Introduction

Pepper (Capsicum spp.) is the most economically significant genus within the family Solanaceae and contains approximately 35 species, of which five major species represent ancient domestication [1]. The production and usage of Capsicum has spread worldwide across civilizations, which subsequently led to selective pressure associated with cultivation under different environmental conditions [2]. According to the recent data (FAOSTAT; http://faostat.fao.org/), worldwide pepper production surpassed 3.68 million hectares (Mhs), yielding a total production of 36.29 million tons (Mts) of green peppers and 4.84 Mts of dried peppers. Pepper fruits are used in food and non-food processing applications in many innovative ways [3,4,5]. The genus Capsicum comprises domesticated and wild species with highly variable morphologies. Systemic classification of the Capsicum genus was achieved by considering morphology, cytogenetics, biochemistry, and ethnobotanical data [6]. Recently, approximately 34 of the 35 species were recognized using comprehensive phylogenetic analysis [7]. In subsequent iterations of the analysis, 11 major clades were meticulously resolved within the Capsicum genus. These clades were then assigned informal designations to aid their identification and characterization [8, 9].

Indeed, archaeological evidence indicates separate domestication events among cultivated pepper species occurred as early as 6000 years ago [10]. Pepper made its voyage from the West Indies to Europe during the late 15th and early 16th centuries, and its rapid distribution to Africa and Asia, including China, holds the earliest documentation of pepper [11, 12]. The course of domestication and controlled breeding yielded non-deciduous pepper cultivars exhibiting a wide array of morphologically and agronomically important traits, including diverse fruit shapes, varying sizes, distinct weights, pendulous fruit orientation, as well as a broad spectrum of pungency levels [13]. More recently, large, sweet, and non-pungent bell peppers have emerged through selective breeding, with the earliest records dating to the 1700s [14].

Capsicum annuum, commonly known as pepper, has been extensively bred and is primarily considered in breeding programs to develop genetically superior genotypes. Selective pressure leads to the advent of genomic selection signals, which are quantified as decreases in nucleotide diversity [15]. Several genomic selection signals have been detected in the pepper genome owing to early domestication [16]. Initial evolutionary shifts within the slender-fruited pepper category were marked by selective pressure acting on extensive assemblages of candidate genes implicated in determination of fruit pungency and/or shape. This selective phenomenon was likely facilitated by the considerable genetic diversity existing within these groups, which was amplified by the absence of genetic bottlenecks [16, 17]. Conversely, the subsequent genomic transition exhibited targeted selection towards a narrower subset of genes. This observation supports the hypothesis that genetic bottlenecks may occur during this transition. This bottleneck scenario is plausibly attributable to the transport of a subset of gene pools via the historical Silk Road route towards East Asia [16].

Wild Capsicum species such as C. chinense, C. frutescens, C. baccatum, and C. pubescence are well-known for their intercrossing characteristics with C. annuum to produce fertile progenies. Thus, to enhance breeding efficiency of C. annuum, inter-breeding of C. annuum with other wild Capsicum species has been performed [18]. Several agronomic and physiological traits found in wild Capsicum species are considered valuable genetic sources to improve the characteristics of C. annuum [19]. Especially, capsaicinoids comprise a group of substances produced only within the Capsicum genus, which is one of the pivotal criteria for hot pepper breeding [20]. Research on various Capsicum species is not only relevant to improve C. annuum, but also to discover novel species. This may lead to domestication and preservation of their genetic pool in Capsicum species and may play a significant role in conservation of Capsicum species.

Despite the economic importance of C. annuum, the availability of wild Capsicum species for use in cultivar breeding is relatively limited within the Solanum family. Recent advances in next-generation sequencing (NGS) technology have made whole-genome sequencing and analysis much more straightforward than in the past. Whole genome-scale datasets encompass a larger number of single nucleotide polymorphisms (SNPs), which can significantly aid in reconstructing well-supported species-specific phylogenetic trees [21]. However, the large genome size of Capsicum species (> 3 Gbp) with long terminal repeat (LTR) sequences hinders its genomic analysis [22,23,24]. Plastid genome assembly and analysis are relatively simple approaches for analyzing the phylogenetic history of a species [25] and are highly useful in resolving taxonomic complexes within and between species and provides insights for breeding programs [26]. In particular, maternal inheritance of plastids provides stable molecular evidence for each species.

Despite the presence of well-annotated genomic sequences for numerous pepper species, our comprehension of the molecular evolutionary trajectory of this crop trails behind that of its closely related counterpart, Solanum. To fill this gap, we sequenced three wild Capsicum species, and compared their genomes to other publicly available plastid genomes of Capsicum species. The findings of this study may provide new guidelines for evolutionary framework of Capsicum species and facilitate the development and utilization of its genetic resources.

Materials and methods

Plant materials and DNA extraction

Seeds of C. chacoense, C. galapagoense, and C. tovarii (Accession Nos. 261663, 236750, and K057567, respectively) were obtained from the National Agrobiodiversity Center, Republic of Korea. These seeds were germinated in a greenhouse and fresh leaves from 40-day-old seedlings were ground in liquid nitrogen using a mortar and pestle. Genomic DNA was extracted using the Exgene Plant SV midi kit (GeneAll Biotechnology, Seoul, Korea) following the manufacturer’s instructions. The quantity and quality of the extracted DNA were measured using a NanoDrop ND-1000 spectrophotometer (Thermo Scientific Inc., Wilmington, DE, USA).

Genome sequencing, assembly, and annotation

An Illumina paired-end DNA library with an average insert size of 500 base pairs (bp) was constructed using the Illumina TruSeq library preparation kit, following the manufacturer’s instructions. Subsequently, the library was sequenced using the MiSeq instrument at LabGenomics (http://www.labgenomics.co.kr/) to generate paired-end reads of 300 bp each. Complete cp. genome sequences were assembled through de novo assembly of low-coverage whole genome sequence (WGS) data. Low-quality sequences (quality score < 20, Q20) were initially filtered out, and the remaining high-quality reads were assembled using the CLC Genome Assembler (version beta 4.6, CLC Inc., Aarhus, Denmark), employing an overlap size ranging from 200 to 600 bp. Next, chloroplast contigs were identified within the initial assembly using BLAST (ver. 2.2.31) search against reference cp. (NCBI accession NC_018552) sequences. The selected contigs were oriented to construct the final complete cp. genome structure, and manual corrections were made to address any ambiguous nucleotides or gaps.

The assembled chloroplast genome was annotated using the web-based program Dual Organellar GenoMe Annotator (DOGMA; http://dogma.ccbb.utexas.edu/) with default parameters. This annotation encompasses the prediction of protein-coding, transfer RNA (tRNA), and ribosomal RNA (rRNA) genes. Next, the positions of intron-containing genes were identified using BLASTN (ver. 2.2.31) alignment of the assembled cp. genome against a reference cp. genome. A graphical map illustrating the arrangement of genes within the cp. genome was generated using CPGView (http://www.1kmpg.cn/cpgview/; Accessed in July 2023). Additionally, cp. genomes of 19 Capsicum species were retrieved from the NCBI database for subsequent comparative analysis (Supplementary Table S1).

Codon usage analysis

A total of 87 inherited protein-coding genes within three Capsicum cp. genomes of C. chacoense, C. galapagoense and C. tovarii were concatenated to facilitate computation of synonymous codon usage analysis. Number of codons (Nc) and relative synonymous codon usage (RSCU) were calculated using the CodonW1.4.2 (Available online: http://downloads.fyxm.net/CodonW-76666.html) program. The Nc, present within individual genes, serves as a straightforward gauge of codon bias [27], with Nc values spanning a spectrum ranging from 20 (indicative of extreme bias) to 61 (totally unbiased). RSCU is the ratio of observed frequency of a specific codon to its anticipated frequency [28]. In situations where synonymous codons, each encoding the same amino acid, were utilized with equal frequency, the RSCU values were approximately 1.0, signifying an absence of bias. In addition, the frequency distribution of amino acids (AA) was quantified and is represented as the percentage of codons encoding a particular amino acid relative to the total codon count.

Comparative analysis of cp. Genomes

Expansion/contraction events in the cp. genomes of 10 Capsicum species (Table 1) were compared using IRScope (https://irscope.shinyapps.io/irapp/). Complete cp. genome sequences of Capsicum species were compared using the mVISTA online tool (http://genome.lbl.gov/vista/mvista/submit.shtml). This analysis was performed using the Shuffle-LAGAN mode [29]. To determine nucleotide diversity in the cp. genomes, the 10 Capsicum cp. genome sequences were aligned using the program MAFFT 7.388 [30]. Ambiguously aligned loci (‘N’) were excluded from the analyses. Unambiguously aligned DNA sequences were used for the nucleotide diversity analysis. To identify hypervariable sites among Capsicum chloroplast genomes, a sliding window analysis was performed using DnaSP v6 software [31]. This analysis utilized a window length of 600 base pairs and a step size of 200 base pairs to calculate the nucleotide diversity (π).

Table 1 Summary statistics for the sequencing and assembly of nine Capsicum species chloroplast genomes
Full size table

Phylogenetic analysis and divergence time estimation

Twenty-two cp. genomes of Capsicum species, including three newly sequenced cp. genomes of species such as C. chacoense, C. galapagoense, and C. tovarii, were used to construct Maximum Likelihood (ML) phylogenic trees employing the Kimura 2-parameter model (with 1000 bootstrap repeats) using MEGA X [32] after aligning whole cp. genome sequences using MAFFT 7.388. The divergence times among Capsicum species were estimated using the Bayesian method implemented in the BEAST program [33]. In BEAST, the lengths of the branches are converted into units of evolutionary time. Essentially, this allows the tree to be placed in a time-based context for a better understanding of evolutionary processes.

In the present study, an uncorrelated relaxed lognormal clock model was employed to enable independent optimization of evolutionary rates for each branch in the phylogenetic tree. The speciation process was represented using the Yule process [34]. Branching rates were determined using the GTR + I + G nucleotide substitution model, and two separate Markov chain Monte Carlo (MCMC) runs were conducted using BEAST, each spanning 100,000,000 generations. Sampling was performed at intervals of 10,000 generations during these runs to ensure comprehensive exploration of the parameter space and convergence of the analysis.

To calibrate the phylogenetic tree of cp. genomes for the 22 Capsicum species accessions, we incorporated estimated dates of speciation events between Capsicum and Solanum species, which were sourced from www.timetree.org. These events were indicated to have occurred approximately 21.3 million years ago (Mya). In our study, Solanum lycopersicum and Lycianthes radiata were designated as outgroup species, and age of the tree roots was constrained by a normal distribution, with a mean of 21.38 Mya. To further analyze the results generated by BEAST, we used Tracer v. 1.6 [35] to explore the output, enabling us to assess various statistical aspects of the phylogenetic analysis. Additionally, TreeAnnotator v.1.8, a component of the BEAST package, was employed to generate maximum clade credibility trees from the post-burn-in trees. This process facilitated determination of 95% posterior density of ages for all nodes within the phylogenetic tree, thus providing crucial insights into the timeline of speciation events among Capsicum species. To visualize the outcomes, maximum clade credibility trees were annotated using FigTree v1.4.4 (http://tree.bio.ed.ac.uk/software/figtree/, Accessed on July 18, 2023). Principal component analysis (PCA) was conducted on all samples utilizing TASSEL 5.0 [36], and the resulting data were visualized using XLSTAT software (www.xlstat.com, Accessed on July 10, 2024).

Results

Plastid genome size and Organization

This study focused on analysis of cp. genomes of three Capsicum species that are wild relatives of domesticated Capsicum species. The HiSeq2500 generated approximately 7.75 GB of paired-end sequence data. The plastid genome coverage for C. chacoense, C. galapagoense, and C. tovarii with the reference genome reached 206x, 184x, and 156x, respectively. The cp. genomes of the three Capsicum species, along with their sizes, reads, GC contents, and averages are shown in Table 1. We obtained complete cp. genome maps of C. chacoense, C. galapagoense, and C. tovarii through a de novo assembly followed by reference-guided assembly using the C. annuum (NC_018552) cp. genome as a reference (Fig. 1A, B, and C). Size of the Capsicum cp. genome was 156,995 bp in C. chacoense, 156,959 bp in C. galapagoense, and 156,818 bp in C. tovarii (Table 1). The complete cp. genome size of Capsicum species was very similar to that of the reference genome (approximately 156 kb).

Fig. 1
figure 1

Structure of the cp. genomes of three Capsicum species (C. chacoense, C. galapagoense, and C. tovarii). The map consists of six tracks. Starting from the center, the first track shows scattered repeats (direct and palindromic) linked by red and green arcs. The second track displays long tandem repeats as blue bars, while the third track shows short tandem repeats (microsatellites) as differently colored short bars based on their sizes. Colors correspond to different repeat sizes: black for complex repeats, green (size 1), yellow (size 2), purple (size 3), blue (size 4), orange (size 5), and red (size 6). The fourth track marks specific regions of the genome. The fifth track depicts the GC content, and the sixth track highlights the genes, named by codon usage bias and colored by their functions. Inner genes are transcribed clockwise, and outer genes are transcribed anticlockwise. The bottom left corner denotes the gene functions

Full size image

Combined with the newly assembled cp. genomes, we collected 19 available cp. genomes for Capsicum species, of which 10 were chosen to characterize genomic features of inter-species cp. genome (Table 1). The lengths of the 10 cp. genomes varied from 156,583 (C.lycianthoides) to 157,390 bp (C.pubescens). As reported previously, all the 10 Capsicum cp. genomes contained two copies of IR regions, each ranging from 25,792 (C. tovarii and C. frutescens) to 25,887 bp (C.pubescens). The IRs were separated by an LSC region ranging from 87,330 (C. chinense) to 87,688 bp (C.pubescens) and an SSC region ranging from 17,853 (C. tovarii and C. frutescens) to 17,929 bp (C. annuum var. annuum). The overall GC content in all the 10 Capsicum cp. genomes was similar to that reported previously, with an average of 37.71% (Table 1).

The three Capsicum cp. genomes sequenced in this study contained 132 genes, comprising 87 protein-coding genes (CDS), eight rRNAs, and 37 tRNAs (Table 2). Seventeen genes containing introns were identified in the cp. genome. Among the annotated genes, 14 (trnA-UGC, trnG-UCC, trnK-UUU, trnL-UAA, trnV-UAC, rps16, rpl2, rpl16, rpoC1, ndhA, ndhB, petB, petD, and atpF) contained one intron each, and three (rps12, ycf3, and clpP) contained two introns each (Table 2). In addition, 19 genes were determined to be duplicated, of which three genes such as rps19, rpl2, and rpl23 were found to have one or two copies. The complete cp. genome sequences with gene annotations have been submitted to NCBI under GenBank accession numbers NC_033525 for C. chacoense, NC_033524 for C. galapagoense, and KX913219 for C. tovarii.

Table 2 List of annotated genes in the cp. Genomes of C. Chacoense, C. Galapagoense, and C. Tovarii
Full size table

Codon usage analysis

Codon usage frequency of cp. genomes of three Capsicum species such as C. chacoense, C. galapagoense and C. tovarii was calculated based on protein-coding genes (CDSs). The number of codons within the protein-coding genes of the three plastid genomes ranged from 26,945 (C. chacoense) to 26,991 (C. tovarii). In all three investigated species, leucine (leu; 2908–2957 occurrences) was the most abundant amino acid that was encoded by six codons, followed by isoleucine (ile; 2142–2249 occurrences) that was encoded by three codons (Supplementary Table S2). Conversely, methionine and tryptophan were encoded by a single codon, with occurrences ranging from 593 to 628 and 485–522, respectively. Notably, these amino acids exhibited no codon usage bias, as reflected by their RSCU values of 1. The UUA codon for leucine displayed the highest RSCU values (1.68–1.86), while the AGC codon for serine exhibited the lowest RSCU values (0.37–0.40) across the three cp. genomes. In the cumulative cp. genomes, a set of 30 codons featured RSCU values exceeding one, indicating their preferred usage. Notably, of these 30 codons, 29 culminated in adenine (A) or uracil (U), highlighting a discernible codon preference within the cp. genome (Supplementary Table S2).

Comparative analysis of junction regions

Capsicum plastid genomes displayed a typical quadripartite structure consisting of a pair of IR regions separated by LSC and SSC regions (Fig. 2). Among the 10 species, the SSC region of C. lycianthoides was the largest, whereas those of C. tovarii and C. frutescens were the smallest. Similarly, the IR regions of C. baccatum were the largest and those of C. lycianthoides were the smallest (Fig. 2). Within all the 10 examined plastid genomes, four distinct boundaries were identified between the regions LSC–IR and SSC–IR. Genes situated in proximity to these boundaries exhibited consistent patterns across all the analyzed chloroplast genomes (Fig. 2). Specifically, regarding the junction between the LSC and IRB regions (JLB), the gene rps19 was observed to traverse from the LSC region to the IRB region; however, different lengths were observed from 61 (C. galapagoense, C. chinense and C. baccatum) to 73 bp (C. lycianthoides) in all the species analyzed (Fig. 2). The rpl22 and rpl2 genes, which were adjacent to rps19 gene at the JLB boundary, were located in the LSC and IRB regions, respectively.

Fig. 2
figure 2

Comparison of boundaries of LSC, SSC, and IR regions among 10 cp. genomes of Capsicum species. The boundary points are labeled as JLB, JSB, JSA, and JLA; each label represents specific junctions related to LSC and IRb, IRb and SSC, SSC and IRa, and IRa and LSC, respectively

Full size image

Similarly, for the junction between the SSC and IRB regions (JSB), two genes, ycf1 and ndhF, were placed next to the boundary. The ndhF gene was located in the SSC region, whereas ycf1 was observed to cross over from the SSC region to IRB region, except for C. pubescens and C. lycianthoides. The ycf1 gene of C. pubescens was absent in the IRB region, and the ndhF gene in SSC was also 12 bp away from the boundary. The ycf1 gene of C. lycianthoides was smaller in size (1043 bp) than the other species and was also located 1 bp away from the boundary. By contrast, the ndhF gene (2222 bp) in C. annuum plastome showed a difference in position and was located adjacent to the JSB boundary; the ycf1 gene was located approximately 1127 bp in the IRB and 21 bp in SSC region.

Junction analysis between SSC and IRA regions (JSA) demonstrated that ycf1 was found to be intact at the JSA boundary and crossed over from the SSC into IRA region. However, the ycf1 gene length in the SSC region varied from 4578 (C. chacoense) to 4650 bp (C. lycianthoides), whereas in the IRA region, all the species were of 1128 bp, except C. lycianthoides (1044 bp) and C. annuum (1127 bp). In contrast, junction analysis between IRA and LSC regions (JLA) showed that both the trnH and psbA genes were placed in the LSC region and a copy of rpl2 gene was precisely located in the IRA region, except C. lycianthoides in which the copy of rpl2 gene was absent in the IRA region.

Comparative sequence variance analysis

We assessed variations in cp. genomes within congeneric Capsicum species using the online alignment tool mVISTA, with C. frutescens as a reference (Fig. 3). The results showed that there was no gene rearrangement in the Capsicum cp. genome, and the sequence variation in the non-coding regions was higher than that of genes in coding region. LSC and SSC had more diverse loci than IR; several regions were variable and were able to distinguish different populations. Nucleotide polymorphisms (Pi) in the 10 Capsicum cp. genomes, were analyzed by sliding window analysis using multiple-sequence alignment, and C. annuum showed high sequence similarity (Fig. 4). A total of 1486 common polymorphic sites were detected in the cp. genomes, with a mean nucleotide diversity (Pi) value of 0.00225. The mean nucleotide diversity (Pi) values of the LSC, SSC, and IR regions were 0.00264, 0.00350, and 0.00122, respectively. Four nucleotide polymorphism hotspots (hotspots) were discovered in intergenic regions (trnT-UGU/trnL-UAA) and four genes (ccsA/ndhD/rps12/ycf1) with Pi values above 0.01 were located in single-copy regions.

Fig. 3
figure 3

A sequence similarity plot was generated using the mVISTA tool to compare the cp. genomes of Capsicum species. KR078312 (C. frutescens) was used as a reference genome. The Y-axis shows percentage of sequence similarity ranging from 50–100%. The direction in which genes are transcribed is indicated by gray arrows. Non-coding sequences (CNS) are depicted as pink bars, while exons are shown as purple bars. Variations in the genome are shown as white peaks

Full size image
Fig. 4
figure 4

Sliding window analysis was performed on complete cp. genomes of 10 Capsicum species. Nucleotide variability across these genomes was evaluated using DnaSP, employing a window length of 600 base pairs and a step size of 200 base pairs. The X-axis denotes different regions of the cp. genome, whereas the Y-axis represents nucleotide diversity (pi) for each window. Five particularly hypervariable regions, characterized by pi values exceeding 0.01, have been identified and labeled

Full size image

Phylogenetic analysis and estimation of divergence time

To investigate phylogenetic relationships of the Capsicum species, alignments of all 22 cp. genome sequences were performed using MAFFT, and a phylogenic tree was created using the Maximum Likelihood (ML) method with a bootstrap repeat of 1,000 under the Kimura 2-parameter model; Solanum lycopersicum and Lycianthes radiata were selected as the outgroup species. The ML phylogenetic analysis showed strong bootstrap support (between 55 and 100) for most branches, revealing five separate clades (data not shown). To determine evolution of Capsicum species, we repeated the ML phylogenetic analysis for divergence time estimation using the BEAST program, based on the uncorrelated relaxed lognormal clock model time using GTR + I + G category from the MAFFT alignment of all the 22 cp. genome sequences. The results of analysis of the MCMC tree related to species divergence times in Capsicum species are shown in Fig. 5. Speciation times were estimated in millions of years to assess the age of the clades.

Fig. 5
figure 5

Phylogenetic analysis and estimation of divergence time of 22 Capsicum species based on maximum-likelihood analyses and Bayesian inference. S. lycopersicum, and L. radiata were used as the outgroup. The number mentioned on each node indicates estimated divergence time in My (Million years), and the number on each clade indicates Bayesian posterior probabilities from the BEAST analyses. Divergence time of Capsicum family was derived from TimeTree5 (http://www.timetree.org/) using the ‘Get an evolutionary timeline’ function

Full size image

Similar to the ML phylogenetic tree, the C. tovarii, accession of this study found with a C. frutescens accession along with C. annuum accessions with strong support in a different clade (Fig. 5). Similarly, C. galapagoense accession sequenced in this study was grouped with the C. chinense and C. eximum clades, and the C. chacoense accession sequenced in this study was shown to be a sister species strongly related to the C. baccatum complex. The divergence time estimation analysis verified that the Capsicum cp. genome diverged from the outgroups S. lycopersicum and L. radiata ~ 25.79 million years ago (Mya). The most recent common ancestor of all Capsicum cp. genome species was estimated to be ~ 20.13 Mya. As reported in an earlier study, the C. pubescens accession is a sister species to both the C. annuum and C. baccatum lineages, which are estimated to diverge at ~ 8.72 Mya. Our estimation of the Capsicum group speciation showed that the divergence time of C. annuum and C. baccatum complexes was about ~ 5.83 Mya. The cultivated varieties within C. annuum complexes such as, C. annuum, C. frutescens, C. tovarii, C. chinense, C. eximum and C. galapagoense. diverged approximately 3.66 Mya, while C. eximum and C. galapagoense subgroups diverged from C. chinense approximately 1.75 and 1.19 Mya respectively. Similarly, the domesticated C. annuum species started diverging ~ 1.86 Mya, while C. frutescens and C. tovarii subgroups diverged from C. annuum approximately 0.88 Mya. Thus, this study suggests that domesticated C. annuum and C. chinense are the most divergent species groups among the analyzed genotypes.

The PCA based on the entire cp. genome sequences showed five significant principal components (Fig. 6). The PCA results revealed that the first three principal components, PCA1, PCA2, and PCA3, accounted for 32.55%, 24.40%, and 30.76% of the total variation. In a manner consistent with the ML phylogenetic tree, the cultivated species within the C. annuum complex formed a closely related cluster. The accessions of C. galapagoense and C. eximium were grouped with C. chinense, whereas the accession of C. chacoense formed a closely related group with the C. baccatum complex.

Fig. 6
figure 6

Principal component analysis (PCA) of 22 Capsicum species based on cp. genome data. S. lycopersicum and L. radiata were used as outgroups

Full size image

Discussion

Typical characteristics of the cp. Genome

Plastid genomes are highly stable, non-recombinant, and uniparentally inherited with low mutation rates, making them a valuable source of information to produce highly reliable phylogenetic trees and to elucidate their evolutionary history [37, 38]. Though exception on different mode of inheritance and genome rearrangements have been reported in plastid genome. Advances in next-generation sequencing (NGS) techniques have significantly expedited and made sequencing of cp. genomes more cost-effective. In this study, we generated complete cp. genome sequences for three Capsicum species (Fig. 1A, B, and C) using NGS. We compared the genomic features, content, and organization of the plastid genomes of 10 different Capsicum species (Table 1) that showed a typical quadripartite structure [39]. The length of the plastid genomes varied from 156,583 (C.lycianthoides) to 157,390 bp (C.pubescens), suggesting that among the Solanaceae family, variation in plastome size is a distinct feature of cultivated Capsicum, as reported previously [40]. The length of the LSC ranged from 86,813 bp (C. lycianthoides) to 87,688 bp (C. pubescens), whereas the SSC region was quite large (18, 522 bp) in the C. lycianthoides cp. genome. However, the length of the IR region was less variable, from 25,624 bp (C.lycianthoides) to 25,910 bp (C. baccatum). The GC content averaged around 37.72% in the 10 complete Capsicum cp. genomes, showing a strong similarity between them. Some instances of gene loss were identified in these cp. genomes (Table 1).

The variation observed in the cp. genome size was significantly affected by multiple factors, including gene loss, variations within IR regions, and differences in intergenic spacer regions, as reported in previous studies [41,42,43]. Within the set of the 10 Capsicum species investigated, 127–132 cp-coding genes were reliably identified. Notably, an absence of a single instance of the mRNA gene rps12 in the cp. genomes of C. annuum, C. baccatum, and C. frutescens was observed. Similarly, single copies of the mRNA genes ycf1 and ycf15 were absent from the cp. genome of C. pubescens. Of particular significance was the observation that there was a complete loss of the mRNA gene psbZ along with deletion of a copy of the trnV-GAC gene in the cp. genome of C. lycianthoides. Similarly, four mRNA genes, psbB, rpl2, ycf1, and ycf15 were found to be lacking, whereas an additional mRNA gene, lhbA, was found to augment the cp. genome structure. Indeed, several notable changes in the cp. genomes of various plant species have been documented [44,45,46]. These include incorporation of the psbZ gene; absence of mRNA genes such as lhbA, infA, rpl22, and rps16; and instances of intron and copy deletions observed in genes such as rpl2, clpP, and rps12. These observations highlight the dynamic nature of evolution of the cp. genome across the plant kingdom.

Codon usage frequency analysis

Codon usage analysis is a fundamental tool for understanding the intricacies of genomic structure, evolutionary dynamics, and selective pressures imposed upon genes [47, 48]. The degeneracy inherent in the genetic code is manifested by its capacity to accommodate multiple codons to encode a singular amino acid. Codons that serve as interchangeable vehicles for a single amino acid are designated as synonymous codons. Regarding the three sequenced Capsicum species, the protein-coding genes showed a preference for codons ending in A or T/U. These codons had RSCU values of > 1 (Supplementary Table S2). Similar findings were reported in the cp. genomes of a diverse array of flowering plants, an observation that can be attributed to the prevalent A/T-rich composition inherent to cp. genomes [49,50,51,52]. Similar codon usage patterns are evident in closely related plant species and those from the same lineage, providing insights into plant evolution. Species within the same genus show similar codon usage, even when compared to larger taxonomic groups such as families [53]. Therefore, studying codon usage is valuable for understanding evolutionary relationships within the Capsicum genus.

Evolution of cp. genome of Capsicum species

In flowering plants, recurrent contraction and expansion patterns at the junctions of SSC and LSC with IR regions have been identified as evolutionary indicators which serve to elucidate the relationships among different taxonomic groups [54]. Examination of boundaries within the cp. genome regions revealed distinct positional variations across the genomes of Capsicum species (Fig. 2). Higher variations observed at the extremities of single-copy and inverted repeats become conspicuous when profoundly diverged lineages are compared. In contrast, organisms that have undergone relatively recent evolutionary processes show reduced dissimilarity within vicinities of the boundaries [52, 55]. In the present study, the genus Capsicum exhibited a discernible pattern similar to this fluctuating motif at the junctures of single-copy and IR boundaries. This similarity could potentially signify recent evolutionary schisms among diverse Capsicum species.

Analysis of divergent hot spots

Analysis using mVISTA revealed a notable discrepancy in variation patterns, with the IR regions displaying reduced variability compared with the LSC and SSC regions. Moreover, the non-coding regions exhibited a higher number of variation sites than the coding regions (Fig. 3). This is in accordance with the observed diversity patterns prevalent in most angiosperms [56,57,58]. Among the analyzed genomes, some showed variations in the ycf1 and ycf2 genes, and intragenic regions of accD. Moreover, the results showed that there were more variable regions in the noncoding region than in the coding region of the Capsicum genome. The nucleotide diversity of the cp. genomes of the Capsicum species was approximately 0.00225 (Fig. 4). The Capsicum species showed a slightly lower average Pi value than that reported in a previous study [59]. In addition, four highly variable regions were identified within the cp. genomes of Capsicum species, making them potentially robust DNA barcodes. These regions encompassed intergenic regions (trnT-UGU/trnL-UAA) and four specific genes (ccsA, ndhD, rps12, and ycf1), all of which exhibited Pi values exceeding 0.01. A previous study reported that the trnN–GUU/ndhF intergenic region is a notable polymorphic hotspot, particularly when employing C. lycianthoides (NC_026551) as a reference genome [59]. Similarly, nucleotide polymorphism hotspots in cp. genomes were reported in other Solanaceae species [44, 60]. The genes with mutational sites, specifically ccsA, ndhD, rps12, and ycf1, are likely to play significant roles in facilitating adaptation of species to various environmental conditions [61]. For example, trans-splicing of plastid rps12 transcripts is crucial for key embryonic development processes, such as cotyledon formation, provascular tissue establishment, and shoot apical meristem organization [62]. These mutational hotspots hold promise as potential loci for the development of novel DNA barcodes aimed to classify plants within the Capsicum genus.

Phylogenetic analysis

The cp. genome is highly conserved in sequence and structure, which helps determine molecular homology, consequently rendering it valuable in the construction of plant phylogenies [63]. In line with this, our investigation encompassed phylogenetic analysis of 22 Capsicum genotypes, including 10 distinct Capsicum species, with S. lycopersicum and L. radiata serving as designated outgroups. A dataset comprising 19 Capsicum cp. genomes was acquired from NCBI, and sequence alignment was performed using the MAFFT v7.475 alignment tool [30]. The evolutionary relationships among these chloroplast genomes revealed potential variation, and the categorization of individual species into clades was remarkably consistent with the findings of previous studies [7, 59]. However, the evolutionary relationships among Capsicum species have not been clearly clarified because of a lack of complete molecular data. Only a few cp. genomes have been published, and they do not include comprehensive species for genomic comparison and phylogenetic analyses [59, 64]. Hence, in this study, we selected three wild relatives in conjunction with domesticated Capsicum species to investigate cp. genome variations.

Cp genomic data may serve as an excellent source for highly accurate estimation of divergence time at the population level [60]. Hence, the divergence time for the 10 Capsicum species was estimated employing the GTR + I + G model of the BEAST2.0 program (Fig. 5). Estimation of the divergence time in the MCMCTree likelihood calculation with an uncorrelated relaxed lognormal clock model showed that the Capsicum species diverged from the outgroup species at ~ 25.79 million years ago (Mya). Using different time partitions, divergence time analysis considers variations in how rapidly genes evolve in different lineages [65]. This facilitates the acquisition of more accurate estimates of both the rates and ages of specific points in the evolutionary tree. C. lycianthoides was found to be the common ancestor of all Capsicum species, with an estimated speciation at ~ 20.13 Mya. The C. lycianthoides species was found in the Andean clade and was a sister clade to the rest of the genus as reported in previous studies [7, 66]. This study successfully divided all the Capsicum clades into major lineages, in alignment with previous reports [66], in which, among the scrutinized genotypes, the domesticated Capsicum species stood out as the most extensively divergent group. It has recently been reported that extensive domestication and selective breeding of C. annuum has led to a notably reduced level of nucleotide diversity in contrast to the other two domesticated species, C. frutescens and C. chinense [67].

In a recent time-calibrated phylogenetic analysis of the Solanaceae family, divergence of the Capsicum clade of both the Solanum and Lycianthes clades was estimated to have occurred at approximately 19 and 13 Mya, respectively [68]. However, significant speciation events within Capsicum likely occurred at approximately 1–3 Mya, contributing to contemporary species diversity within the genus. Domestication has led to modification of plant traits by influencing both the morphology and functionality of plants within agroecosystems [69]. Hence, it has been documented that cultivated plants exhibit a greater degree of trait variability than their wild progenitors [70,71,72]. The Phaseolus species (tepary beans) serves as a distinct model for investigating crop evolution; [73] analyzed SNP markers in both wild and domesticated P. acutifolius species and found that species of domesticated tepary beans clustered together, based on their origin. Similarly, [74] found a significant reduction in genetic diversity in lima beans (P. lunatus) owing to domestication. Several researchers have used different molecular markers and nucleotide data to show a reduction in the genetic diversity of various crop species compared to their wild progenitors [15, 75, 76].

Conservation and breeding implications for Pepper

Understanding the genetic variation in pepper species, particularly through cp. genome analysis, provides critical insights for both conservation and breeding efforts [25]. The domestication and intensive breeding of pepper over the past 5,000 years have significantly shaped its genetic landscape. According to the PCA and phylogenetic analysis, the cultivated species/varieties appeared as one structured monophyletic group. The cp. genome variation in pepper species offers valuable insights that are crucial for both conservation and breeding. By leveraging this knowledge, we can ensure the long-term sustainability and productivity of pepper crops, safeguarding them against future challenges while enhancing their value for both producers and consumers.

Conclusions

Our study effectively demonstrates the utility of cp-derived phylogenetic analysis in elucidating species relationships within taxonomically intricate and highly diverse groups, such as Capsicum species. Through comprehensive sequencing of complete cp. genomes (156–157 kbp), we achieved a high-resolution phylogenetic framework with robust support values. Our findings contribute significantly to the existing body of knowledge, emphasizing that the utilization of complete cp. genome sequences enhances phylogenetic resolution of Capsicum species, especially at the infrageneric taxonomic level. This improved resolution has unveiled narrowed levels of genetic variability within Capsicum species with possible genetic bottlenecks, as previously reported [11, 77]. Furthermore, it is evident that selective pressures exerted by early agriculturalists and contemporary crop-breeding practices have led to a notable reduction in the inherent genetic variability of cultivated crops. In conclusion, in the context of evolutionary analysis, it is observed that the early domesticated species of C. annuum exhibit a higher degree of divergence, very similar to their divergent phenotypes, when compared with their counterparts. These findings significantly broaden our understanding of Capsicum domestication and diversification and might play a pivotal role in advancing future research in the fields of evolution and phylogenetics related to Capsicum species.