Wild relatives of domesticated Capsicum represent substantial genetic diversity and thus sources of traits of potential interest. Furthermore, the hybridization compatibility between members of Capsicum species complexes remains unresolved. Improving our understanding of the relationship between Capsicum species relatedness and their ability to form hybrids is a highly pertinent issue. Through the development of novel interspecific hybrids in this study, we demonstrate interspecies compatibility is not necessarily reflected in relatedness according to established Capsicum genepool complexes. Based on a phylogeny constructed by genotyping using simple sequence repeat (SSR) markers and with a portion of the waxy locus, and through principal component analysis (PCA) of phenotypic data, we clarify the relationships among wild and domesticated Capsicum species. Together, the phylogeny and hybridization studies provide evidence for the misidentification of a number of species from the World Vegetable Center genebank included in this study. The World Vegetable Center holds the largest collection of Capsicum genetic material globally, therefore this may reflect a wider issue in the misidentification of Capsicum wild relatives. The findings presented here provide insight into an apparent disconnect between compatibility and relatedness in the Capsicum genus, which will be valuable in identifying candidates for future breeding programs.
Citation: Parry C, Wang Y-W, Lin S-w, Barchenger DW (2021) Reproductive compatibility in Capsicum is not necessarily reflected in genetic or phenotypic similarity between species complexes. PLoS ONE 16(3): e0243689. https://doi.org/10.1371/journal.pone.0243689
Editor: Dengcai Liu, Sichuan Agricultural University at Chengdu, CHINA
Received: November 27, 2020; Accepted: February 17, 2021; Published: March 24, 2021
Copyright: © 2021 Parry et al. This is an open access article distributed under the terms of the Creative Commons Attribution License, which permits unrestricted use, distribution, and reproduction in any medium, provided the original author and source are credited.
Data Availability: The data used in this study are available in the World Vegetable Center repository, HARVEST (DOI: 10.22001/wvc.73914).
Funding: Funding for this research was provided by the Ministry of Science and Technology (MOST) of Taiwan (Project ID:107-2311-B-125 -001 -MY3) and long-term strategic donors to the World Vegetable Center: Taiwan, UK aid from the UK government, United States Agency for International Development (USAID), Australian Centre for International Agricultural Research (ACIAR), Germany, Thailand, Philippines, Korea, and Japan.
Competing interests: The authors have declared that no competing interests exist.
The genus Capsicum (n = 12 or 13) is comprised of about 35 diploid species including five domesticated species: C. annuum L., C. baccatum L., C. chinense Jacq., C. frutescens L., and C. pubescens Ruiz & Pav. . All members of the genus originate in the Americas; however, the crop is produced worldwide with the majority of production occurring in Asia . The genetic and phenotypic diversity across the genus is significant, and thus represents a valuable resource for crop improvement . The primary limitations to improving productivity and quality of Capsicum are abiotic and biotic stresses, many of which lack sources of host tolerance or resistance . Furthermore, as a widely consumed crop with cultural and culinary value across global cuisines, there is high demand for Capsicum . There is therefore significant incentive to overcome challenges to cultivation, one means of doing so being the introgression of resistance to the various stresses that limit production of Capsicum species.
Understanding interspecies compatibility and identifying barriers to hybridization is essential to the design of introgression breeding programs. Capsicum species are divided among 11 clades [4,5] and grouped into three complexes—Annuum, Baccatum and Pubescens—based on their relative reproductive compatibility [6–8]. There is understood to be relatively low reproductive compatibility between species complexes , while unknown mechanisms of unilateral incompatibility have previously been demonstrated . Barriers to hybridization may include failure of the pollen grain to germinate or the pollen tube to develop, or may be post-zygotic: embryo death or inviability, such as that caused by untolerated aneuploidy [11,12]. The pre- and post-zygotic barriers to hybridization between genetic complexes in Capsicum remains largely unresolved , however, a number of cross-complex hybridizations have been achieved [13–18]. This suggests isolation between complexes is not absolute, and there is therefore potential for introgression breeding, or design of genetic bridge strategies in order to best exploit this genetic variation.
In contrast to other Solanaceae crops, including tomato (Solanum lycopersicum L.) , potato (S. tuberosum L.)  and to a lesser extent eggplant (S. melongena L.) , introgression breeding using wild species has been relatively underutilized in Capsicum; . The wild progenitor, C. annuum L. var. glabriusculum (Dunal) Heiser & Pickersgill is a potential source of disease resistance, with reported resistance to Beet curly top virus (BCTV: Curtovirus) [23,24]. Members of the wild species C. chacoense (Hunz.) and C. rhomboideum (Dunal) Kuntze have been identified as being resistant to powdery mildew (Leveillula taurica) . Recently, an accession of C. galapagoense Hunz. has been proposed to be a potential source of resistance to the insect pest, whitefly, based on trichome density and type (M. Rhaka, pers. comm.). However, despite extensive hybridization no successful progeny have so far been developed . These results are surprising because C. galapagoense has been reported as part of the C. annuum clade, and readily hybridize with C. annuum accessions [5,7]. One reason for unsuccessful hybridization attempts may be misidentification; several genebanks have incorrectly reported accessions identified as C. galapagoense which are, in fact, C. frutescens (P.W. Bosland, pers. comm.). Such misidentification presents a challenge to utilizing knowledge of the relatedness of Capsicum species and their ability to hybridize. Although the genetic diversity and variation within wild populations of Capsicum has been studied [5,27–31], the pool of phenotypic data for wild Capsicum species remains limited . There also remains a lack of access to publicly available germplasm representing the diversity of wild Capsicum . There is therefore an immediate need to better understand the role of wild Capsicum species in future breeding programs.
The objectives of this study were to elucidate the relationship between interspecies compatibility and relatedness through extensive interspecific hybridization and the construction of a phylogeny. We aimed to clarify the relationships among the wild and domesticated Capsicum species included in the study, and confirm the identities of several World Vegetable Center genebank accessions.
Materials and methods
Thirty-eight accessions of 15 species of Capsicum were chosen for this experiment (Table 1). Most of the accessions in our experiment have been previously karyotyped and have 12 chromosomes, with the exceptions of C. eshbaughii Barboza (n = unknown), C. minutifolium (Rusby) Hunz. (n = unknown) and C. rhomboideum (n = 13). The accessions were provided to the World Vegetable Center, having been collected from diverse locations and deposited into collections at either the World Vegetable Center Genebank, the World Vegetable Center Pepper Breeding Collection in Tainan, Taiwan, the United States Department of Agriculture—Agriculture Research Service National Plant Germplasm System, or the Chile Pepper Institute, New Mexico State University, Las Cruces, NM USA. Of each accession, two biological replications were used wherever possible for phenotyping and genotyping, although due to poor germination, four accessions (NMCA50034, PBC 556, PBC 1892, NMCA50064) did not have a biological replicate.
All experiments were conducted at the World Vegetable Center, Shanhua, Tainan, Taiwan (lat. 23.1°N; long. 120.3°E; elevation 12 m). Prior to sowing, all seed was treated with trisodium phosphate (TSP) and hydrochloric acid (HCl) following the methods of Kenyon et al. , which has been observed to reduce germination rates. Seeds were sown into 72-cell plastic trays of sterilized peat moss. Trays were placed in a climate-controlled greenhouse for germination at 28 ± 3°C with a 12-hour photoperiod and ≈95% relative humidity. At the 4–6 true leaf stage, the seedlings were transplanted into pots and moved to a greenhouse without climate control. Plants were irrigated twice daily and regularly fertilized with Nitrophoska (Incitec Pivot Fertilisers, Victoria, Australia) during the experimental period.
The accessions were morphologically characterized according to the Descriptors of Capsicum Manual  for the following characteristics: mature leaf length, mature leaf width at widest point, leaf color, density (if present) of leaf pubescence, leaf shape, lamina margin, stem color, stem shape, density (if present) of stem pubescence, nodal anthocyanin color, node length, anther color, anther length, filament length, corolla color, corolla spot color, corolla shape, corolla length, stigma exsertion, flower position, tillering, leaf density, fruit length, fruit width, fruit pedicel length, neck at base of fruit. Quantitative traits were the mean of 10 values measured across replicates. Qualitative traits were scored according to the IPGRI Descriptors of Capsicum manual  based on observations of both plant replicates. Accessions with incomplete data were excluded from analysis of phenotypic data. To identify trends in traits between species, the quantitative traits were used for principal component analysis (PCA) using the R packages, ‘factoextra’  and ‘ggfortify’  for PCA analysis with scaling. The scores of qualitative traits were analyzed using an unweighted pair group method with arithmetic mean (UPGMA) hierarchical cluster analysis. Bootstrap resampling was applied to clustering with 1,000 iterations.
Reciprocal hybridizations were attempted among all combinations of accessions throughout the experimental period. Ability to hybridize in reciprocal was used to confirm previous reports of relatedness and ability to hybridize species across clades and complexes. The fruits of successful hybridizations were collected upon ripening. Within three days of harvest, the seeds were extracted from the fruits and dried for at least 1 week. Five seeds each of 112 crosses of interest were sown into 72-cell plastic trays containing sterilized peat moss. The trays were placed in a greenhouse without climate control and irrigated twice daily and observed daily for 12 weeks to assess germination. A chord diagram was produced in R using the package ‘circlize’  to visualize successful crosses for which seed was obtained. A heat map was produced in R using the package ‘ to visualize the percentage of seeds germinated after 12 weeks.
For genotyping, DNA was isolated from young, actively growing leaves from plants of each accession using the modified cetrimonium bromide (CTAB) extraction method . Using 27 Simple Sequence Repeat (SSR) markers, DNA was amplified by PCR, for which each well of a 96-well microtiter plate contained 2 μl of template DNA, 0.4 μl of primer (0.2 μl each forward and reverse), 0.1 μl of AmpliTaq Gold DNA polymerase, 0.4 μl of deoxyribonucleotides, 1.5 μl of 10× PCR Buffer II Gold buffer (Thermo Fisher Scientific, Waltham, MA, USA), and sterile water to a final volume of 15 μl. The reactions were carried out in a thermal cycler (Single Block Alpha Unit, DNA Engine®, Bio-Rad Laboratories, Berkeley, CA, USA) with an annealing temperature of 55°C. The electrophoresis of amplified products was performed on 6% acrylamide gels at 160 Volts for 30 minutes (Thermo Electron Electrophoresis EC250-90, Thermo Fisher Scientific). The results were visualized under UV light using UVITEC Imaging Systems (Cleaver Scientific, Warwickshire, UK) following staining with ethidium bromide. Electrophoresis was repeated whenever the clarity of the bands or their exact size was uncertain.
Gels were scored for each primer pair using a binary method: each accession was scored for presence (1) or absence (0) of amplicons of each size. The data were processed in R using the packages, ‘proxy’ and ‘shipunov’  to produce a dendrogram with bootstrapping for the assessment of the relatedness between the individual accessions. A distance matrix was produced using the Dice index, and an unweighted pair group method with arithmetic mean (UPGMA) hierarchical cluster analysis was carried out. Bootstrap resampling was applied to clustering with 1,000 iterations.
Further molecular analysis to clarify the identification of some accessions included the study of the waxy gene region of six accessions, VI051012 (C. tovarii); VI051011 (C. galapagoense, potentially C. annuum); VI012574 (C. chacoense, potentially C. annuum); PBC 1892 (C. galapagoense); VI013161 (C. eximium Hunz); and PBC 556 (C. frutescens), using the primer pair, 860F and 2R . The chosen accessions were those expected to need clarification due to possible misidentification, based on molecular and morphological data. The waxy region was amplified by PCR as before, with an annealing temperature of 60°C. The quality of the products were evaluated by running on a 2% agarose gel with EtB‘out’ (Yeastern Biotech Co. Ltd., Taipei, Taiwan) at 100 Volts for 50 minutes, then visualized using a Microtek Bio- 1000F gel imager (Microtek International Inc., Hsinchu, Taiwan). The PCR products were sequenced by Genomics Biotechnology Co., Ltd. (New Taipei City, Taiwan) by the Sanger sequencing method. Low-quality nucleotides were manually removed throughout the resulting sequence, including approximately the first and last 60 nucleotides. The sequences were aligned using NCBI nucleotide BLAST  and a consensus sequence constructed using the CAP contig assembly program from BioEdit . The sequences, including that of the publicly available S. lycopersicum GBSS sequence (gene ID: 101259777) as the outgroup, and the waxy sequences of 15 Capsicum species deposited in NCBI (accession numbers: KP747352.1, KP747351.1, KP747358.1, KP747354.1, KP747353.1, KP747360.1, KP747310.1, KP747309.1, KP747359.1, KP747314.1, KP747306.1, KP747357.1, KP747311.1, KP747320.1, KP747361.1) (National Center for Biotechnology Information (NCBI) , were aligned using multiple sequence alignment tool, Clustal MAFFT . The resulting dendrogram was visualized using Interactive Tree of Life (iTOL) version 5.7 .
To clarify the phylogeny of the wild and domesticated Capsicum species in the sample, UPGMA clustering was applied to the genetic variation captured by the SSR molecular markers. The C. baccatum and C. chinense group accessions were distinct from the C. annuum group, with 76% bootstrap support (Fig 1). The C. baccatum accessions made up a significant group, being closely clustered with the C. praetermissium Heiser & P.G. Sm. accessions. This grouping was adjacent to a large group comprised of closely clustered C. chinense accessions with the C. galapagoense accession PBC 1892, as well as C. eshbaughii, C. eximium and C. frutescens, separated from the C. baccatum group with a relatively low confidence interval. Within this grouping, C. chinense accession, PBC 1820, was distinct from its counterparts, with 92% bootstrap support. Furthermore, the C. chinense species accessions were relatively separate from the accessions at the periphery of this grouping, the C. galapagoense accession PCB 1892, the C. frutescens accession PBC 556, and the C. eximium accession VI013161. The grouping of the C. eximium accession VI013161 with C. frutescens was similar in clustering from the waxy gene sequence (Fig 2). These accessions were thus more similar to each other than they were similar to the C. galapagoense accession PBC 1892, and this was a distinct grouping from other sequenced accessions.
‘Height’ represents dissimilarity, derived from ‘dice method’. Bootstrap resampling applied to clusters, represented as percent confidence interval. Numbers following the hyphen indicate replicates.
Sequences of accessions that begin with “KP” were obtained from NCBI, while those that begin with “PBC” or “VI” were from this experiment. The waxy sequence (gene ID: 101259777) of tomato (Solanum lycopersicum) was used as the root of the tree.
To provide further evidence of this phylogeny, we analyzed the waxy gene sequence of a sample of accessions in this study (VI051012, VI051011, VI012574, VI013161, and PBC 1892), that of a range of Capsicum species, and S. lycopersicum, available publicly (Fig 2). In this analysis, C. baccatum accessions were similarly clustered closely with C. pubescens, and with C. chacoense.
To better understand the crossing relationship between species, reciprocal hybridizations were performed between each combination of accessions (Fig 3). Members of C. baccatum and C. praetermissium hybridized as either the female or male parent with at least one accession of each other species with the exception of C. rhomboideum (Fig 3). However, of the sample of seeds selected for sowing, only the cross between VI014924 and PBC 1969 germinated (Fig 4). Hybridizations were not achieved between C. galapagoense as either parent with accessions of C. tovarii, C. flexuosum, C. minutiflorium, C. cardenasii, C. eshbaughii, and C. rhomboideum species. Capsicum eshbaughii hybridized more readily as the female parent, but failed to hybridize in either direction with accessions of C. eximium, C. frutescens, C. galapagoense, C. tovarii, C. flexuosum, and C. rhomboideum. The majority of C. frutescens hybrids were achieved with C. annuum accessions, but successful hybridizations were found across a broad species range. Of the sample of seeds sown, 80% of the PBC 556 × PBC 1970 cross seeds germinated (Fig 4).
Direction of arrow represents successful hybridizations in the male-female direction from which fruit was harvested.
The C. annuum group, which was adjacent to C. baccatum, consisted of the closely clustered C. annuum species accessions: PBC 1799, AVPP9905, PBC 1899, PBC 1867, and PBC 196, along with C. annuum var. glabriusculum PI 574547, and C. chacoense VI012574 (Fig 1). Neighboring this group was a cluster comprised of the C. eximium accession VI012964, the C. eshbaughii accession NMCA90006, the C. annuum accessions PBC 142 and VI029657, the C. annuum var. glabriusculum accession PI 674459, the C. tovarii Eshbaugh et al. accession VI051012, and the C. galapagoense accession VI051011. Based on clustering of the waxy gene sequence, we found C. tovarii accession VI051012 to be clustered broadly with C. chacoense, C. galapagoensis, C. eximium and C. eshbaughii, and with two other C. tovarii accessions (Fig 2). The C. galapagoense accession VI051011 was distinct from its counterpart, PBC 1892, and another C. galapagensis accession, KP747306.1 (Fig 2).
Accessions of C. annuum hybridized in both directions with one or more accessions of all species except C. rhomboideum and C. minutifolium (Fig 3). Thirteen of the hybrids sown germinated well (Fig 4). Capsicum annuum var. glabriusculum hybridized in either direction with at least one accession of every species except C. rhomboideum, and 7 out of 10 of those sown germinated (Fig 4). Accessions of C. chacoense also hybridized broadly, but not with C. frutescens × chinense, C. minutifolium or C. rhomboideum, and seven of the 25 hybrids sown germinated. More than one cross was achieved between C. tovarii and an accession of every species except C. eshbaughii, C. eximium, C. galapagoense, C. minutifolium and C. rhomboideum. Of these crosses, VI012574 × VI051012 and NMCA90030 × VI051012 germinated with 40% and 100% efficiency, respectively.
With 80% confidence interval, C. chacoense accession VI012900 and C. cardenasii Heiser & P.G. Sm. accession NMCA90030 were clustered separately from the C. baccatum and C. annuum groups (Fig 1). NMCA90035 clustered distinctly from its C. cardenasii counterpart, with bootstrap support of 57%. The C. flexuosum Sendtn. accessions NMCA50034 and NMCA50030 clustered closely together, with high bootstrap support (99%); adjacent was the C. minutifolium accession NMCA50053, and in a separate cluster, the C. rhomboideium accession NMCA50064, which was the most distinct grouping, separated from its neighbors with 100% confidence. This was supported by waxy sequences which showed C. rhomboideum to be most similar to the outgroup, S. lycopersicum, and more broadly clustered with C. flexuosum (Fig 2).
Capsicum cardenasii species hybridized with every species except C. frutescens, C. flexuosum, C. galapagoense and C. rhomboideum (Fig 3), and the five of the nine hybrids sown germinated well (Fig 4). No hybrids were achieved with C. flexuosum or C. minutifolium as female parents, however C. flexuosum hybridized as the male parent with C. chacoense, C. annuum var. glabriusculum, C. baccatum, C. tovarii, C. annuum and C. frutescens, while C. minutifolium hybridized with C. cardenasii, C. eshbaughii, and C. annuum var. glabriusculum, and of the crosses sown, only VI012574 × PBC 124 germinated (Fig 4). No successful hybrids were achieved with C. rhomboideum in either direction (Fig 3).
We applied principal component analysis to the quantitative phenotypic data collected to understand phenotype across Capsicum species (Fig 5). The first two components account for 59.9% of the total variation. The C. baccatum accessions made up a group along with PBC 196 and VI01668, due to their correlated fruit and flower characteristics (pedicel length, fruit width, fruit length, anther length, filament length, corolla length) (Fig 5). Capsicum annuum accessions made up a less distinct group, along with the wild progenitor C. annuum glabriusculum, and the other domesticated species C. chinense, C. frutescens, C. frutescens × chinense along with C. eximium (Fig 5).
The UPGMA clustering of the accessions’ qualitative phenotypic data more closely mirrored the genetic relatedness based on SSR molecular markers, especially for the C. baccatum and C. praetermissum accessions (Fig 6), which formed two groupings (VI012528, VI029696, VI029697; and PBC 80, VI014924, PBC 81). However, we found C. annuum did not form a unique clade, highlighting the phenotypic diversity of this domesticated species (Fig 6).
Understanding the relatedness between accessions of Capsicum species, and the extent to which they hybridize is key in identifying candidates for the introgression of traits of interest into commercial varieties. Our results mirror the widely accepted species phylogeny; clustering based on genotyping is centered around C. annuum, C. baccatum and C. chinense complexes [4,6,7]. Similarly, our results of phenotyping the accessions also support previously described genetic complexes . Generally, accessions from the domesticated species (C. annuum, C. chinense, and C. frutescens) were nearer the origin of the score plot, with the exception of members of the domesticated C. baccatum, which clustered further away from the other domesticated species (Fig 5). Conversely, members of the wild species were further from the origin, indicating greater diversity (Fig 5). Although we measured different phenotypic traits, our findings contradict those of Luna-Ruiz et al. , who found greater levels of diversity among domesticated species for capsaicinoids. Interestingly, based on hybridization success rates, there was a weak relationship between relatedness and crossability, which is in contrast with previous understanding that compatibility between complexes is low . This suggests potential for crop improvement with wild relatives of domesticated species using genetic bridge strategies.
Genotyping using SSR markers evenly distributed across the genome provides evidence of the level of relatedness between wild and domesticated species, and has become a valuable tool for this purpose in many species [46–53]. The use of SSR markers is particularly useful when little is known about the species in question, as in the study of wild species of Capsicum, which are relatively poorly understood. We have supplemented genotyping using SSR markers with targeted sequences of the waxy gene. The sequence of this single-copy nuclear gene encoding the granule-bound starch synthase (GBSS, also known as waxy) protein has been previously utilized in elucidating phylogenies in Capsicum [5,54,55], and has proven useful in understanding interspecies relationships here.
When phylogeny, interspecific compatibility and phenotype are considered in concert, the identity of a number of accessions included in this study may be questioned. The issue of misidentification of Capsicum species has been raised previously, with several genebanks incorrectly reporting accessions C. frutescens as C. galapagoense (P.W. Bosland, pers. comm.). Thorough characterization is important in supporting conservation of genetic material and identifying gaps in genebank collections . Only 12% of national vegetable germplasm collections have been characterized biochemically, while 65% have been characterized morphologically . Thorough characterization is therefore key in understanding the reproductive relationships between Capsicum species.
Clustering based on SSR genotyping revealed a close relationship between C. baccatum accessions (Fig 1), as expected for this well-accepted domesticated species. Capsicum praetermissium accessions are also grouped within this complex, shown by both SSR and waxy genotyping, which supports previous findings  and suggestions that C. praetermissium in fact comprises a subgroup of C. baccatum . Capsicum praetermissium is thought to have diverged prior to domestication of C. baccatum, but has not yet been utilized in breeding domestic C. baccatum accessions . We found C. praetermissium readily hybridized with C. baccatum (Fig 3) in line with the findings of Emboden Jr. , and thus offers potential as a genetic resource.
Capsicum chinense species accessions comprised a significant cluster, which included C. chinense, C. frutescens, C. eshbaughii, and C. galapagoense (Fig 1). The grouping of C. chinense adjacent to C. baccatum was in line with a recent study that also used SSR molecular markers to characterize Capsicum species . Conversely, our analysis of publicly available waxy sequences found C. chinense to be grouped with C. annuum var. glabriusculum, supporting findings of Pickersgill et al.  and Ince et al. , who grouped C. chinense within the C. annuum complex. Furthermore, in this study, a total of 20 crosses were achieved between C. annuum (including the wild progenitor C. annuum var. glabriusculum), and C. chinense, 13 of which had a C. chinense female parent (Fig 3). Seeds from two of these crosses were sown (VI029446 × PBC 1969 and PI 152225 × NMCA90030) and germinated well (Fig 4). This contrasts to previous work that reports a barrier to reproduction between C. annuum and C. chinense . However, Costa et al.  found that crosses between C. chinense and C. annuum accession were possible. These findings highlight the genetic variation that exists in Capsicum species, as well as the variability in compatibility, and its dependence on accession selection.
The grouping of C. frutescens in the C. chinense complex (Fig 1) was in line with previous findings of the close relationship of these species . A number of researchers argued their identities as sister species within the annuum clade [57,63] including Walsh and Hoot , who similarly used the waxy gene sequence in order to delineate phylogenetic relationships among Capsicum species. Furthermore, we found C. frutescens hybridized readily with both members of the C. baccatum and C. annuum clades, as well as with C. chinense (Fig 3). Of the three C. frutescens hybrids selected for sowing, 80% of the PBC 556 × PBC 1970 hybrid seeds germinated (Fig 4). The relationship of C. eshbaughii to this clade, and its pairing with C. eximium both in SSR and waxy genotyping (Figs 1 and 2) was consistent with its previous placement in the ‘Purple Corolla clade’ . Furthermore, this was supported by Carrizo Garcia et al  and Walsh and Hoot , whose use of waxy gene sequencing demonstrated C. eximium as a divergent species, distinct from C. annuum. Capsicum chinense formed hybrids with other members of this grouping (Fig 3), and 100% of seeds from the C. chinense and C. eximium cross germinated (Fig 4).
Interestingly, the C. eximium and C. cardenasii accessions in our study appeared distantly related (Fig 1). This contradicts the relationship seen between accessions of these species in waxy sequencing, and previous reports of these species as members of the C. pubescens complex [64,65]. Furthermore, their phenotypes correlated closely with C. annuum accessions (Fig 5). This raises the question of the validity of the identification of accessions VI013161 and VI012964 as C. eximium.
The C. annuum accessions comprise a major grouping adjacent to the C. baccatum group (Fig 1). A sample of C. annuum accessions (PBC 1799, PBC 196, PBC 1867 and AVPP9905) formed a tightly clustered group, indicating genetic similarity. They also display highly correlated phenotypes, forming a cluster along with accessions from other domesticated species (Fig 5). The C. annuum accessions PBC 142 and VI029657 were in an adjacent group (Fig 1), therefore may be considered part of the wider C. annuum complex, along with C. chacoense accession VI012574, C. galapagoense, VI051011, and C. tovarii accession VI051012. The presence of C. chacoense (VI012574) in this group, distant from the second C. chacoense accession (VI012900) included in this study, highlights its possible misidentification. Sequencing clustered VI012574 closely with C. galapagoense accession VI051011, which may be considered a member of the C. annuum complex (Fig 2). Principal component analysis (Fig 5) revealed VI012574 was grouped with C. annuum accessions, away from its counterpart, while UPGMA analysis further highlights this disparity. Direct observation of the phenotypes emphasizes the similarity between the morphology of VI012574 and typical C. annuum features. This includes upright growth, elongated fruits, and relatively large flowers with blue anthers.
The C. galapagoense accession, PBC 1892 was grouped with the wider C. baccatum cluster (Fig 1), conflicting previous findings that C. galapagoense is derived from a C. annuum progenitor population . No successful hybridizations were achieved between PBC 1892 and PBC 556 (C. frutescens), which clustering suggested were closely related. The second C. galapagoense accession included in the study, VI051011, was distant from PBC 1892 in waxy sequence (Fig 2), and grouped within the C. annuum complex in the SSR analysis (Fig 1). It also displayed a distinctly different phenotype to that of PBC 1892; PBC 1892 had a compact growth habit, very small fruits, flowers and leaves, and densely pubescent stems and leaves, typical of C. galapagoense descriptions. Conversely, VI051011 had a morphology similar to that of C. annuum, reflected in its close proximity to the PCA origin, along with C. annuum accessions. Eight hybridizations were achieved between VI051011 and C. annuum accessions, and of the selected hybrid seeds sown, 20% germinated. The close clustering of VI051011 with the C. annuum complex, their similar morphology and their ability to hybridize suggests likely misidentification of this accession.
There were five further clusters consisting of C. chacoense, C. cardenasii, C. flexuosum, C. minutifolium, and C. rhomboideum respectively, which had increasingly distant relation to the three major species complexes (Fig 1). Although C. chacoense has been previously grouped within the C. baccatum clade [5,64], this wild species has an apparently distant relationship with C. baccatum, supported by analysis of waxy sequencing. Capsicum cardenasii was similarly distantly related to other clades. Other studies [60,64,65] also found C. chacoense and C. cardenasii not to be closely related to any major clade. Furthermore, VI012900 hybridized readily with members of both C. annuum and C. baccatum clades (Fig 3). Both C. chacoense and C. cardenasii accessions (with the exception of PI 159236 and PI 15225) lay on the periphery of the PCA plot, clustering with neither C. baccatum or C. annum groups. This suggests C. cardenasii and C. chacoense accessions are not members of either C. baccatum or C. annuum clades. In their waxy sequence analysis, Walsh and hoot  similarly demonstrated the distinction of C. chacoense from either C. annuum or C. baccatum groups.
Capsicum flexuosum, C. minutifolium and C. rhomboideum were distantly related to the major clades in analysis of both waxy sequence and SSR data (Figs 1 and 2), consistent with the body of literature [5,59,66]. The C. flexuosum accession (NMCA50034) was also distinct in phenotype from other accessions (Figs 4 and 5). A small number of hybridizations were achieved between C. flexuosum and C. minutifolium with members of both C. annuum and C. baccatum clades. However, no hybridizations were achieved between C. rhomboideum and any other accession. This finding is supported by the sequence dissimilarity of the waxy gene obtained from NCBI, where C. rhomboideum clustered with the tomato outgroup and not the other Capsicum species (Fig 2). This low success rate of hybridization with a C. rhomboideum parent is likely caused by differences in chromosome number, resulting in abnormal chromosomal pairing and disrupting meiosis; however, more studies are needed to confirm this.
The results reported here highlight the extent of phenotypic diversity in Capsicum species, the complexity of Capsicum phylogeny, and the similarly complex reproductive relationships between Capsicum species. The evidence suggesting the incorrect identification of VI013161, VI012964, VI012574, and VI051011 may highlight a broader issue of misidentification of Capsicum in genebanks. Thorough characterization of Capsicum genetic material taking a multifaceted approach is therefore important for the development of future breeding programs. Furthermore, the generation of diverse hybrids among accessions of all species included in this study (with the exception of C. rhomboideum) demonstrates the possibility for introgression of a diverse range of traits of interest directly or through the design of bridge crossing strategies. Wild relatives of domesticated Capsicum species therefore represent significant potential for future breeding programs, and should not be discounted on the basis of their assumed relatedness to domesticated species.
We thank Dr. Paul Bosland of Chile Pepper Institute, New Mexico State University, USA for providing Capsicum accessions.
- 1. Khoury CK, Carver D, Barchenger DW, Barboza GE, van Zonneveld M, Jarret R, et al. Modelled distributions and conservation status of the wild relatives of chile peppers (Capsicum L.). Divers Distrib. 2020;26(2):209–225. https://doi.org/10.1111/ddi.13008.
- 2. Barchenger DW, Bosland PW. Wild chile pepper (Capsicum L.) of North America. In: Greene S, Williams K, Khoury C, Kantar M, Marek L, editors. North American Crop Wild Relatives, New York: Cham: Springer International Publishing; 2019; p. 225–42.
- 3. Barchenger DW, Naresh P, Kumar S. Genetic resources of Capsicum. In: Ramchiary N, Kole C, editors. The Capsicum Genome. New York: Springer Nature; 2019; p. 9–23.
- 4. Bosland PW, Votava EJ. Peppers: Vegetable and spice capsicums. 2nd ed. 2012, Wallingford. CABI.
- 5. Carrizo García C, Barfuss MHJ, Sehr EM, Barboza GE, Samuel R, Moscone EA, et al. Phylogenetic relationships, diversification and expansion of chili peppers (Capsicum, Solanaceae). Ann Bot, 2016;118(1):35–51. https://doi.org/10.1093/aob/mcw079.
- 6. Emboden WA Jr. A preliminary study of the crossing relationships of Capsicum baccatum. Butler Univ Bot Stud. 1962;14:108–114. https://doi.org/10.1080/00231940.1962.11757638.
- 7. Pickersgill B. Relationships between weedy and cultivated forms in some species of chili peppers (genus Capsicum). Evol. 1971;25(4):683–691. https://doi.org/10.2307/2406949.
- 8. Tong N, Bosland PW. Capsicum tovarii, a new member of the Capsicum baccatum complex. Euphytica. 1999;109(2):71–77. https://doi.org/10.1023/A:1003421217077.
- 9. van Zonneveld M, Ramirez M, Williams DE, Pretz M, Meckelmann SW, Avila T, et al. Screening genetic resources of Capsicum peppers in their primary center of diversity in Bolivia and Peru. PLoS ONE, 2015;10(9). pmid:26402618
- 10. Onus and Pickersgill. Unilateral Incompatibility in Capsicum (Solanaceae): Occurrence and Taxonomic Distribution. Ann Bot. 2004;94:289–295. pmid:15229125
- 11. da Silva Moneiro CE, Santana Pereira TN, de Campos KP. Reproductive characterization of interspecific hybrids among Capsicum species. Crop Breed Appl Biotechnol. 2011:11(3). https://doi.org/10.1590/S1984-70332011000300006.
- 12. de Souza Macedo V, García Dávila MA, de Castro GR, Garzón Bautista YM, Maria Caetano C. Cytogenetic evaluation of chili (Capsicum spp., Solanaceae) genotypes cultivated in Valle del Cauca, Colombia. 2017:66(4)612–617. https://doi.org/10.15446/acag.v66n4.59162.
- 13. Yoon BJ, Yang DC, Do JW, Park GP. Overcoming two post-fertilization genetic barriers in interspecific hybridization between Capsicum annuum and C. baccatum for introgression of anthracnose resistance. Breed Sci., 2006;56(1):31–38. https://doi.org/10.1270/jsbbs.56.31.
- 14. Pickersgill B. Cytogenetics and evolution of Capsicum L. Chromosome engineering in plants: genetics, breeding, evolution, part B. Amsterdam: Elsevier; 1991.
- 15. OECD. Consensus document on the biology of the Capsicum annuum complex (chili peppers, hot peppers and sweet peppers). Paris (France): OECD; 2006.
- 16. Costa LV, Lopes R, Lopes MTG, de Figueiredo AF, Barros WS, Alves RRM. Cross compatibility of domesticated hot pepper and cultivated sweet pepper. Crop Breed Appl Biotechnol. 2009;9:37–44.
- 17. Eggink PM, Tikunov Y, Maliepaard C, Haanstra JP, de Rooij H, Vogelaar A, et al. Capturing flavors from Capsicum baccatum by introgression in sweet pepper. Theor Appl Genet. 2014;127(2):373–390. pmid:24185820
- 18. Kamvorn W, Techawongstien S, Techawongstien S, Theerakulpisut P. Compatibility of inter-specific crosses between Capsicum chinense Jacq. and Capsicum baccatum L. at different fertilization stages. Sci Hortic. 2014;179:9–15. https://doi.org/10.1016/j.scienta.2014.09.003.
- 19. Lin T, Zhu G, Zhang J, Xu X, Yu Q, Zheng Z, et al. Genomic analyses provide insights into the history of tomato breeding. Nat Genet. 2014;46(11):1220–1226. pmid:25305757
- 20. Hirsch CN, Hirsch CD, Felcher K, Coombs J, Zarka D, van Deynze A, et al. Retrospective view of North American potato (Solanum tuberosum L.) breeding in the 20th and 21st centuries. G3-Genes, Genom, Genet. 2013;3(6):1003–1013. pmid:23589519
- 21. Gramazio P, Prohens J, Plazas M, Mangino G, Herraiz FJ, Vilanova S. Development and genetic characterization of advanced backcross materials and an introgression line population of Solanum incanum in a S. melongena background. Frontiers Plant Sci. 2017;8:1477.
- 22. Mongkolporn O, Taylor PJW. Capsicum. In: Kole C., editor. Wild crop relatives: genomic and breeding resources. Heidelberg, Berlin: Springer; 2011. p. 43–57. https://doi.org/10.1007/978-3-642-20450-0_4.
- 23. Bosland PW. Sources of curly top virus resistance in Capsicum. HortScience 2000;35(7):1321–2.
- 24. ]Jimenez RC. Utilizing Wild Capsicum annuum germplasm for breeding resistance to beet curly top virus (genus: Curtovirus, family: Geminiviridae) in Cultivated Pepper (Capsicum annuum L.) [PhD Thesis]. University of California, Davis; 2019.
- 25. McCoy JW, Bosland PW. Identification of resistance to powdery mildew in chile pepper. HortScience 2019;54:4–7.
- 26. Lin TH, Lin SW, Wang YW, van Zonneveld M, Barchenger DW. Environmental influence on inter- and intraspecific crossability and self-pollination compared to heat treatment of wild and domesticated Capsicum species. HortScience 2020;55(9):S109–S110.(abstr).
- 27. Loaiza-Figueroa F, Ritland K, Laborde Cancino JA, Tanksley SD. Patterns of genetic variation of the genus Capsicum (Solanaceae) in Mexico. Plant Syst Evol. 1989;165(3–4):159–188. https://doi.org/10.1007/BF00936000.
- 28. Votava EJ, Nabhan GP, Bosland PW. Genetic diversity and similarity revealed via molecular analysis among and within an in situ population and ex situ accessions of chiltepín (Capsicum annuum var. glabriusculum). Conserv Genet. 2002;3(2):123–129. https://doi.org/10.1023/A:1015216504565.
- 29. Oyama K, Hernández-Verdugo S, Sánchez C, González-Rodríguez A, Sánchez-Peña P, Garzón-Tiznado JA, et al. Genetic structure of wild and domesticated populations of Capsicum annuum (Solanaceae) from northwestern Mexico analyzed by RAPDs. Genet Resour Crop Evol. 2006;53(3):553–562. https://doi.org/10.1007/s10722-004-2363-1.
- 30. Aguilar Meléndez A, Morrell PL, Roose ML, Kim SC. Genetic diversity and structure in semiwild and domesticated chiles (Capsicum annuum; Solanaceae) from Mexico. Amer J Bot. 2009;96(6):1190–1202. pmid:21628269
- 31. Cheng J, Qin C, Tang X, Zhou H, Hu Y, Zhao Z, et al. Development of a SNP array and its application to genetic mapping and diversity assessment in pepper (Capsicum spp.). Sci Rep, 2016;6:1–11 pmid:28442746
- 32. Kenyon L, Hanson P, Kumar S, Shih SL, Hsieh MH, Chen HY, et al. Treatment for cleaning small seed lots of tomato and pepper seeds of surface contamination with viroids. The World Vegetable Center. 2017 Available from: https://worldveg.tind.io/record/74053?ln=en.
- 33. IPGRI, AVRDC and CATIE. Descriptors for Capsicum (Capsicum spp.). International Plant Genetic Resources Institute, Rome, Italy; the Asian Vegetable Research and Development Center, Taipei, Taiwan, and the Centro Agronómico Enseñanza, Turrialba, Costa Rica; 1995.
- 34. Kassambara A, Mundt F. factoextra: extract and visualise the results of multivariate data analyses. R package version 1.0.7.; 2020.
- 35. Horikoshi M., Tang Y, Dickey A, Genié M, Thompson R, Seltzer L, et al. ggfortify: Data visualization tools for statistical analysis results. R package version 0.4.11. 2020.
- 36. Gu Z, Gu L, Eils R, Schlesner M, Brors B. circlize Implements and enhances circular visualisation in R. Bioinformatics. 2014;30(19):2811–2812 pmid:24930139
- 37. Wickham H, Chang W, Henry L, Pedersen TL, Takahashi K, Wilke C, et al. ggplot2: Create elegant data visualisations using the grammar of graphics. R package version 3.3.2.; 2020.
- 38. Meyer D, Buchta C. proxy: Distance and similarity measures. R package version 0.4–24.; 2020 https://CRAN.R-project.org/package=proxy.
- 39. Shipunov A, Murrell P, D’Orazio M, Turner S, Altshuler E, Rau R, et al. shipunov: miscellaneous functions from Alexey Shipunov. R package version 1.12; 2020. https://cran.r-project.org/web/packages/shipunov/index.html.
- 40. Altschul SF, Gish W, Miller W, Myers EW, Lipman DJ. Basic local alignment search tool. J Molec Biol. 2009;215(3)403–10. https://doi.org/10.1016/S0022-2836(05)80360-2.
- 41. Hall TA. Bioedit: a used-friendly biological sequence alignment editor and analysis program for Windows 95/98/NT. Nucleic Acids Symp Ser. 1999;4:195–98 https://doi.org/10.14601/PHYTOPATHOL_MEDITERR-14998U1.29.
- 42. National Center for Biotechology Information (NCBI) [Internet]. Bethesda 9MD): National Library of Medicine (US), National Center for Biotechnology Information; 1988 [cited 2021 Feb 05]. https://www.ncbi.nlm.nih.gov.
- 43. Katoh K, Standley DM. MAFFT multiple sequence alignment software version 7: improvements in performance and usability. Mol Biol Evol. 2013;30(4):772–780. pmid:23329690
- 44. Letunic I, Bork P. Interactive Tree Of Life (iTOL): an online tool for phylogenetic tree display and annotation. J. Bioinform. 2006;23(1):127–128. pmid:17050570
- 45. Luna-Ruiz J de J, Nabhan GP, Aguilar-Meléndez A. Shifts in plant chemical defenses of chile pepper (Capsicum annuum L.) due to domestication in Mesoamerica. Frontiers Ecol Evol. 2018;6:48 https://doi.org/10.3389/fevo.2018.00048.
- 46. Li YH, Li W, Zhang C, Yang L, Chang RZ, Gaut BS, et al. Genetic diversity in domesticated soybean (Glycine max) and its wild progenitor (Glycine soja) for simple sequence repeat and single‐nucleotide polymorphism loci. New Phytol. 2010;188: 242–253. pmid:20618914
- 47. Nicolaï M, Cantet M, Lefebvre V, Sage-Palloix AM, Palloix A. Genotyping a large collection of pepper (Capsicum spp.) with SSR loci brings new evidence for the wild origin of cultivated C. annuum and the structuring of genetic diversity by human selection of cultivar types. Genet Resour Crop Evol. 2013;60:2375–2390.
- 48. Rai VP, Kumar R, Kumar S, Rai A, Kumar S, Singh M, et al. Genetic diversity in Capsicum germplasm based on microsatellite and random amplified microsatellite polymorphism markers. Physiol. Mol. Biol. Plants. 2013;19:575–586. pmid:24431527
- 49. Olango TM, Tesfaye B, Oagnotta MA, Pè ME, Catellani M. Development of SSR markers and genetic diversity analysis in enset (Ensete ventricosum (Welw.) Cheesman), an orphan food security crop from Southern Ethiopia. BMC Genet. 2015;16. pmid:25887754
- 50. Huang L, Wu B, Zhao J, Li H, Chen W, Zheng Y, et al. Characterization and Transferable Utility of Microsatellite Markers in the Wild and Cultivated Arachis Species. PLoS One. 2016;11(5). pmid:27243460
- 51. Carvalho SIC, Agassi CF, Ribeiro CSC, Reifschneider FJB, Buso GSC, Faleiro FG. Genetic variability of a Brazilian Capsicum frutescens germplasm collection using morphological characteristics and SSR markers. Genet. Mol. Res. 2017;16(3). pmid:28692124
- 52. Ali A, Pan YB, Wang QN, Wang JD, Chen JL, Gao SJ. Genetic diversity and population structure analysis of Saccharum and Erianthus genera using microsatellite (SSR) markers. Sci Rep. 2019;9. pmid:30674931
- 53. Gioia T, Logozzo G, Marzario S, Spagnoletti Zeuli P, Gepts P. Evolution of SSR diversity from wild types to U.S. advanced cultivars in the Andean and Mesoamerican domestications of common bean (Phaseolus vulgaris). PLoS One. 2019;14(1). https://doi.org/10.1371/journal.pone.0211342.
- 54. Walsh BM, Hoot SB. Phylogenetic relationships of Capsicum (Solanaceae) using DNA sequences from two noncoding regions: the chloroplast atpB ‐ rbcL spacer region and nuclear waxy introns. Int J Plant Sci. 2001;162(6):1409–1418.
- 55. Jarret RL, Dang P. Revisiting the waxy locus and the Capsicum annuum complex. GA J Sci. 2004;62:117–133.
- 56. FAO. The second report on the state of the world’s plant genetic resources for food and agriculture. FAO, Rome; 2010 http://www.fao.org/3/i1500e/i1500e00.htm.
- 57. Albrecht E, Zhang D, Mays AD, Saftner RA, Stommel JR. Genetic diversity in Capsicum baccatum is significantly influenced by its ecogeographical distribution. BMC Genet 2012;13 pmid:22866868
- 58. Albrecht E, Zhang D, Saftner RS, Stommel JR. Genetic diversity and population structure of Capsicum baccatum genetic resources. Gene Resources Crop Evol. 2011;59(4):517–38. https://doi.org/10.1007/s10722-011-9700-y.
- 59. Guzmán FA, Moore S, de Vicente MC, Jahn MM. Microsatellites to enhance characterization, conservation and breeding value of Capsicum germplasm. Genet Resour Crop Evol. 2020;67(3):569–585. https://doi.org/10.1007/s10722-019-00801-w.
- 60. Ince AG, Karaca M, Onus AN. Genetic relationships within and between Capsicum species. Biochem Genet. 2010;48(1–2):83–95. pmid:19916044
- 61. Campos KP, Pereira TNS, Costa FR, Sudré CP, Monteiro CES, Rodrigues R. Interspecific Hybridization among cultivated germplasm in Capsicum. 2004 Naples, Florida, 17th International Pepper Conference. p.20.
- 62. Shiragaki K, Yokoi S, Tezuka T. Phylogenetic analysis and molecular diversity of Capsicum based on rDNA-ITS region. Horticulturae 2020;6(4):87.
- 63. Tam SM, Lefebvre V, Palloix A, Sage-Palloix AM, Mhiri C, Grandbastien MA. LTR-retrotransposons Tnt1 and T135 markers reveal genetic diversity and evolutionary relationships of domesticated peppers. Theor Appl Genet. 2009;119(6):973–989. pmid:19618162
- 64. McLeod MJ, Guttman SI, Eshbaugh HW, and Rayle RE. An electrophoretic study of evolution in Capsicum (Solanaceae). Evol. 1983;37(3):562–574.
- 65. Ibiza VP, Blanca J, Cañizares J, Nuez F. Taxonomy and genetic diversity of domesticated Capsicum species in the Andean region. Genet Resour Crop Evol., 2012;59(6):1077–1088. https://doi.org/10.1007/s10722-011-9744-z.
- 66. Choong CY. DNA Polymorphisms in the study of relationships and evolution in Capsicum [dissertation]. Reading, UK: Univ. Reading; 1998.