Introduction

Like other Solanaceae genera, Physalis L. includes a number of commercially important edible and ornamental species. With the exception of P. alkekengi L., most of the 90 species in the genus are native to the Americas, and their distribution ranges from the USA through Mexico and into South America, including the Antilles. Many weedy and cultivated species have also been introduced to Old World tropical regions (Martínez 1998; Whitson and Manos 2005). With 50 native species, Mexico is the center of diversity (D’Arcy 1991; Waterfall 1967). In natural areas and traditional agroecosystems, the fruit of at least ten wild species are collected for consumption, but only Physalis philadelphica Lam. has been domesticated and cultivated in the region for centuries.

The high diversity of Physalis species observed in Mexico parallels the presence of high inter- and intraspecific morphological variation. This variation is observed mostly in widely distributed species or those that are managed and cultivated, making taxonomic identification more challenging.

Physalis species are currently identified by morphological characteristics. A number of studies have been done using morphological data to characterize species, to study the systematics of some sections (Martínez 1998, 1999a; Seithe and Sullivan 1990; Sullivan 1985; Vargas-Ponce et al. 1999, 2001, 2003), to determine the infrageneric taxonomy (Martínez 1999b), and to understand the phylogenetic relationships between Physalis and the genera related to the physaloid group (Axelius 1996; Estrada and Martínez 1999). Sequence data from part of the nuclear Waxy gene, the internal transcribed spacer of the nrDNA (ITS) and chloroplast regions (Martínez 1998; Whitson and Manos 2005) have also been used to address the phylogeny of Physalis, related genera, and the Solanaceae (Olmstead et al. 2008).

The PCR-based technique using inter-simple sequence repeats (ISSR) is widely used because of its technical simplicity and reproducibility and the high polymorphism of the markers (Arif et al. 2009; Bornet and Branchard 2001; Reedy et al. 2002; Zietkiewicz et al. 1994). ISSRs provide extremely useful data for estimating genetic diversity (Ge et al. 2005; Nan et al. 2003; Petros et al. 2007; Gomes et al. 2009; Vargas-Ponce et al. 2009), phylogenetic analyses (Matos et al. 2001; Mort et al. 2003; Wolfe and Randle 2001) for studying the inter- and intraspecific relations among plants (Archibald et al. 2006; Petros et al. 2008; Wolfe and Randle 2001). The objective of the present study was to determine whether ISSR markers provide sufficient resolution for interspecific discrimination of Mexican species of Physalis and to assess their utility for inferring interspecific relationships.

Materials and Methods

Plant Material

Samples were taken from 12 wild Physalis populations, representing eight species and five taxonomic sections of the genus. The plants were collected in the states of Aguascalientes, Coahuila, Colima, and Jalisco, Mexico (Table 1). Three species are perennial, and five are annual. Leaf material was collected and dried in silica gel from 11 to 20 individuals per population. Physalis virginiana Miller, Physalis cordata Miller, Physalis cinerascens (Dunal) Hitch., Physalis lignescens Waterfall, and Physalis sulphurea Waterfall were represented by a single population, while Physalis angulata L. and Physalis lagascae Roem. and Schult. were represented by two populations each. The highly morphologically variable P. philadelphica Lam. was represented by three populations (Table 1).

Table 1 Eight Physalis species and 12 populations included in the study
Full size table

DNA Extraction and PCR Amplification

DNA was extracted from 20 mg of leaf samples following the CTAB method of Doyle and Doyle (1987). The extracted DNA of five individual plants from the same population was mixed to create a genetic pool representing each population using the approach of Martínez et al. (2008), and Gilbert et al. (1998). Eleven primers were used in the initial screening; six were selected because they showed the highest potential to distinguish between species (Table 2). PCR was carried out twice by each ISSR primer with all Physalis samples included in the study. Only those bands that were present in both runs were scored and included in the data set. Each 20-μl PCR reaction mixture contained 60-ng genomic DNA, 10 mM buffer, 2.5 mM MgCl2, 0.3 mM dNTPs, 0.2 μM primer, 0.5 unit Taq polymerase, and HPLC water. The PCR reactions were performed in a PTC-100 thermal cycler (MJ Research, Inc.) programmed for an initial step of 4 min at 95°C; followed by 40 cycles of 1 min at 95°C, 1 min annealing temperature (see Table 2), 2 min extension at 72°C; and a final 15 min extension at 72°C.

Table 2 ISSR primers used in the analysis, number of polymorphic loci, and discriminatory resolution of each primer (Simpson’s Index)
Full size table

Amplification products were separated in 6% 19:1 bis-acrylamide gels containing 7 M urea and TBE buffer IX. A 100-bp molecular marker standard was included in each gel to estimate band sizes. Electrophoresis was run at 70 V in a dual adjustable slab gel unit (CBS Scientific) until the bromophenol blue indicator dye had traveled 20 cm. The products were visualized with the silver staining technique of Sanguinetti et al. (1994).

Data Analysis

A digital imaging system (Kodak Gel Logic 100) was used to record gel images as TIFF files. ISSR banding patterns were analyzed with Phoretix 1D image analysis software (TotalLab). Bands were automatically assigned, then manually adjusted based on the resulting images. Each unique fragment size, considered as a locus for each primer, was scored as present (1) or absent (0) for each sample. The polymorphic loci percentage (Pl) by primer across samples was calculated using the POPGENE v.1.31 program (Yeh and Boyle 1999). Each primer’s discrimination potential was expressed in terms of the Simpson’s diversity index \( \left[ {{h_j} = \sum {\left( {1 - \sum {{p_i}^2} } \right)/n} } \right] \), where p i is the frequency of the ith allele, and n corresponds to the number of loci detected by each primer (Hunter and Gaston 1988; Lüdtke et al. 2009). A value of 1.0 indicates that the primer is able to discriminate between all samples, and a value of 0.0 indicates that all samples are identical. From the 101 polymorphic bands, only 74 were chosen for the genetic relation analysis because of their clarity and constant amplification in all samples. These bands were generated with the primers (GA)8YG, (CA)6RY, (CA)6RG, and RY(GACA)3. Genetic similarity among all samples was calculated using the Nei and Li coefficient (Nei and Li 1979), which excludes absent bands from the analyses. Genetic similarity values were used to estimate genetic relations, and then represented with a dendrogram generated by the unweighted pair-group mean analysis method using the Freetree (Pavlicek et al. 1999). Relative support for the relations was assessed with bootstrap analyses (1,000 replicates; Felsenstein 1985) using the Freetree (Pavlicek et al. 1999) and Tree View (Page 1996) programs. Another genetic relation analysis was applied to a small subset of Physalis samples of Physalis section Angulatae (four species/eight populations, see Table 1) to assess the potential utility of ISSR data for inferring fine relations among taxa showing close genetic relations.

Results and Discussion

The six ISSR primers amplified 101 scorable bands (Table 2). The number of bands per primer varied from 10 to 21, with an average of 16.8 fragments per primer. The amplified product ranged from 200 to 1,600 bp, while the number of products amplified per primer varied from 10 [(CT)8RG] to 21 [(GA)8YG]. All primers produced 100% polymorphic bands, a very high interspecific percentage that has also been reported for Polygala (Polygalaceae), Guizotia (Asteraceae), Morus (Moraceae), and Agave (Agavaceae) (Lüdtke et al. 2009; Petros et al. 2007, 2008; Prasanta et al. 2008; Vargas-Ponce et al. 2009).

ISSR—Taxonomic Discrimination Among Physalis Species

All six primers produced different fingerprinting profiles for each species (Fig. 1). In addition, four of the primers generated species-specific bands. Primer (CA)6RG generated a unique band for P. philadelphica (580 bp), and primer (CA)6RY produced a specific band for P. cordata (1,100 bp). Likewise, primer RY(GACA)3 yielded a band for P. cinerascens (590 bp) and another for P. sulphurea (700 bp). Last, the primer (CT)8RC resulted in a unique band for P. cinerascens (700 bp), P. philadelphica (1,350 bp), and P. sulphurea (1,400 bp). Thus, ISSR markers can be useful in generating ISSR genetic fingerprints specific to each Physalis species. Indeed, ISSRs have been useful in the genomic fingerprinting of diverse plant groups Reedy et al. (2002).

Fig. 1
figure 1

Inter-simple sequence repeat (ISSR) banding pattern obtained on acrylamide gels for 12 populations representing eight Physalis species with the primer. (GA)8 YGM = DNA Molecular size marker (100 bp ladder) in the last lane. Lane/s: 1, P. virginiana; 2, P. cordata; 3, P.sp; 4, P. cinerascens; 5, P. lignescens; 6, P. sulphurea; 7–9, P. philadelphica; 10–11, P. angulata; 12–13, P.lagascae

Full size image

Banding patterns showed clear interspecific differentiation and intraspecific similarity, demonstrating the value of ISSRs in taxonomic discrimination of Physalis species (Fig. 1). Simpson’s index values varied from 0.483 (primer RY(GACA)3) to 0.579 (primer (CT)8RC; Table 2) showing moderate to high potential to discriminate between samples.

Because Simpson’s diversity index is sensitive to the number of groups or samples and size of the group, the variation may be attributed to the unequal number of individuals per sample in the study (Dillon et al. 1993). However, Simpson’s index has been successfully used to discriminate among microorganisms (Dillon et al. 1993; Harth-Chu et al. 2009) and plants (Agostini et al. 2008; Lüdtke et al. 2009), with values slightly greater than those obtained here for Physalis.

Genetic Similarity and Relations

Genetic similarities, as shown by ISSRs, did not parallel the morphological variation and ecological preferences observed for the Physalis species included in this study. Genetic interspecific similarity values, as calculated with the Nei and Li coefficient, ranged from 0.20 to 0.57. The highest values were observed for the P. lignescens/P. sulphurea comparison (0.57) and the lowest for the P. cinerascens/P. virginiana comparison (0.20). The genetic similarity between P. lignescens and P. sulphurea contrasts with their growing habits and habitat preferences. Physalis lignescens is a perennial herb growing in coniferous forests, whereas P. sulphurea is an annual herb found in flood plains (Vargas-Ponce et al. 2003). On the contrary, P. cinerascens and P. virginiana, although morphologically different, had a very low genetic similarity value, maybe reflecting that both species are perennial herbs growing in a tropical, dry forest. In future studies, sampling must be widened to better represent the genus Physalis, all of its taxonomical sections, and the diversity in biological form and gradients in its ecological and geographical habitats.

The genetic similarity values between species pairs in Physalis section Angulatae ranged from 0.24 to 0.59: P. philadelphica/P. angulata (0.32–0.59), P. philadelphica/P. lagascae (0.24–0.39), P. philadelphica/P. sulphurea (0.25), and P. lagascae/P. angulata (0.36–0.45). Intraspecific similarity values were highest for the three species of the Physalis section Angulatae: P. angulata (0.71), P. philadelphica (0.63), and P. lagascae (0.55). Similar values have been reported for Polygala (Polygalaceae), Cunila (Lamiaceacea), Guizotia (Asteraceae), and other genera (Reedy et al. 2002; Agostini et al. 2008; Petros et al. 2008; Lüdtke et al. 2009).

The genetic relations phenogram (Fig. 2) that included all samples showed two principal clusters diverging at the 0.31 phenon level. Cluster one (I) included P. cinerascens (Physalis sect. Stellatae), P. sulphurea (Physalis sect. Angulatae), P. lignescens (Physalis sect. Coztomatae), and P. cordata (Physalis sect. Epeteiorhiza). Cluster two (II) included P. angulata, P. lagascae, and P. philadelphica of Physalis sect. Angulatae, and one population of P. virginiana (Physalis sect. Lanceolatae). Neither cluster corresponds to the presently accepted intrageneric taxonomic classification (Martínez 1999b). As mentioned already, cluster II encompasses P. angulata, P. lagascae, and P. philadelphica, but excluded P. sulphurea, which belongs to the same section (Fig. 2a).

Fig. 2
figure 2

Genetic similarity relationship among eight Physalis species (a) and among Physalis species of Section Angulatae (b). The bootstrap values are shown on the branches (1,000 permutations)

Full size image

The analysis of distantly related lineages (e.g., P. cinerascens, P. lignescens, and P. virginiana) might generate incongruence between genetic similarity values and taxonomic section affiliation. As a result, high bootstrap values support the intraspecific nodes but not the interspecific ones. The cluster analysis of Physalis sect. Angulatae had better resolution and higher boostrap values (Fig. 2b), suggesting an advantage to including closely related species in the analyses. Two important aspects of the genetic similarity analyses were found in the phenograms. First, samples corresponding to the same species grouped together. Second, the genetic differentiation among Physalis species corresponded to the observed morphological variation between them, indicating that distinctive morphological traits can be used to differentiate the species. For example, the small corolla and the small yellow anthers of P. angulata separate it from P. philadelphica (Vargas-Ponce et al. 2003).

The results demonstrate that ISSR markers are useful in estimating genetic relations in Physalis. The markers are also an effective tool for documenting interspecific variability, discriminating species, and assessing intraspecific similarity and differentiation. Moreover, the development of this highly versatile marker for Physalis opens the way for implementing a wide range of studies needed to understand, use, and conserve species of husk tomato.