Fresh leaves of P. cordata were collected in the field and immediately dried with silica gel for further DNA extraction. The cpDNA was isolated based on Shi et al. (2012) and stored at the Laboratorio Nacional de Identificación y Caracterización Vegetal (LaniVeg) at the University of Guadalajara (voucher JS571, Table 1). DNA quality was assessed by spectrophotometry in a NanoDrop 2000 (Thermo Fisher Scientific). DNA integrity was determined by electrophoresis in a 1% agarose gel, and DNA quantity was analyzed by fluorometry in a Qubit 2.0 (Thermo Fisher Scientific). The sample was sequenced using the Ion Torrent platform following the manufacturer’s protocol. The cpDNA was fragmented by sonication and used to prepare the library following the standard Ion Torrent Personal Genome Machine (PGM) protocol (200 bp fragments). The library was quantified by qPCR. The template was amplified in Ion OneTouch2 and enriched in OneTouch2 ES. Sequencing was performed using the Ion PGM Hi Q View Sequencing Kit. Raw data are available under the BioProject number PRJNA870909 in NCBI.

The quality of the raw reads was evaluated in FastQC 0.11.7 (Andrews 2010). Removal of low-quality reads was based on the Phred parameter (> 20) in Trimmomatic (Bolger et al. 2014). Reads were mapped to the plastome of P. philadelphica (Table 1) to exclude reads of nuclear and mitochondrial origin in Bowtie2 2 2.3.5 (Langmead and Salzberg 2012). Putative plastome reads were assembled de novo with SPAdes (Bankevich et al. 2012). Plastome coverage and assembly quality were performed in Quast (Gurevich et al. 2013). The complete plastome sequence was manually evaluated and corrected with IGV 2.5.0 (Thorvaldsdóttir et al. 2013). Automated annotation was performed in GeSeq (Tillich et al. 2017). tRNA genes were confirmed with tRNAscan-SE (Chan and Lowe 2019) and the remaining using BLAST in GenBank. The circular representation of the plastome was obtained in OGDraw 1.3.1 (Greiner et al. 2019).

The complete sequence of the plastome of P. cordata was compared with the plastomes of seven Physalis species: P. angulata, P. chenopodiifolia, P. minima, P. peruviana, P. philadelphica, P. pruinosa L., and P. pubescens. The cpDNA of P. chenopodiifolia and P. philadelphica were stored in the LaniVeg. Accession numbers, references and voucher or DNA number of Physalis species are listed in Table 1. The comparison included the genome sequence total and each region’s size, gene number and functional classification, nucleotide content, and number and size of introns.

The sequences of the eight plastomes were aligned in MAFFT (Katoh and Standley 2013) for various analysis. Sequence identity between coding and non-coding regions was assessed in mVista (Frazer et al. 2004) using Shuffle-LAGAN mode without modifying the pre-established values of the remaining parameters and using P. cordata as a reference. The limits of the LSC/IRs and SSC/IRs regions of the eight Physalis plastomes and A. officinarum (MH045575) were evaluated in IRscope (Amiryousefi et al. 2018) using the “Manual files” option and default settings. To assess nucleotide differences between coding and intergenic regions, nucleotide diversity (π) was calculated using DnaSP v. 6.12.03 (Rozas et al. 2017).

Forward, reverse, and palindromic repeat sequences in the plastomes were identified in REPuter (Kurtz et al. 2001) under the parameters of repeat unit (RU) length ≥ 21 bp, repeat identity ≥ 90%, and a Hamming distance of two. In addition, microsatellites present in each of the eight plastomes were identified with the MIcroSAtellite (MISA) identification tool (Beier et al. 2017). The search parameters were at least 10 RUs for mononucleotides, six for dinucleotides, and five for tri-, tetra-, penta-, and hexanucleotides.

To investigate the type of selection that has acted on Physalis plastome genes, we calculated the ratio of non-synonymous (Ka) and synonymous (Ks) substitutions. The Ka/Ks ratios of 51 genes that showed variation were evaluated. The aligned sequences were analyzed in KaKs_Calculator 2.0 (Wang et al. 2010). The 11^th genetic code (-c 11) was used. Ka/Ks ratios > 1, Ka/Ks = 1, and Ka/Ks < 1 suggested positive, neutral, and purifying selection, respectively.

To obtain a phylogenetic perspective on the relationships of P. cordata and the other seven species of Physalis sensu stricto we used A. officinarum as outgroup. The sequences of nine plastomes were aligned in MAFFT (Katoh and Standley 2013). The evolutionary model of the whole plastome dataset without partitions was estimated in jModelTest 2.1.10 (Darriba et al. 2012). GTR + I + G was the best evolutionary model. Finally, a maximum likelihood (ML) analysis was conducted in Garli 2.01 (Bazinet et al. 2014) with 1,000 bootstrap replicates.

The Physalis cordata plastome is 157,000 bp long and presents a quadripartite structure, with an LSC region of 87,267 bp, an SSC region of 18,501 bp, and two IRs of 25,616 bp (Fig. 1). The GC content was 37.52%, with a higher content in IRs (43.08%) than in the LSC (35.57%) and SSC (31.26%) regions (Table 2). There were 115 genes and five pseudogenes, including 80 genes coding for proteins, 31 for tRNA, and four for rRNA. Twenty-two duplicate genes were identified in IRs. Nineteen introns were present in 17 genes, two genes with two introns (clpP and ycf3) and the remainder with one (atpF, ndhA, ndhB, petB, rpl16, rpl2, rps12, rpoC1, rps16, trnA-UGC, trnG-GCC, trnI-GAU, trnK-UUU, trnL-UAA, and trnV-UAC). The rps12 gene (small ribosomal protein 12) was the only gene that was trans-spliced. This result implies that it has an intron, the first exon (5’ end) is in the LSC region, and the second (3’ end) is in IRb; therefore, it is duplicated in IRa (Table 3).

Figure 1. 

Plastome map of Physalis cordata. Genes located outside the outer circle are transcribed in the clockwise direction, whereas genes within the circle are transcribed in the counterclockwise direction. Genes with introns were marked with (*). Genes belonging to different functional groups are color-coded. Darker gray dashed area in the inner circle indicates GC content while lighter gray corresponds to the AT content of the plastome.

The comparison of the plastome of P. cordata with those of P. angulata, P. chenopodiifolia, P. minima, P. peruviana, P. philadelphica, P. pruinosa, and P. pubescens showed that all plastomes presented the typical quadripartite structure and genetic organization (Table 2). The sizes of the plastomes were variable, ranging from 156,692 bp in P. minima to 157,007 bp in P. pubescens. The regions also varied in size; the LSC region was 86,845 bp in P. minima and 90,977 bp in P. angulata, the SSC region was 18,393 bp in P. peruviana and 18,503 bp in P. minima, and the IRs were 23,667 bp in P. angulata and 25,695 bp in P. peruviana. The total GC content was similar in all species (37.51% in P. philadelphica and up to 37.56% in P. pruinosa), and by region, the total GC content was higher in the IRs (43.03% in P. minim up to 43.19% in P. pruinosa), intermediate in the LSC region (35.57% in P. cordata, P. chenopodiifolia, P. peruviana, and P. pubescens and up to 35.7% in P. pruinosa), and lower in the SSC region (31.26% in P. cordata and up to 31.4% in P. angulata and P. minima).

The plastome of P. cordata presented 115 genes. This number is only shared with P. philadelphica since P. angulata, P. minima, P. peruviana, P. pruinosa, and P. pubescens have 114 genes and P. chenopodiifolia 113 genes. Of the species sharing 113 genes, P. cordata and P. philadelphica differed in the presence of the trnP-GGG gene, and P. chenopodiifolia was lacking orf188. All species presented 22 genes in IRs and the rps12 gene was trans-spliced (Table 3). Of the shared genes, 103 were the same size, and 10 showed variation between species (accD, petB, psbB, psbC, psbH, rpl16, rpoC2, ψycf1, ycf1, ycf2, and ycf5). The number and distribution of introns were identical, 19 in 17 genes; however, 12 introns ranged from three to 99 bp (Suppl. material 1: Table S1).

The comparison of the limits of the LSC/IR and SSC/IR regions of the eight Physalis plastomes and A. officinarum showed some variations (Fig. 2). At the LSC/IRb boundary, the rps19 gene can be located at the end of the LSC region and continue at the beginning of the IRb (P. chenopodiifolia, P. cordata, P. minima, P. philadelphica, P. pubescens, and A. officinarum), presenting the second exon of rpl2 in the LSC region and the first in the IRb (P. peruviana and P. pruinosa) or the rpl23 gene in the LSC region and the trnM-CAU in the IRb (P. angulata). At the limit of IRb and the SSC region, two variations were observed: ψycf1 was in the IRb and ended at the beginning of the SSC region (P. angulata, P. chenopodiifolia, P. cordata, P. peruviana, P. philadelphica, P. pruinosa, and A. officinarum), followed by the ndhF gene or the final sequence of ψycf1, which ended at the limit of the IRb, and in the SSC region, the ndhF gene (P. minima and P. pubescens). In the SSC/IRa limit, the eight Physalis species and A. officinarum presented the ycf1 gene. Finally, the IRa/LSC limit showed three variations: the IRa may have the trnM-CAU gene and rpl23 (P. angulata) in the LSC region; the IRa may have the second exon of rpl2 and the trnH-GUG in the LSC region (P. chenopodiifolia, P. cordata, P. minima, P. philadelphica, P. pruinosa, and A. officinarum); or the IRa may have the first exon rpl2 and the second exon in the LSC region (P. peruviana and P. pruinosa) present. In addition, an extension of the LSC region and contraction in IRs were identified in P. angulata and P. pruinosa.

Figure 2. 

Comparison of border regions of plastomes of Physalis species and Alkekengi officinarum.

The identity between the plastome of P. cordata and those of the other seven Physalis species was high. Identical sequences were mainly found in coding regions, and the greatest divergence was in the intergenic regions. The comparison between regions showed that the LSC and SSC regions were more divergent than were IRs. Introns also exhibited greater variation than the exons. The most divergent genes were ycf1 and ycf2, as well as the intergenic regions trnH-GUG-psbA and trnL-UAA-trnF-GAA (Fig. 3).

Figure 3. 

Comparative plots of identity among Physalis species. The percentage of identity ranges from 50 to 100% and is shown in the vertical axis. Gray arrows indicate genes with their orientation and position of their transcription in the reference plastome (P. cordata). Plastome regions are color coded as blue blocks for the conserved coding genes (exon), turquoise for introns and red blocks for non-coding sequences in intergenic regions (CNS).

The sequences of 51 genes and 75 intergenic regions showed variation. The lowest variation in genes was one change in 14 genes, and the highest variation was 173 changes in ycf1. The lowest variation in intergenic regions was one change in 16 of them, and the highest was in trnL-UAA-trnF-GAA with 42. The average value of π was lower in the genes than in the intergenic regions (Suppl. material 1: Fig. S1). The nucleotide diversity in the genes varied from π = 0.00016 in ndhB to π = 0.01038 in ycf1 and in the intergenic regions from π = 0.0003 in rps7-trnV-GAC to π = 0.02671 in trnL-UAA-trnF-GAA. In general, 14 regions presented π values > 0.005, which included four genes (trnD-GUC, trnW-CCA, ndhE, and ycf1) and 10 intergenic regions (trnH-GUG-psbA, trnfM-CAU-rps14, trnL-UAA-trnF-GAA, petA-psbJ, rps18-rpl20, infA-rps8, rpl16-rps3, rpl32-trnL-UAG, trnL-UAG-ccsA, and ndhG-ndhI).

The repeated sequences in the plastome ranged from 35 in P. philadelphica to 49 in P. cordata (Suppl. material 1: Fig. S2). The most abundant type of repetition was forward (19 in P. philadelphica up to 29 in P. cordata), followed by palindromic (five in P. philadelphica up to 23 in P. angulata) and finally reverse (one in P. philadelphica up to three in P. angulata, P. minima, P. peruviana, and P. pruinosa). The number of microsatellites fluctuated from 52 in P. peruviana to 62 in P. angulata (Suppl. material 1: Fig. S3). Mono-, di-, and trinucleotide URs (repeat units) were present in all eight species; tetranucleotides were absent in P. peruviana, and pentanucleotides and hexanucleotides were only present in P. angulata and P. pruinosa. The types of UR with the highest number were T and A mononucleotides. In contrast, the mononucleotide C is present in a single region in five species (P. chenopodiifolia, P. cordata, P. minima, P. philadelphica and P. pubescens), and G was not found in any regions. In turn, the region with the highest number of microsatellites was the LSC region, followed by the SSC region, and then IRs.

In 51 genes, eight showed values of Ka/Ks > 1, indicating that they are under positive selection (cemA, ndhB, ndhJ, ndhK, psaC, rbcL, rpoA, and ycf1, Fig. 4). Six genes -cemA, ndhB, ndhJ, ndhK, rbcL, and rpoA- are under positive selection in all eight species, psaC in five (P. chenopodiifolia, P. cordata, P. peruviana, P. philadelphica, and P. pruinosa), and ycf1 in four (P. angulata, P. peruviana, P. philadelphica, and P. pruinosa). The values in ndhB, ndhJ, ndhK, and psaC were slightly higher than 1, compared to those in cemA, rbcL, rpoA, and ycf1, which presented higher values (Ka/Ks = 1.516 to Ka/Ks = 6.029). The other 43 genes showed Ka/Ks values <1, which indicates they are under purifying selection.

Figure 4. 

Non-synonymous substitution (Ka), synonymous substitution (Ks) and Ka/Ks values of genes of eight Physalis species. Blue bars indicate Ka values, orange bars indicate Ks values, and gray bars indicate Ka/Ks ratios. Genes with Ka or Ks values equal to 0 are not shown.

The ML phylogeny recovers P. minima as a sister to the seven other Physalis species included in this study (BS = 100; see Fig. 5). We recovered P. philadelphica as sister to the clade containing P. pruinosa, P. angulata, P. peruviana, and P. chenopodiifolia (BS = 91). This group was in turn sister to the clade formed by P. pubescens and P. cordata (BS = 91).

Figure 5. 

Phylogenetic tree based on Maximum Likelihood (ML) using the plastome sequence of Physalis species. Bold lines indicates bootstrap support values (BS) = 100, values < 100 are shown above the branches.

The plastome of P. cordata analyzed here presents the typical quadripartite structure and the same order of genes as has been found for other species of the genus. However, the species vary in the total size and the size of the regions. In general, the average size of plastomes in Physalis is 156,814 bp, and the difference between the largest (P. pubescens, 157,007 bp) and the smallest plastome (P. minima, 156,692 bp) was 315 bp. Phylogenetically, closely related species tend to be homogeneous in size and their regions (Daniell et al. 2016; Mohanta et al. 2020). In the studied Physalis, however, this is true in terms of the size of the plastome but not with regions. The expansion of the LSC regions and contraction of IRs found in P. angulata and P. pruinosa is not an isolated evolutionary event; it has characterized the evolutionary history of other groups of angiosperms such as Indigofera L. (Fabaceae, Oyebanji et al. 2020), Passiflora L. (Passifloraceae, Pacheco et al. 2020), and Corydalis DC. (Papaveraceae, Xu and Wang 2021). Expansions and contractions in regions have occurred multiple times and in different lineages; there are models that indicate this may be due to single or double-strand breaks or be promoted by multiple inversions along with several rounds of expansions and/or contractions (Zhu et al. 2016). In Physalis the total GC content shows a minimal variation range of only 0.05%, ranging from 37.51% (P. philadelphica) to 37.56% (P. pruinosa). These values are similar to those documented in other genera of Solanaceae, such as Atropa L., Capsicum L., Nicotiana L., and Solanum L. (Kaur et al. 2014; Magdy et al. 2019), and families such as Asteraceae and Saxifragaceae (Zhong et al. 2019; Liu et al. 2020).

The Physalis species studied have between 113 to 115 genes; 113 of these are completely shared, with the same distribution and number of introns. The difference in the number of genes is based on the presence of trnP-GGG in P. cordata and P. philadelphica and the absence of orf188 in P. chenopodiifolia. We suggest that the genes that are not shared are the product of loss events during the evolutionary process. In addition, the size of 10 genes was different in at least one of the eight species. For example, in P. philadelphica, the second exon of the petB gene differs by three bp with respect to those of the other species. Additionally, gene sizes are variable among the eight species, as occurs for ycf1, which varies from six to 114 bp. In the eight species, there were 17 genes with 19 introns, 15 genes with one intron, and two genes with two introns (clpP and ycf3). Physalis does not have an intron in the petD gene (gene of the cytochrome b6-f subunit 4 complex), unlike that which occurs in other genera of Solanaceae such as Atropa, Capsicum, Datura L., Nicotiana, Solanum, and Withania Pauquy (Kaur et al. 2014; Mehmood et al. 2020a) and in families belonging to other orders such as Oxalidaceae (Li et al. 2021) and Lamiaceae (Zhao et al. 2020). Similarly, 12 introns presented differences ranging from three bp in atpF and petB to 99 bp in trnI-GAU. In Physalis, the difference in the sizes of exons and introns does not impact the total size or the regions; however, changes in intergenic regions could contribute to the unequal sizes.

Variation in the boundaries of plastome regions is a relatively common evolutionary process that occurs in different plant groups (Huang et al. 2020). In the Physalis species studied, this variation is present in three types, distinguished by the presence of different genes at the IRs/LSC boundaries (Fig. 2). The first type is most common and it is found in P. chenopodiifolia, P. cordata, P. minima, P. philadelphica, P. pubescens, and A. officinarum. In these species the rps19 gene starts at LSC and ends at IRb at LSC/IRb boundary; furthermore, at IRa/LSC boundary, the second exon of the gene rpl2 is at IRa and the trnH-GUG gene at LSC. The second type is found in P. peruviana and P. pruinosa, which had the intron of the rpl2 gene at both of the IR/LSC boundaries. The third and most distinctive type is present in P. angulata, here there is change in the order of the genes, with rpl23 located at LSC and trnM-CAU at IRs (Fig. 2). The changes in P. angulata and P. pruinosa may be a product of the expansion of LSC and contraction of IRs about 2 kb (Table 2) which contrasts with the rest of the Physalis species analyzed, other Solanaceae genera, and several land plant families with average sizes of 25 kb in IRs (Kaur et al. 2014; Ruhlman and Jansen 2014; D’Agostino et al. 2018; Zhao et al. 2020; Yang et al. 2021a). Our results are somewhat similar to those reported by Feng et al. (2020), where the boundaries of the four regions reported in P. angulata, P. minima, P. peruviana, and P. pubescens are the same as those observed in the first type identified by us for P. chenopodiifolia, P. cordata, P. minima, P. philadelphica, and P. pubescens. In contrast, A. officinarum differs in the IRs/LSC boundaries by the presence and position of rpl2 gene (Fig. 2). Furthermore, this species exhibits an expansion of LSC (ca. 1.2 kb) and a contraction of IRs (ca. 0.7 kb) like that in P. pruinosa. Boundary variations are heritable and provide information on evolutionary and speciation processes. These mutations can be traced throughout the evolutionary process and used as evidence of shared ancestry (Stettler et al. 2021). In the Physalis species analyzed here there appears to be no evolutionary pattern that characterizes the boundaries of the four regions; future studies are necessary to identify a particular pattern in the genus.

The variation between plastomes, in some cases, is limited due to their low rate of evolution, so repetitive regions and microsatellites can reveal interspecific variation (D’Agostino et al. 2018). In the case of repetitive regions, their divergence has been correlated to a precursor of inversions and rearrangements, so their analysis allows for different types of studies (Weng et al. 2014). In Physalis, repetitive regions mostly have sizes of 30–39 bp. This result coincides with those found in other genera of Solanaceae (Nicotiana, Mehmood et al. 2020b; and Withania, Mehmood et al. 2020a) and even in phylogenetically distant families such as Moraceae (Souza et al. 2020) and Poaceae (Wang et al. 2021). In turn, microsatellites have been used for the identification of plants and in analysis of population genetics and relationships between cultivars of the same species (Bassil et al. 2020). In Physalis, the most abundant URs are mononucleotides T and A, this may be the result of the high content of T and A in the plastome in relation to G and C. Most microsatellites are in the LSC region, which is probably because this region is longer than the SSC region and IRs. Additionally, microsatellites occur mainly in non-coding regions rather than in coding regions. The microsatellites identified in Physalis plastomes could be useful as potential molecular markers.

The evolutionary history of species is shaped by two main factors: mutation, which generates new genotypes, and selection, which determines the probability that new genotypes will be fixed or eliminated (Marcos and Echave 2020). If selection fixes the mutations, then the patterns of polymorphism, divergence, and gene expression are modified (Johri et al. 2020). Mutations, based on the effect of amino acid coding, are classified as Ka and Ks. Their relationship (Ka/Ks) allows us to understand the independent evolutionary history of each gene and determine if it is under positive selection (Ka/Ks > 1), purifying/stabilizing (Ka/Ks < 1), or neutral selection (Ka/Ks = 1) (Menezes et al. 2018; Yang et al. 2020a). In Physalis, most of the genes analyzed were under purifying selection. This implies that these regions of the plastome are maintained in terms of size and nucleotide content and that the variants that could modify the functions of the encoded proteins are eliminated (Cho et al. 2021; Yang et al. 2021a). However, eight genes were under positive selection, either in the eight species or in some of them. This result implies that certain allelic variants are fixed and benefit the optimization of physiological processes and adaptive advantages to the environment (Cho et al. 2021). Under this condition, the genes cemA, ndhB, ndhJ, ndhK, rbcL, and rpoA occurred in the eight species; psaC occurred in five species (P. chenopodiifolia, P. cordata, P. peruviana, P. philadelphica, and P. pruinosa), and ycf1 occurred in four species (P. angulata, P. peruviana, P. philadelphica, and P. pruinosa).

Throughout the evolutionary history of the plastome, most genes have been under purifying selection due to functional limitations (Yang et al. 2020b). However, positive selection can act on those that encode proteins involved in environmental adaptive processes or during the domestication process (D’Agostino et al. 2018; Li et al. 2020b; Yang et al. 2020b). In Physalis, the eight genes that are under positive selection can confer certain advantages. The genes ndhB, ndhJ, and ndhK (NADH-dehydrogenase subunits B, J, and K) protect against stress caused by high concentrations of light, stabilizing the NADH complex and adjusting the photosynthetic rate, in addition to delaying plant growth because of drought (Yang et al. 2021a). The cemA gene (protein that envelops the chloroplast membrane) contributes to the absorption of more CO₂ by chloroplasts (Chen et al. 2020). The rbcL gene (large subunit of RuBisCO) increases the transfer of electrons during the process of photosynthesis, as well as the catalytic activity on CO₂ (Piot et al. 2018; Gui et al. 2020). The psaC gene (subunit of photosystem I), which occurred in the six species that are under positive selection, increases the photosynthetic rate when plants are exposed to high concentrations of ambient light (Fischer et al. 1998). The rpoA gene (alpha subunit of RNA polymerase) increases the transcription and expression of plastomic photosynthetic genes so that a plant develops correctly (Mehmood et al. 2020b). Finally, the ycf1 gene (membrane protein) is essential for cell survival and improves the construction of the cell membrane and the importation of photosynthetic proteins that contribute to the environmental adaptation process (Ye et al. 2018; Wang et al. 2020). This gene is differentially expressed in P. angulata, P. cordata, P. philadelphica, and P. pruinosa. In contrast to that which occurs in other genera, such as in Citrus L. (Carbonell-Caballero et al. 2015), in Physalis, it is not possible to associate the differential expression of genes with biological and ecological characteristics shared between the species analyzed (Suppl. material 1: Table S2). This can be the result of the historical evolutionary process of the species, such as in the case of ycf1 (Carbonell-Caballero et al. 2015; Jiang et al. 2018; Yang et al. 2021b).

Coding and non-coding regions of plastomes both tend to have a high degree of conservation (Daniell et al. 2016; Tonti-Filippini et al. 2017). But some variable regions are routinely used in the construction of phylogenetic hypotheses, phylogeographic analysis, and population genetics (Kim et al. 2020; Li et al. 2020a; Zhang et al. 2020; Zhao et al. 2021). Our results show that the π values in the coding and non-coding regions in Physalis are lower than those documented in other genera of Solanaceae, such as Nicotiana (Mehmood et al. 2020b) and Capsicum (D’Agostino et al. 2018). In Physalis, previous phylogenetic analyses have not resolved the relationships between species due to the presence of polytomies or low support values. These studies have only included one to five of the following regions: matK, rbcL, ndhF, psbA-trnH, rpl32-trnL, trnL-trnF, trnS-trnG, and ycf1 (Olmstead et al. 2008; Särkinen et al. 2013; Zamora-Tavares et al. 2016; Feng et al. 2018; Deanna et al. 2019). Use of more plastome regions in phylogenetic analyses has the potential to help clarify species level relationships. We recommend using regions of the plastome with values of π > 0.005 (trnD-GUC, trnW-CCA, ndhE, and ycf1 and the intergenes trnH-psbA, trnfM-rps14, trnL-trnF, petA-psbJ, rps18-rpl20, infA-rps8, rpl16-rps3, rpl32-trnL, trnL-ccsA, and ndhG-ndhI).

The phylogenetic perspective we obtained confirms the usefulness of the plastome as a source of information for conducting phylogenetic studies in Physalis, despite the limited number of species studied. In comparison with other studies that include partial nucleus and chloroplast sequences (Whitson and Manos 2005; Zamora-Tavares et al. 2016; Deanna et al. 2019), our analysis had high support values and polytomies are not present. In this study, P. minima is rescued as a sister species to the remaining seven. This partially agrees with Deanna et al. (2019) where P. minima is recovered as sister to the great majority of Physalis species. In contrast, in the study of Whitson and Manos (2005) P. minima forms a clade with P. angulata, P. cordata, and P. pubescens. Similar to Deanna et al. (2019), in our work P. angulata and P. pruinosa maintain a sister species relationship, while in Zamora-Tavares et al. (2016) P. pubescens is sister to P. angulata. Furthermore, the phylogenetic relationship of the Physalis species studied based on the plastome does not reflect groupings according to the chromosome number, as P. angulata, P. minima, P. peruviana, and P. pubescens have n = 24, while the other species have n = 12. This agrees with the results of Rodríguez et al. (2021), who showed that the genera Physalis, Quincula Raf. and Chamaesaracha (A.Gray) Benth. & Hook comprise a lineage with asymmetric karyotypes. For its part, A. officinarum has a symmetric karyotype (Rodríguez et al. 2021) and is an independent lineage. Moreover, since Physalis includes 95 species, the inclusion of a large number of species is needed to elucidate its evolutionary history and to analyze if it has a correlation with their ecological affinities and the life history of the species.