Identiﬁcation and Genetic Diversity Analysis of Edible and Medicinal Malva Species Using Flow Cytometry and ISSR Molecular Markers

: The Malva genus contains species that reveal therapeutic properties and are mostly important in medicine and the functional food industry. Its breeding, cultivation, and utilization are based on proper germplasm / plant identiﬁcation, which is di ﬃ cult using morphological features. For this reason, we applied ﬂow cytometry and inter simple sequence repeat polymerase chain reaction (ISSR-PCR) for fast and accurate species identiﬁcation. Genome size estimation by ﬂow cytometry was proposed as the ﬁrst-choice method for quick accession screening. Out of the 12 tested accessions, it was possible to identify six genotypes based on genome size estimation, whereas all species and varieties were identiﬁed using ISSR markers. Flow cytometric analyses revealed that Malva species possessed very small (1.45–2.77 pg / 2C), small (2.81–3.80 pg / 2C), and intermediate (11.06 pg / 2C) genomes, but the majority of accessions possessed very small genomes. Additionally, this is the ﬁrst report on genome size assessment for eight of the accessions. The relationships between the investigated accessions showed the presence of two clusters representing malvoid and lavateroid group of species. Flow cytometry and ISSR molecular markers can be e ﬀ ectively used in the identiﬁcation and genetic characterization of Malva species. genetic variation Malva accessions. The results of this study can be useful in mallow breeding programs, planning of conservation strategies, germplasm collection, and taxonomy of the genus.

The Malva genus is morphologically very diverse, but some species are hardly distinguishable based on morphological features [29]. Several studies have been conducted to clarify the taxonomic affiliation of Malva species using different features, such as molecular data (nuclear ribosomal DNA (rDNA), internal transcribed spacer (ITS) region, intron-exon splice junction (ISJ), and inter simple

Plant Material
Twelve Malva accessions were used in this study ( Table 1). The seeds of 11 species were obtained from GRIN-ARS-USDA (Germplasm Resources Information Network, Agricultural Research Service, United States Department of Agriculture) gene bank, while seeds of M. sylvestris var. mauritiana (L.) Boiss. were collected from the Medicinal and Cosmetic Plant Garden of the Nicolaus Copernicus University, Collegium Medicum, Bydgoszcz, Poland. The seeds were sown in pots with 1:2 sand and commercial humus mixture in a growth chamber for 16/8 h at 26/18 • C (day/night).

Genome Size Estimation
Leaves of Malva accessions were prepared for nuclear DNA content analysis as previously described by Jedrzejczyk and Rewers [45] with some modifications. Plant tissue (about 0.5-1.0 cm 2 ) was chopped in 1 mL of nuclei isolation buffer (200 mM TRIS, 4 mM MgCl 2 •6H 2 O, 0.5% (v/v) Triton X-100) with the addition of propidium iodide (PI, 50 µg/mL) and ribonuclease A (RNase A, 50 µg/mL). The prepared suspension of nuclei was analyzed using a CyFlow SL Green (Partec GmbH, Münster, Germany) flow cytometer. The nuclear DNA content of 5000-7000 nuclei was measured for each sample on a linear scale. The fluorescence intensity histograms (CV = 5.28%-6.57%) were evaluated using FloMax software (Partec GmbH, Münster, Germany). Leaves of Zea mays L. CE-777 (5.43 pg/2C; [46]) and Glycine max (L.) Merr. "Polanka" (2.50 pg/2C; [47]) were used as internal standards (Table 2). For each Malva accession, six randomly selected plants were analyzed. Genome size was determined using the linear relationship between the mean ratio of the 2C peak positions of the Malva accessions and the internal standard on the histogram of fluorescence intensities. The obtained 2C DNA values (pg) were converted to megabase pairs (Mbp) of nucleotides using the following formula: 1 pg = 978 Mbp [48]. The results were analyzed using a one-way analysis of variance (ANOVA) and a Duncan's post-hoc test (p < 0.05; Statistica v. 13.3, StatSoft, Poland).

Genomic DNA Extraction
Total DNA was isolated from 0.12 g of fresh mallow leaves from three randomly selected plants per accession using a GeneJet Plant Genomic DNA Purification Mini Kit (Thermo Scientific, Waltham, MA, USA). After DNA extraction, the quality and quantity of DNA samples were assessed by spectrophotometry, and the quality was additionally assessed by 1% (w/v) agarose gel electrophoresis.

ISSR-PCR Amplification
Twenty-five ISSR primers (Genomed, Warsaw, Poland) were examined, out of which 13 generated stable and clear band patterns and were used for PCR amplifications ( Table 3). The PCR reactions were carried out in a total volume of 12.5 µL containing 30 ng of genomic DNA template, 0.1 U/µL Taq DNA polymerase, 4 mM MgCl 2 , 0.5 mM of each dNTP, 10 µM ISSR primer, and sterile deionized water. Reactions were performed using T100 Thermal Cycler (Bio-Rad, Hercules, CA, USA) and the following conditions: an initial denaturation at 94 • C for 5 min, then 1 min denaturation at 94 • C (40 cycles), followed by 1 min annealing at 48.5-67.2 • C (depending on the ISSR primer, Table 3), 2 min elongation at 72 • C, and 7 min of final extension at 72 • C. The PCR products were separated on 1.5% (w/v) of agarose gel electrophoresis. The size of the obtained bands was estimated by comparison to a 3000 bp DNA ladder (GenoPlast Biochemicals, Rokocin, Poland). The gels were visualized and archived via GelDoc XR+ (Bio-Rad, Hercules, CA, USA).

ISSR Marker Analysis
The ISSR amplicons were scored as presence (1) or absence (0) of bands in the form of a binary matrix. Only reproducible and clear bands were counted. Monomorphic and polymorphic bands were calculated for each ISSR primer. The polymorphic information content (PIC) was determined using the following equation: PIC = 1 − p 2 − q 2 , where p is the band frequency and q is no-band frequency [50]. Genetic distances were calculated for all accessions according to Nei and Li [51]. Cluster analysis was performed with the UPGMA algorithm (unweighted pair group method with arithmetic average) and the Treecon ver. 3.1 software [52]. To estimate the statistical support of the dendrogram branches, the data were bootstrapped using 2000 replications, and values higher than 50% were mapped ( Figure 1).

ISSR Marker Analysis
The ISSR amplicons were scored as presence (1) or absence (0) of bands in the form of a binary matrix. Only reproducible and clear bands were counted. Monomorphic and polymorphic bands were calculated for each ISSR primer. The polymorphic information content (PIC) was determined using the following equation: PIC = 1 -p 2 -q 2 , where p is the band frequency and q is no-band frequency [49]. Genetic distances were calculated for all accessions according to Nei and Li [50]. Cluster analysis was performed with the UPGMA algorithm (unweighted pair group method with arithmetic average) and the Treecon ver. 3.1 software [51]. To estimate the statistical support of the dendrogram branches, the data were bootstrapped using 2000 replications, and values higher than 50% were mapped ( Figure 1).

Figure 1
The UPGMA (unweight pair group method with arithmetic average) dendrogram computed using genetic distance matrix based on ISSR markers. Only bootstrap values > 50% are indicated. Scale indicates genetic distance, and the values in pg represent genome size (2C) of the particular species.

Genome Size
The genome size of the Malva accessions was different and ranged from 1.45 pg/2C in M. parviflora to 11.06 pg/2C in M. alcea (Table 2 and Figure 2), which corresponds to 1418 and 10,817 Mbp, respectively. The mean nuclear DNA content (2C) determined for all Malva accessions reached 3.23 pg/2C, which relates to 3159 Mbp. In three varieties of M. varticillata, the differences between the

Genome Size
The genome size of the Malva accessions was different and ranged from 1.45 pg/2C in M. parviflora to 11.06 pg/2C in M. alcea (Table 2 and Figure 2), which corresponds to 1418 and 10,817 Mbp, respectively. The mean nuclear DNA content (2C) determined for all Malva accessions reached 3.23 pg/2C, which relates to 3159 Mbp. In three varieties of M. varticillata, the differences between the smallest and the highest 2C DNA content was 0.13 pg. The obtained data of genome size for all accessions were separated into three groups according to the classification of Soltis et al. ( [49]; Table 2). Six accessions were classified into a group with very small genomes, with 2C DNA content ranging from 1.

ISSR Markers
Twelve Malva genotypes were examined with 13 ISSR primers, which revealed unambiguous and reproducible band patterns. The primers yielded 243 polymorphic loci, with the size of the bands ranging from 212 (ISSR-1) to 2392 (ISSR-7) bp. The percentage of polymorphism amounted to 100% for all investigated primers ( Table 3) (Table 3). The ISSR-4 and ISSR-42 primers revealed to be useful to distinguish all the investigated species, while the ISSR-5 and ISSR-6 primers enabled the discrimination of species and also M. verticillata varieties (Figure 3).

ISSR Markers
Twelve Malva genotypes were examined with 13 ISSR primers, which revealed unambiguous and reproducible band patterns. The primers yielded 243 polymorphic loci, with the size of the bands ranging from 212 (ISSR-1) to 2392 (ISSR-7) bp. The percentage of polymorphism amounted to 100% for all investigated primers ( Table 3) (Table 3). The ISSR-4 and ISSR-42 primers revealed to be useful to distinguish all the investigated species, while the ISSR-5 and ISSR-6 primers enabled the discrimination of species and also M. verticillata varieties (Figure 3). The genetic distance was calculated to reflect the relationships between the tested accessions. The close relationship between Malva accessions was expressed by a low degree of the genetic distance. The values of genetic distance between the studied accessions ranged from 0.22 to 0.81 (Table 4)  The genetic distance was calculated to reflect the relationships between the tested accessions. The close relationship between Malva accessions was expressed by a low degree of the genetic distance. The values of genetic distance between the studied accessions ranged from 0.22 to 0.81 (Table 4)

Discussion
Distinguishing Malva species from each other and closely related genera is mainly based on morphological analysis. However, within the genus, some species (e.g., M. parviflora, M. pusilla, M. nicaeensis, and M. neglecta) are regularly misidentified based on morphological features [38]. Moreover, the phylogenetic relationships and taxonomic organization of the Malva genus are still unclear. Therefore, molecular analysis can be useful in species identification and study of the genetic relationships between Malva species/subspecies/varieties. Flow cytometry was used for nuclear DNA content estimation as a first step of Malva species identification and genetic characterization. To the best of our knowledge, this is the first study to determine the 2C DNA content for eight of the studied accessions. The 2C DNA content measured for all accessions ranged from 1.45 to 11.06 pg, with more than half the accessions (54%) possessing very small genomes. Most of the studied Malva accessions are hexaploids with small and numerous chromosomes, which makes chromosome counting challenging, so the published numbers (40-44) could be incorrect [3]. This means that it is possible several hexaploid species have the same chromosome number. In this study, the hexaploid level was observed for all accessions with very small genomes pg/2C) based on flow cytometric analysis. Therefore, to identify all studied accessions and to investigate genetic relationships between them, ISSR molecular markers were applied. The usefulness of these markers in species identification and detection of genetic diversity have been proven in studies of several other herbal genera (e.g., Trigonella L., Thymus L., Ocimum L., Origanum L., Mentha L.; [45,[60][61][62][63]). Until now, molecular studies using ISSR markers conducted in the Malva genus have only included a few species [31,32]. All primers used in ISSR-PCRs for the Malva genus revealed 100% polymorphism between all accessions. Therefore, it was possible to identify all tested species. Moreover, for M. verticillata taxon, it was possible to distinguish all studied varieties. The usefulness of most of the used ISSR primers was also confirmed in Ocimum, Origanum, and Mentha identification [45,62,63].
The systematics of the Malva genus and closely related genera is complicated. Moreover, the relationships obtained from molecular studies do not confirm traditional classification [3,29]. So far, only molecular analysis relying on rDNA ITS sequences and ISSR markers have shed light on taxonomical relationships between Malva species [3,[29][30][31]. Phylogenetic analyses of rDNA ITS sequences indicated the presence of two well-supported clusters within the mallow species (malvoid and lavateroid clades), which is consistent with the presented data. In this study, malvoid clade (I) was represented by five species: M. neglecta, M. nicaeensis, M. parviflora, M. sylvestris var. mauritiana, and M. verticillata. Moreover, within this cluster, two subgroups were detected. The first subgroup contained all M. verticillata varieties, while the second subgroup included two species: M. nicaeensis and M. parviflora. An earlier study reported that the two latter species are difficult to identify using morphological features [39]. Our work also confirmed a close relationship between M. parviflora and M. nicaeensis. Both M. sylvestris var. mauritiana and M. neglecta were placed outside of the subgroups. The lavateroid group (II) consisted of five species: M. cretica ssp. althaeoides, M. aegyptia, M. alcea, M. moschata, and M. tourneforiana. Within this group M. alcea together with M. moschata and M. tournefortiana created one subcluster, which confirms earlier phylogenetic analysis [29]. M. aegyptia and M. cretica ssp. althaeoides were placed outside of this subgroup. An earlier phylogenetic study revealed problems in the affiliation of those two species [29], while seed image analysis grouped them into separate Bibracteolatae sections [42]. Within the investigated genotypes, M. sylvestris var. mauritiana and M. alcea were the most genetically diverse. Nevertheless, the phylogenetic data presented in this study mostly agree with the division of mallow species proposed by Escobar Garcia et al. [29] on malvoid and lavateroid sections of the Malva genus. The data are also consistent with the study of Celka et al. [31], although the authors proposed the affiliation of the two clusters to Malva and Bismalva sections.

Conclusions
This study proved the potential of flow cytometric genome size determination and ISSR marker analyses in the identification and evaluation of genetic relationships between Malva species/varieties. Out of the 12 studied accessions, it was possible to identify six genotypes using flow cytometry, whereas all species and varieties were identified by the ISSR-PCR method. This is the first report about the application of these two techniques to estimate the genetic variation and genetic characterization of Malva accessions. The results of this study can be useful in mallow breeding programs, planning of conservation strategies, germplasm collection, and taxonomy of the genus.