoniensis, and Abrus pulchellus subsp. Mollis). Some genes showed evidence of positive selection at a Ka/Ks ratio > 1 (Figure S5). These genes include rpl20, rpl23, rps16, ycf2, and ycf4. In contrast, the majority of Ks values were less than 1 among all the genes (Figure S5).

3.9. Phylogenetic Relationships

The phylogenetic tree of Papilionoideae (Figure 8) created on the basis of the chloroplast genomic sequences revealed that almost all nodes obtained Bayesian posterior probabilities (PP = 1). This finding indicates that almost all nodes were highly supported clades, as demonstrated in this investigation. In the present work, the largest and most diverse clade (nonprotein amino acid-accumulating clade [NPAAA]) included both the Millettioid clade and the Hologalegina clade. The tribe Abreae is nested within the Indigofereae + Millettioid clade. The Indigofereae and Millettioid clades are composed of several tribes, including Dicocleae, Desmodieae, Indigofereae, and Psoraleeae, with well-supported nodes (PP = 1). The Abreae genus Abrus appears as a monophyletic clade (PP = 1) with four species, all highly supported (PP = 1). A. pulchellus subsp. Cantoniensis is nested as a sister to A. precatorius, and the two cluster together with A. pulchellus subsp. Mollis and Abrus bottae, with maximum support values (PP = 1). These genera are phylogenetically close to C. cathartica, composing a subclade highly supported (PP = 1), indicating that Canavalia (Phaseoleae) is a sister to Abrus (Abreae).

A strong phylogenetic relationship is recognized for Robinioids (Sesbanieae, Loteae, Robinieae) and the IRLCs (Wisterieae, Cicereae, Fabeae, Hedysareae, and Galegeae) within the Hologalegina group; most of these clades are strongly supported (PP = 1). The Robinidae clade is well supported (PP = 0.5), except for some subclades. In the latter case, the Dalbergioid clade (Amorpheae and Dalbergieae) is a sister to the NPAAA clade, and the generalist clade, which includes Borngniartieae, Podalyieae, and Crotalarieae, is located at the base of the tree.

4. Discussion

The chloroplast genome of A. bottae is 152,540 bp in length and has a four-region structure, including a large single-copy (LSC) region that is 83,508 bp in length, small single-copy (SSC) region 17,681 bp in length, and two inverted repeat (IRa and IRb) regions 25,676 bp in length. This finding is consistent with the average genome length in the subfamily Papilionoideae of 140–160 kb [44] and with the average genome length in angiosperms of 107–218 kb [45]. The chloroplast genome of angiosperms has a conserved quadripartite arrangement characterized by two inverted repeats (IRs) interspersed between a large single-copy region (LSC) and a small single-copy region (SSC) [46].

The GC content in A. bottae was 37.73%, but it was lower in A. pulchellus subsp. Mollis (36.49%), A. pulchellus subsp. Cantoniensis, and A. precatorius (36.5%). Such small variation in the percentage of GCs within a genus has been observed in some legumes [47]. A total of 130 genes were identified, including 111 unique genes that were observed in Abrus botta. This finding is largely consistent with the number of genes reported in both species of A. pulchellus subsp. Cantoniensis and A. precatorius [16].

In the Abrus bottae cp genome, the ycf3 and clpP genes have two introns, whereas the remaining genes have only one intron. In addition, the trnK-UUU gene hosts the longest intron (2499 bp). These characteristics are consistent with what has been recorded in other species of the genus A. pulchellus subsp. Cantoniensis, as indicated by Xu et al. [16].

Analysis of proteins encoded by A. bottae revealed the presence of 24398 codons, with leucine emerging as the most prevalent amino acid. In contrast, cysteine was the rarest. This is similar to what has been observed in A. pulchellus subsp. Cantoniensis and A. precatorius [16], and it is also found in most angiosperms [16,48]. Among the A. bottae having 28 codons with RSCU values greater than 1, A/U-endings exhibited the most codon usage (85,7%). Codon-usage-bias A/U-endings are common in other angiosperms [49,50].

RNA editing is a post-transcriptional modification that may cause changes in the coding sequences of primary RNA transcripts. Thus, the detection of RNA editing sites in chloroplasts is expected to reveal important clues for evolution [45].

Its editing sites were more prevalent in the rpoC2, ycf2, and ndhF genes. The rpoC2 gene is an important transcriptional regulator of chloroplast genes, and the majority of these genes are essential for photosynthesis [51], suggesting that ycf2 may be essential for cellular survival. On the other hand, the NADH dehydrogenase subunits are essential for the electron transport chain, which leads to ATP production and is also a part of the photosynthesis system in plants. RNA editing likely plays a key role in the function of the NDH protein complex, which potentially contributes to increasing photosynthesis efficiency and is under positive selection during the course of evolution [44,52].

In total, in A. bottae, most conversions (64.91%) of RNA editing led to amino acid sequence alterations that promoted conversion from hydrophilic to hydrophobic acids, such as histidine to tyrosine (H→Y), leucine to phenylalanine (L→F), threonine to isoleucine (T→I), or proline and serine to leucine (P or S→L) and alanine to valine (A →V). Analogous findings have also been observed in the cp genomes of other species of Abrus, i.e., “A. pulchellus subsp. Cantoniensis and A. precatorius”. It has been shown that RNA editing mainly occurs at the first or second position of the codon that results in transitions of amino acids that are hydrophilic to hydrophobic and polar to nonpolar, leading, in turn to, an increase in the hydrophobicity of the protein [53].

During our study of the Abrus spp., we found long repeats under 30 bp and those at 30–60 bp. Repetitive sequences are involved in many different biological processes. Examples include regulating gene expression; organising the genome; DNA replication; and recombination and repair mechanisms [53]. The ycf2 gene possessed the longest repeats, with six palindromic repeats and five forward repeats. An equivalent observed was reported in the species A. pulchellus subsp. Cantoniensis [16], which could be indicative of lineage-specific genome reorganization and genetic differentiation in the Abreae tribe.

Molecular markers serve essential functions in genetic variation research at the DNA level. They provide a consistent and reliable means for genetic evaluation, especially when morphological characters cannot always be trusted because of environmental influences. The chloroplast DNA simple-sequence repeats (cpSSRs) have become instrumental resources thanks to their excellent repeatability, enormous polymorphism, and robust stability [54]. This is especially important given the rapid and unprecedented progress in plant genomic research. So far, cpSSRs have made much headway in investigations related to plant population genetic diversity, analyses of population structure, classification of populations, and their biogeographic distributions. For example, polymorphic cpSSR markers were identified and used to investigate the genetic diversity as well as the population genetic structure of endemic Paeonia suffruticosa species in China [54]. Götz et al. [55] utilized chloroplast genomes from economically significant Juglans species to develop a highly polymorphic and reliable set of cpSSR markers. They were then used to test genetic differentiation among species, within species and for population genetic analysis. Zhou et al. [56] developed cpSSR markers from the chloroplast genome of Orchidantha chinensis T. L. Wu, a species endemic to China, to examine its genetic variation. Furthermore, the research by Hladnik et al. [57] utilized SSRs identified in the chloroplast genome of Helichrysum italicum (Roth) G. Don, a Mediterranean medicinal plant with significant potential to detect its genetic diversity. Furthermore, cpSSR markers were used to evaluate the genetic diversity and population structure of Phoebe zhennan and the impacts of environmental factors on its population structure [58], which is one of China’s most precious and best protected tree species. In the present study, a total of 335 simple-sequence repeats (SSRs) have been identified within the chloroplast genome of A. bottae, with observed counts ranging from 323 to 344 in plants of this genus investigated so far, and predominantly, they are single-nucleotide (79.65–85.37%). The chloroplast genomic data of Abrus species can be used to establish cpSSR markers that will be instrumental in the exploration, identification, and definition of medicinal species that will address taxonomic issues on a species or even interspecific level and will also allow us to assess the genetic diversity and population structures within the Abreae lineage.

On the other hand, the main type of mononucleotide repeat observed was A/T. This finding is consistent with previous research that found most chloroplast SSRs are made up of A or T nucleotides [53]. It is worth mentioning that the repeats rarely involve G-C bases. This probably arises from G-C bases forming three hydrogen bonds, making them more stable and less susceptible to disruption. Thus, they are less likely to engage in genetic recombination processes [53].

In the present study, the majority of SSRs were located in the LSC region followed by the SSC region, whereas relatively few were located in the IR regions in the four Abrus species; these observations are also supported [16] by the cp genome of Abrus pulchellus subsp. Cantoniensis.

Studies focused on the contraction and expansion of the IR region in angiosperms have revealed changes in the length and location of genes at the border such as ycf1, rps19, and ndhF [59], which contribute to variations in the plastome length [44]. This structural contraction of the IR region could account for the expansion of the trnH gene in the LSC region, the ycf1 gene in the SSC region, and the rps19 gene in the LSC region in A. bottae.

The ycf1 gene plays an essential role in the plant life cycle, has a high level of varia-tion among the plant cp genomes, and is thus regarded as a potential candidate locus for studying angiosperm phylogeny [59]. Comparative analysis of the IR boundaries in four Abreae species revealed striking disparities in the lengths of the ycf1 genes in the IRa region (A. bottae, A. pulchellus subsp. Cantoniensis, and A. precatorius).

The Fabaceae family Papilionoideae subfamily is a particularly suitable model for assessing the evolutionary mechanisms of cp genomes because remarkable genomic reorganization has been observed within this clade [47]. Nevertheless, the sizes of the LSC, SSC, and IRs, as well as the changes in the IR boundaries, GC contents, and SSR distributions, displayed small variations in the Abrus genus and were similar to those of other legume genera [43,45,46].

Notably, the genes with Pi values greater than 0.04 were located mainly in the LSC region, thus supporting the conservation and stability of the IR region, as it has been suggested that a copy-dependent repair mechanism is applied between IRs in chloroplasts [52]. This may explain the apparent decrease in sequence divergence of the IR region compared with that of the SC region. Moreover, significantly greater values of Pi (0.04–0.06) were also detected for the atpE, atpF, atpH, atpI, ndhC, ndhK, psaB, psbC, psbM, psbZ, rpoB, rpoC1, rpoC2, rps14, rps16, rps2, rps4, ycf3, trnD-GUC, trnF-GAA, trnG-GCC, trnG-UCC, trnL-UAA, trnS-GGA, trnT-GGU, and trnT-UGU genes using nucleotide diversity analysis. These loci furnish a substantial amount of genetic diversity that is important for phylogenetic and biodiversity studies. A previous study revealed that some genes (atpF, ndhC, rps16, rpoC1, and rpoC2) are positively selected and have a high degree of evolutionary divergence in different species [46]. Hence, they should focus on identifying molecular markers and new DNA barcodes for the Abrus species.

On the other hand, Ka/Ks is a standard measure of sequence diversity between different species or taxa of uncertain evolutionary distance and nucleotide substitution, natural selection, and advantageous mutation in genes subjected to selective pressure [48]. Specifically, within the Abrus species, the rpl20, rpl23, rps16, ycf2, and ycf4 genes exhibited Ka/Ks values > 1, which indicated positive selection, whereas the ycf2 gene presented the highest rates of nucleotide substitutions among the Abrus spp. Similar results were also reported in a number of Fabaceae species, and these results are suspected to be related to adaptation to arid habitats [46]. Moreover, the results of positive selection on the ycf4 gene revealed that Lathyrus, Pisum, and Vavilovia, as members of tribe Fabeae, have undergone some adaptive evolutionary patterns [52].

Although the strategies used to address the systematic and phylogenetic problems of the Fabaceae family are important for the discovery of phylogenetic signals, they may not provide a comprehensive understanding of specific relationships that are more complex. These problems have been attributed to the relative diversification of these lineages [60,61] as well as palaeopolyploidy [62].

In this context, the Legume Phylogeny Working Group (LPWG) produced the largest legume phylogeny (covering 3696 species) and proposed a novel classification of six subfamilies LPWG [63]. In addition, other metrics, such as the use of the chloroplast genomes of 187 species [9], were used. Moreover, a phylogenetic tree of the Papilionoid subfamily was established with 244 taxa with complete plastomes on the basis of available whole plastome sequences, including 237 papilionoids and 174 genera [10].

To study the phylogenetic relationships of Abrus bottae, an endemic species from the Arabian Peninsula, an in-depth MrBayes analysis was built for 46 taxa, including the tribal levels of the Papilionoideae subfamily (Fabaceae), on the basis of chloroplast genome sequences. Most values of node PP are supported posterior probabilities (PPs = 1). The phylogenetic tree constructed revealed that the largest group was the nonprotein amino acid-accumulating clade (NPAAA). This clade could be broken down into two major subclades: Millettioid and Hologalegina. This finding is congruent with all prior molecular phylogenetic estimates [10].

Indigofereae is a sister group of the Millettioid clade, which includes taxa from many tribes, such as Dicocleae, Desmodieae, Indigofereae, and Psoraleeae, and has strongly supported relationships, with support values (PPs = 1). The four species of Abrus included within the tribe Abreae under the Indigofereae + Millettioid clade are monophyletic and have well supported evolutionary relationships (PPs = 1). Abrus pulchellus subsp. Cantoniensis is a sister of Abrus precatorius, and it forms a monophyletic relationship with Abrus pulchellus subsp. Mollis and Abrus bottae posterior probabilities (PPs = 1). These findings are in agreement with those of a recent study [14]. In addition, the current phylogenetic tree demonstrated that the genus Abrus is more closely related to Canavalia cathartica, with notable support (PP = 1) that is in agreement with previous viewpoints on the shared ancestor of the genera Abrus and Canavalia [9,16].

The Hologalegina clade has two subclades, the Robinoid clade (Sesbanieae, Loteae, and Robinieae) and the IRLC clade (Wisterieae, Cicereae, Fabeae, Hedysareae, and Galegeae), which is in agreement with earlier findings [10,64]. Most of these clades are strongly supported (PP = 1). Some subbranches of the Robinoid clade, however, showed weaker support (PP = 0.50). The exact phylogenetic position of this plastome structure within Sesbania or related groups in the Robinoid clade is poorly resolved [10].

Our analysis revealed that the Dalbergioid clade (species from the tribes Amorpheae and Dalbergieae) was also a sister to the NPAAA clade. The clade consisting of the genera, which includes taxa from the Borngniartieae, Podalyieae, and Crotalarieae tribes, was placed at the base of the papilionoid tree and was the base of the two lineages.

5. Conclusions

Abrus bottae is limited to a certain geographic region in southern Arabia. It is placed under the genus Abrus. This genus has many taxonomic issues related to the number of species in the genus as well as the systematics of its species. In this study, the whole-chloroplast genome of the endemic species Abrus bottae was sequenced, and a comparative analysis was performed to clarify its basic architecture, conservation, and divergence from closely related species of the Abreae tribe.

Noncoding gene regions are more conserved than noncoding regions are, with intergenic repeat (IR) regions being more conserved than large single-copy (LSC) or small single-copy (SSC) regions. Our study suggested that approximately 335 simple chloroplast sequence repeats (cpSSRs) have the ability to function as molecular markers to study intraspecific variation and support the identification of interspecific relationships among the Abrus species.

The species showed some differences in the organization of genes in the IR (especially at the borders of the IR-SC), likely because of the expansion and contraction of the IR. The IR and LSC regions of A. bottae were slightly smaller than those of the other Abrus species. Nevertheless, these species maintained a relatively constant structural pattern and nearly identical gene content.

The constructed phylogenetic tree supported the close phylogenetic relationships between Abrus species and between Abrus and Canavalia species. This result provides evidence for the phylogenetic relationships within the subfamily Papilionoideae, as demonstrated by earlier studies. The results of this research provide a valuable reference that will facilitate further species identification, clarification of evolutionary relationships, and future genetic resource development for the Abrus genus and, subsequently, help in resolving the classification difficulty of this genus.