Background

Since thousands of years, medicinal plants have been used in health-care systems [1, 2]. The correct identification of these plants is required for their safe use. Medicinal plants were traditionally classified by experienced experts based on morphological characteristics. However, it may take an expert many years to accurately identify these plants, and only a few taxonomists have been able to do so for around 1000 species so far [3]. Furthermore, the number of trained experts has been decreasing as the workload on identification has increased. Simultaneously, the number of adulterants and substitutes on the market is increasing, resulting in variation in the quality of medicinal plants [4, 5]. Thus, identifying medicinal plants is a significant challenge, posing safety risks to both the traditional medicine industry and consumers. As a result, a quick and accurate method for authenticating these plant species is required. DNA barcoding is a novel method for species identification that has the capacity to overcome the aforementioned challenges [3].. This method is based on sequence differences within short, standardized DNA regions, with the primary goal of both authenticating and discovering new species [3, 6, 7]. DNA barcoding is an efficient and low-cost technique that could boost the work of specialists while also making species validation accessible to nonspecialists [8,9,10,11]. Furthermore, it is independent of both morphological characteristics and expert experiences, allowing for faster, more subjective, and accurate results. As a result, DNA barcoding has quickly become the focus of research, and it has been recognized as a powerful tool for species identification and the construction of phylogenetic trees [12, 13].

The Apiaceae family contains approximately 250–440 genera and a range of 3300–3700 species that are widely distributed throughout the temperate zones of both hemispheres. Selinum belongs to Apiaceae is perennial, aromatic, and medicinal plant ranging from 1 to 2 m in length [14]. Selinum species are mainly found in Western Himalayas and Himalayan foothills of India, West Pakistan, Bhutan, China, and Nepal. Some of its species have also been reported from Europe, specifically in the UK, Poland, Bosnia, and Croatia. Ecological and floral diversity type studies so far do not present the molecular characterization of the Selinum spp. [15, 16]. Molecular characterization may further provide database for intensive pharmacological studies in Selinum.

It has proven difficult to classify species within the Selinum genus because there are no clear morphological synapomorphies and the genus is plagued by synonymies, in which a species is referred to by a different name somewhere else [17]. Several other Apiaceae genera, such as Angelica L., Ferula L., and Coriandrum sativum L., are also treated separately; however, no phylogenetic studies are available in Selinum.

Two species of Selinum, viz S. tenuifolium Wall ex C. B. Clarke (Milk Parsley) and S. vaginatum C. B. Clarke (Bhutkeshi), have been identified on the basis of morphology in India. The National Center for Bioinformatics database and UniProt database reveal that studies have only been done on specified regions of the gene or proteins of the study with pharmaceutical or cosmetic motives. However, they do not provide molecular characteristics of Selinum spp. [18]. Currently, rbcL genes are widely used for phylogenetic analysis within angiosperm families and subclasses, as well as among seed plant groups. However, matK evolves 3 times faster than rbcL in angiosperms, potentially providing more phylogenetic information and improved phylogenetic resolution among the taxa studied [19, 20]. The present study therefore aims to assess the molecular characteristics of Selinum spp. using matK and rbcL plastid genes in order to derive a phylogenetic relationship within the sp. of Selinum as well as with other members of Apiaceae.

Methods

Sample collection

The plants were sampled from twelve accessions of Selinum collected from different locations representing two states in the Northern Himalayan region of India, namely Himachal Pradesh (HP) and Uttarakhand (UK). Seven S. vaginatum C. B. Clarke accessions were collected from various locations, viz SV1 — S. vaginatum, Chipla kedar, 2691 m (UK); SV2 — S. vaginatum, Rohtang valley, 2137 m (HP); SV3 — S. vaginatum, Dhartula, 2378 m (UK); SV4 — S. vaginatum, Cheena Peak, 1937 m (UK); SV5 — S. vaginatum, Dhartula, 1819 m (UK); SV6 — S. vaginatum, Chipla Kedar, 2236 m (UK); and SV7 — S. vaginatum, Gulaba 2781 m (HP), and five accessions of S. tenuifolium Wall ex C. B. Clarke were collected from different locations (ST1 — S. tenuifolium, Kothi, 2158 m (HP); ST2 — S. tenuifolium, Rohtang valley, 2350 m (HP); ST3 — S. tenuifolium, Chopta trek, 2150 m (UK), ST4 — S. tenuifolium, Chopta trek, 2593 m (UK), ST5 — S. tenuifolium, Chopta trek, 3178 m (UK)) and stored at −4 °C for molecular studies.

For two target genes, matK and rbcL, species disparity was accomplished through BLASTn analysis and phylogenetic inference based on maximum likelihood and neighbor-joining method. The disparity index was used to assess the dependability of phylogenetic tree regenerations. The maximum composite likelihood with Kimura 2-parameter model was used to estimate evolutionary divergence between and within all accessions.

Laboratory reagents, plasticware, and glassware

SIGMA (Sigma-Aldrich, Inc., 3050 Spruce Street, St. Louis, MO, USA) and Gyan Scientific analytical grade reagents and chemicals were used in this study (Lucknow, India). Borosil glassware and Axygen plasticware were used.

Sterilization of plasticware and glassware

The glassware and plasticwares used for our experiment purpose were washed, cleaned, and sterilized as per the standard protocol recommended for molecular biology researches. All the glassware and plasticware were autoclaved before use, and UV (ultraviolet) irradiation was used as a standard laboratory practice to decontaminate the apparatus. Carryover of DNA from such a previous amplification of the same target almost always results in false-positive amplification. To prevent this, PCR reactions were set up in a separate room than that used for post-PCR manipulations. Autoclaved deionized water (Millipore ZLXS50034, Millipore Bangalore, India) was used for conducting all experiments.

Genomic DNA extraction

Five S. tenuifolium accessions and seven S. vaginatum accessions were collected, and their genomic DNA was isolated from young leaves and used as a template for a PCR reaction using a DNA isolation kit (QIAGEN, DNeasy® Plant Handbook). With the respective universal primers, the target DNA regions, namely matK and rbcL, were amplified.

Bioinformatics tools

The software used for analyzing the chromatograms of amplicons generated after sequencing by ABI 3130 Genetic Analyzer (Applied Biosystems Inc., USA) was as follows:

  1. a)

    BioEdit program version 3 (Naim & Mahboob. 2020) [21]

  2. b)

    Tamura 3-parameter model (Olmstead & Palmer 1994) [22]

  3. c)

    MEGA version 6.06 (Hartl & Frost 1999) [23]

Barcoding gene for the study

To evaluate intraspecific and interspecific variations among the accessions of genus Selinum, two genes of DNA barcodes (rbcL, matK) were selected to identify diversity among selected species.

PCR amplification with matK and rbcL primers

A reaction mixture containing 50 μl of Taq PCR Master Mix to conduct PCR, 22 μl of distilled water, 1 μl of forward primer (10 μM), 1 μl of reverse primer (10 μM), and 1 μl of DNA template (50–80 ng/l) was used to conduct PCR. The PCR conditions are mentioned in Table 1, and primer details are given in Table 2.

Table 1 PCR condition
Full size table
Table 2 matK and rbcL primer details
Full size table

The PCR products were tested on a 1.5% agarose gel with ethidium bromide in 1X TAE buffer (pH 8.0) to ensure that they were correct. Molecular Imager® GelDocTMXR software was used to examine and archive the gel. With Quantity One software, bands were scored and evaluated (Bio-Rad). By comparing the size of the products to 500 bp DNA Ladder H3 RTU, the size of the products was determined (GeneDirex, cat no. DM003-R500) (Figs. 1 and 2). The nucleotide sequence was converted into an amino acid sequence after the PCR products were verified by electrophoresis, purified, and sequenced using the following comprehensive sequencing procedures (Table 3). The nucleotide and amino acid sequences were aligned using the system software aligner tool to produce a phylogenetic tree based on two statistical data analyses, bootstrapping and pairwise distance, from both nucleotide and amino acid sequences. Table 4 shows the sequencing and features in greater detail.

Fig. 1
figure 1

Agarose gel showing amplification of rbcL gene

Full size image
Fig. 2
figure 2

Agarose gel showing amplification of matK gene

Full size image
Table 3 Sequencing reaction
Full size table
Table 4 Sequencing and chemistry details
Full size table

Results

Evolutionary and phylogenetic analysis

After DNA sequencing, the twelve accessions of Selinum collected from different locations and their molecular sequences of matK and rbcL markers were evaluated using BLAST method. As a statistical method and a phylogeny test, aligned sequences were used to create phylogenetic trees using the maximum composite likelihood model and bootstrap resampling, respectively. Gene sequences from the twelve Selinum accessions were used to create the most likely phylogenetic trees. The numbers at branch nodes are bootstrap values, which show the percentage of bootstrap iterations that support the tree at that specific point of divergence for each phylogenetic tree node. The more the topology of the phylogenetic tree is supported, the higher the bootstrap value. In the phylogenetic tree, species that are adjacent to each other are closely linked. ST5, SV4, ST4, and SV5 have been demonstrated to be closely related (cluster 1) and have a moderate bootstrap support value of 49, whereas ST2, ST1, ST3, SV2, SV6, and SV7 have been grouped together (cluster 2) and have a bootstrap value of 41, showing that this grouping is poorly supported. ST5 is a distant common ancestor of all the accessions. ST3, SV3, SV6, ST4, and SV5 have also been grouped together (cluster 3), while SV7, SV1, SV4, ST1, ST2, and SV2 have been grouped together (cluster 4) based on rbcL gene sequences, and both groupings are related to SV1 and SV3, which are clustered separately. Cluster 3 has a moderate bootstrap support rating of 43, whereas cluster 4 has a high bootstrap value of 53, indicating that it is strongly supported. The evolutionary trees of matK and rbcL are surprisingly similar. ST5, SV4, ST4, and SV5 have been grouped together based on matK genes, and all four accessions form a monophyletic group (cluster 1). The phylogenetic tree built using the rbcL genes shows the similar pattern (cluster 4). ST2, ST1, ST3, SV2, SV6, and SV7 have all been demonstrated to be closely related using matK genes (cluster 2). Such species’ close association can also be seen in phylogenetic trees based on the rbcL genes (cluster 4). ST2 has been shown to be closely associated with ST1, ST3, SV2, SV6, and SV7 (cluster 2) based on matK genes, whereas SV7 has been shown to be closely related to SV1, SV4, ST1, ST2, and SV2 (cluster 4) based on rbcL genes, as shown in Fig. 3.

Fig. 3
figure 3

Comparative analysis of 12 different accessions of Selinum from matK and rbcL markers

Full size image

Assessment of genetic diversity in Selinum using matK and rbcL markers

Nucleotide variation

The length of the matK gene sequence ranged from 740 to 855 base pairs (bp), while the rbcL gene ranged from 582 to 638 bp. DNA base composition represented as G + C content, where G + C percentage in all species is 40% on average. AT (adenine-thymine) was dominant over GC (guanine-cytosine) in the composition of nucleotides from both genes. In rbcL, GC content was 56% and AT was 44%, while in matK, GC content was 65% and AT was 35%, respectively (Fig. 4 and Table 5).

Fig. 4
figure 4

GC content distribution of matK (upper panel) and rbcL (lower panel) in sequences gene

Full size image
Table 5 Nucleotide variation in matK and rbcL chloroplast genes in Selinum species
Full size table

Phylogenetic relationship between the Selinum species

Based on partial rbcL gene sequences and matK gene sequences, the tree construction findings were midpoint rooted. Molecular phylogenetic trees for the rbcL and matK gene constructed using the maximum likelihood (ML) methods showed approximately identical clustering (Fig. 5). The trees showed that two main divisions of the chosen accessions from different locations of Selinum group were clustered together. Four clusters emerged from a cluster analysis of data gathered from the matK gene in twelve Selinum accessions from various regions. Angelica sylvestris was in the first cluster (ST1). The second cluster included Peucedanum ostruthium (ST4), Peucedanum ostruthium (SV5), and Angelica sylvestris (ST2). The third cluster included Seseli libanotis (ST3) and Peucedanum ostruthium (SV4), while the fourth cluster included Angelica sylvestris (ST5), Seseli libanotis (SV6), Seseli libanotis (SV7), Angelica sylvestris (SV2), Angelica sylvestris (SV3), and Peucedanum ostruthium (SV1) as shown in Fig. 5, and cluster was given in Table 6.

Fig. 5
figure 5

1000 bootstrap replications of phylogenetic tree of partial matK gene sequences

Full size image
Table 6 Selinum cluster ID and names for matK
Full size table

Clustering analysis from the rbcL gene, using S. vaginatum as the out-group and rbcL phylogenetic trees, found two primary clades and one out-group clade. Seseli libanotis is the first clade (SV1). Clade two consist of four accessions: Peucedanum palustre (ST5), Peucedanum palustre (ST1), Peucedanum palustre (SV2), and Peucedanum palustre (ST2), while clade 3 consists of seven accessions, Peucedanum palustre (ST3), Peucedanum palustre (SV7), Peucedanum palustre (SV6), Peucedanum palustre (SV5), Peucedanum palustre (ST4), Peucedanum palustre (SV4), and Peucedanum palustre (SV3) as shown in Fig. 6 and Table 7.

Fig. 6
figure 6

Using 1000 bootstrap replications, phylogenetic tree of partial rbcL gene sequences

Full size image
Table 7 Selinum cluster ID and names for rbcL
Full size table

Inter- and intraspecific relationship between matK and rbcL

Based on phylogenetic analysis, twelve accessions, including ST1, ST2, ST3, ST4, ST5, SV1, SV2, SV3, SV4, SV5, SV6, and SV7 from different geographical locations of India, were slightly different from all other accessions and were grouped together (Fig. 3). The samples were also divided into two clades using the maximum likelihood method. Clade 1 was made up of a monophyletic clade that included all of the samples under consideration. The first subgroup was further split into two clades. Clade 1 nested all the matK samples under it, while only a single rbcL sample was under clade 2. Similarly, group 2 also has two main clusters, both having largely rbcL samples.

The intraspecific and interspecific analysis of twelve accessions from different locations of Selinum infer that the intraspecific variation between the accessions is very less and interspecific variation as exhibited by matK and rbcL marker showed variations of different degrees among the species (Fig. 7). matK markers showed maximum interspecific variation (97%) and rbcL markers only 3% (Fig. 8). Intraspecific variation was also maximum in matK (61%) marker fallowed by rbcL (39%) (Fig. 9).

Fig. 7
figure 7

Interspecific and intraspecific relationships between matK and rbcL gene of twelve accessions of S. tenuifolium and S. vaginatum collected from different locations

Full size image
Fig. 8
figure 8

Interspecific relationships between matK and rbcL gene of twelve accessions of S. tenuifolium and S. vaginatum collected from different locations

Full size image
Fig. 9
figure 9

Interspecific relationships between matK and rbcL and between matK and rbcL gene of twelve accessions of S. tenuifolium and S. vaginatum collected from different locations

Full size image

Discussion

The current study found that the DNA identification approach appropriately differentiated all samples up to species level, and that matK and rbcL are reliable identification markers [21]. It is an accurate method to amplify and sequence, and plastid DNA has played a significant role in building evolutionary relationships and defining limits of species in plants. It is also mostly uniparentally inherited and appears to have little to no recombination, resulting in a small size, structure, and gene order [22]. It is critical to understand the causes of evolution by which processes of genetic polymorphisms across species are converted into genetic divergence between species and genetic diversity by investigating the quantity of nucleotides and patterns of nucleotide variation between and within species. Evolutionary mechanisms, such as selection, recombination, mutation, and population structures, are affected by these diversities [23].

An earlier DNA barcoding research of Chenopodium murale produced similar results. According to their findings, the specimen was identified as C. murale using BLAST for both the matK and rbcL genes with 100% sequence matching. Using BOLD, the rbcL gene showed a high degree of similarity to various taxa, including Chenopodium ambrosioides, Chenopodium album, and Chenopodium ficifolium, ranging from 96.3 to 100% [24]. The DNA barcode is clearly distinguish between species, with no overlap between intraspecific and interspecific divergence ideally [25]. Furthermore, the effectiveness of any DNA barcoding approach is determined by the degree of intraspecific and interspecific divergence in a single locus or a set of loci [26]. According to the China Plant BOL Group (2011), the combination rbcL + matK separates 40% of the sampled species in the matched dataset for the Apiaceae family. Furthermore, among all samples, rbcL and matK had 32.1% and 38.6% species identification efficiencies, respectively [27]. According to previous research, both rbcL and matK can be used to identify plants at the species level, because both have a high amplification and sequencing success rate [28]. These two fragments have high identification success rates at the genus and family level [29]. Previous studies reported that genetic distance across accessions could be determined, and that large differences in dissimilarity values could be attributed to genetic differences [30]. For primers matK and rbcL, for instance, grouping based on genetic distance revealed that one group contains different accessions, while the other contains accessions showing wide variations [31]. Plant DNA barcoding is best accomplished by combining coding and noncoding genetic markers. The highly conserved rbcL gene and the more variable matK gene are the most widely explored markers in many researches [32, 33].

In this study, all twelve accessions from different locations of Selinum were examined, and they all belong to the parsley family, Apiaceae. Five-petalled white flower subgenera were easily distinguished based on the anatomical structure of the leaf. Despite the fact that several morpho-anatomical classification methods for the sections and subsections of subgenus Selinum as well as subgenus Angelica have been offered, the links between the subsections and their evolutionary processes are still being contested.

India is a rich source of medicinal plants, and there is no data to identify these plants on molecular basis. Selinum is a medicinal plant with potential therapeutic value. This study has been conducted to identify Selinum using molecular markers because as we go from lower to higher altitude, there are certain changes in morphology and growth within plants which may not be morphologically evident but exist at genetic level. In India, only two species of Selinum, viz S. tenuifolium Wall ex C.B. Clarke and S. vaginatum C. B. Clarke, have been identified on the basis of morphology without any molecular basis ever utilized to authenticate the morphological markers for proper identification of these plants. Various synonyms have been found in Selinum and the ambiguity regarding species with various synonyms. Selinum species are also hard to distinguish morphologically, because certain species have been given different names by different authors instead of a single genuine name. Therefore, DNA barcoding in this plant would be appropriate tool to establish the correct identification of this species or to differentiate species or genus from other members of this family.

The use of morphological features alone is insufficient to adequately define the genus. When investigating interspecific variations, molecular, anatomical, and biochemical markers are more useful than gross morphological ones. A phylogenetic tree can be formed by comparing the same gene sequence across species within a genus, which can support or provide fresh insights into the present taxonomy and rule out the confusion caused by synonymy in Selinum.

Conclusion

The species boundaries of S. tenuifolium and S. vaginatum were assessed using phylogenetic tree using twelve accessions obtained from various places in the Indian Himalayas of Himachal Pradesh and Uttarakhand. The sequencing data collected from the samples used during this study together with other samples retrieved from NCBI were analyzed phylogenetically by using maximum likelihood methods resolved essentially identical topologies. The trees that resulted from this method of analysis provided strong evidence of phylogenetic relationships among the samples investigated in the present study. The samples that were being investigated consistently grouped together forming a monophyletic clade with very strong support, suggesting that the investigated samples and Selinum evolved from a recent common ancestor of Apiaceae family. This finding implies that the genetic divergence in Selinum is phylogenetically informative at the species level. Despite the fact that all of the samples under research were monophyletic, paraphyly was notably noticeable between the samples under investigation and the other Selinum samples obtained from NCBI. In reference to the altitude, the genetic diversity maintained a threshold level of only 3% margins encoded by the comparative analysis of matK and rbcL.