Karyology and Genome Size Analyses of Iranian Endemic Pimpinella (Apiaceae) Species

Pimpinella species are annual, biennial, and perennial semibushy aromatic plants cultivated for folk medicine, pharmaceuticals, food, and spices. The karyology and genome size of 17 populations of 16 different Pimpinella species collected from different locations in Iran were analyzed for inter-specific karyotypic and genome size variations. For karyological studies, root tips were squashed and painted with a DAPI solution (1 mg/ml). For flow cytometric measurements, fresh leaves of the standard reference (Solanum lycopersicum cv. Stupick, 2C DNA = 1.96 pg) and the Pimpinella samples were stained with propidium iodide. We identified two ploidy levels: diploid (2x) and tetraploid (4x), as well as five metaphase chromosomal counts of 18, 20, 22, 24, and 40. 2n = 24 is reported for the first time in the Pimpinella genus, and the presence of a B-chromosome is reported for one species. The nuclear DNA content ranged from 2C = 2.48 to 2C = 5.50 pg, along with a wide range of genome sizes between 1212.72 and 2689.50 Mbp. The average monoploid genome size and the average value of 2C DNA/chromosome were not proportional to ploidy. There were considerable positive correlations between 2C DNA and total chromatin length and total chromosomal volume. The present study results enable us to classify the genus Pimpinella with a high degree of morphological variation in Iran. In addition, cytological studies demonstrate karyotypic differences between P. anthriscoides and other species of Pimpinella, which may be utilized as a novel identification key to affiliate into a distinct, new genus – Pseudopimpinella.


INTRODUCTION
The genus Pimpinella is one of the largest genera in the family Apiaceae, subfamily Apioideae, and tribe Pimpinelleae, with approximately 170-180 species distributed throughout Europe, Asia, Africa, and South and North America (Pimenov and Leonov, 1993;Pu and Watson, 2005). The members of this genus are annual, biennial or perennial, semibushy, and aromatic plants with cordate-ovoid or oblong-ovoid fruits with five filiform ribs on each cordate-ovoid or oblongovoid leaves (Pu and Watson, 2005). Nearly 23 species of this genus are grown in Iran, six of which are endemic (P. pastinacifolia, P. tragioides, P. khorasanica, P. deverroides, P. khayyamii, and P. anisactis) (Mozaffarian, 2003). Their habitats are dry slopes, rocky crevices, fields, meadows, mountain pastures, grasslands, steppes, and dry open woodlands 1,000-2,200 m above sea level.
Studies suggest that the genus Pimpinella is highly diverse, and the taxonomic delimitation of the genus has not yet been resolved (Zhou et al., 2008(Zhou et al., , 2009Downie et al., 2010). The last revision of the genus was made by Wolff (1927) based on the petal color, fruit and petal vestiture, and life history. The genus was subdivided into three sections: Reutera, Tragium, and Tragoselinum. It has since been realized that morphological markers do not explain the systematic relationship among Pimpinella species (Fereidounfar et al., 2016). Hence, investigation of various aspects, including karyological observations and genome size estimates, may be useful in establishing systematic and evolutionary relationships, resolving taxonomic ambiguities, and gaining a better understanding of the way they diverged from each other (Dobigny et al., 2004;Knight et al., 2005;Bancheva and Greilhuber, 2006;Guerra, 2008;Bainard et al., 2013).
Nuclear DNA content is under strict genotypic control within the defined limits. Thus, it appears that such a variation correlates with evolutionary and systematic considerations Greilhuber et al., 2005;Knight et al., 2005;Doležel et al., 2007;Bainard et al., 2013). Variation of intra/interspecific genome size may reflect karyotypic differences, such as differences in the case of chromosome number and size (Bennett et al., 2008). Greilhuber et al. (2005) the DNA content of the unreplicated haploid chromosomal supplement, n, (1 C-value), and the amount of DNA per basic chromosome number, x, (1 Cx-value, regardless of generative polyploidy, aneuploidies, or other factors). Variation in chromosome number in the Pimpinella genus can indicate intra-and inter-specific differences in genomic DNA quantities.
No detailed information is available regarding the DNA C-value, karyology, and ploidy levels of Pimpinella species, as cytological investigations have mainly concentrated on reporting chromosome numbers. Therefore, this research reports for the first time the karyotype criteria and genome size of 16 Iranian species of Pimpinella.

Plant Materials
The seeds of 16 Iranian endemic species of Pimpinella were collected during the growing season in their natural habitats from different locations in Iran. Only S8 was collected from two geographical locations. The species code and geographical descriptions, including latitude, longitude, altitude (m), mean temperature ( • C), and mean rainfall (mm), are shown in Table 1 and Figure 1.

Cytogenetic Analysis
Actively growing roots of approximately 1-2 cm were cut and pretreated in a 0.002 M solution of 8-hydroxyquinoline for 4 h at 25 • C and fixed in ethanol: glacial acetic acid (3:1, v/v) for 24-36 h at 4 • C. Chromosome preparations from root tip cells were performed, as described by Abdolmalaki et al. (2019). For each species, ten root tips were flooded with ice-cold water twice (5 min each time), followed by 0.01 M citrate buffer twice for 5 min. Between 1 and 1.5 mm of the root tips (meristematic parts) were then digested with a 30 µl enzyme mixture containing 1% pectolyase (Sigma P3026), 0.7% cellulose (CalBiochem219466), 0.7% cellulose R10 (Duchefa C8001), and 1% cytohelicase (Sigma C8274) dissolved in 0.01 M citrate buffer with pH 4.8 for 1 h. After digestion, meristems were washed twice with citrate buffer (5 min each time) and once with ethanol for 5 min to remove the enzyme mixture. Ethanol was changed with a 70 µl fixative solution (9: glacial acetic acid/1: absolute methanol). The root tips were carefully taped using a dissecting needle until a cell suspension formed. Seven microliters of the cell suspension were then dropped onto each glass slide in a box lined with 50% humidity; it was left to dry slowly and stored in 70% ethanol. A drop containing 1 µg/ml DAPI (4 , 6-diamidino-2-phenylindole) was added to the cell area, and a coverslip was applied. High-resolution chromosome images were taken using a Nikon A1Si Laser Scanning Confocal Microscope (Nikon Instruments Inc., Japan). Chromosome measurements and karyotypic features were studied based on five well-prepared metaphase plates from different individuals.

Flow Cytometric Assessment (FCM)
The 2C-DNA content of each Pimpinella species was estimated using flow cytometry (FCM). Flow cytometry experiments were performed using the propidium iodide (PI) staining method. A leaf of Solanum lycopersicum cv. Stupick with a 2C DNA value of 1.96 picograms (pg) (Doležel et al., 1992) was used as an internal reference standard. In brief, 1 cm 2 of leaves of each Pimpinella species, along with 1 cm 2 of the young leaves of the standard, were used to nuclei isolate by chopping with a sharp razor blade in 1 mL of woody plant buffer (Loureiro et al., 2007) in a Petri dish, supplemented with 50 µg ml −1 propidium iodide (PI), polyvinylpyrrolidone (PVP 10) and 50 µg ml −1 RNase. The nuclear suspension was passed through a 30 µm mesh nylon filter and then analyzed using a Cyflow Space flow cytometer (Partec GmbH, Münster, Germany) equipped with a 532 nm green highgrade solid-state laser. For each species, 5,000-10,000 nuclei per G1 peak were measured for DNA content estimates. Five different individuals per species were analyzed using linear amplification. Histograms with a coefficient of variation (CV) lower than 3% were evaluated using the FlowJo software (Version 10.6.2, Treestar, Ashland, OR, United States). Nuclear DNA content was calculated according to the following formula: Sample 2C DNA content (pg) = (Sample G 1 peak mean/standard G 1 peak mean) × standard 2C DNA amount (pg). As Doležel et al. (2003) stated, picogram values were converted to megabase pairs (Mbp), in which 1 pg of DNA represents 978 Mbp.

Statistical Analysis
The data were subjected to variance analysis (ANOVA) using the GLM procedure of the SAS software (SAS, 2003) based on a completely randomized design (CRD) with five replications for both flow cytometry and karyological data. In both cases, the normal distribution of residuals and the homogeneity of variances were approved. For mean comparisons, Tukey's test was utilized (Seijo and Fernández, 2003). Multivariate statistical analysis (Srivastava, 2002) was carried out in the Minitab software package (Minitab 17, 2010) on standardized data (mean = 0, variance = 1). A principal component analysis (PCA) was performed based on a data matrix to estimate the participation of the karyotypic parameters in the species classification (Mirzaghaderi et al., 2010). Based on karyotypic parameters, cluster analysis was carried out using the unweighted pair-group method arithmetic mean (UPGMA) and the Euclidean distance (Abedi et al., 2015). The cophenetic correlation coefficient (r) was computed to specify the goodness-of-fit of the clusters to the original data. Dot plots of mean 2C-values and means of karyotypic TCV and X values were generated, reflecting the presence of 4C DNA in a metaphase cell during mitotic division.
In conclusion, four (S%, DRL%, CV%, and DI%) among the five karyotypic symmetrical groups tested confirmed that among all 17 Pimpinella populations examined, S12 and S9 appear to have the most asymmetrical and symmetrical karyotypes, respectively.
it is not. Hence, another explanation for this chromatin body would be that it is a B-chromosome.
The UPGMA dendrogram -constructed using the matrix of karyotype similarities (Figure 4) -displayed four major clusters. The first cluster is comprised of S1, S2, S5, and S10, distinguished by the shortest complements but high RL and AR. In this cluster, species with 18 chromosomes form a subgroup. The second cluster comprises 11 species with both 20 and 22 chromosomes, which are characterized by a high L, F%, and TL. The S16 species, with 40 chromosomes with the lowest F%, form the third cluster. The fourth cluster contained S8E2, characterized by a high S, CI%, r-value, and the lowest AR.
The first two PCs in the karyotypic parameter's principal component analysis (PCA) account for 80.8% of the cumulative variation, and they were shown in a 2-dimensional image (Figure 5). The first component (50.9%) emphasizes the position of the centromere, while the second component (29.9%) accentuates variation in complement length. The resulting species arrangement from this test entirely matches with those obtained from the UPGMA clustering method.
Seventeen populations of 16 Pimpinella species were analyzed using FlowJo software version 10.6.2 to calculate the amount of DNA in the nucleus. The acquired histograms for estimating the nuclear DNA content of each species contained two peaks; the right peaks refer to the known internal reference standard (Solanum lycopersicum cv. Stupick), and the left peaks refer to the unknown Pimpinella species (Figure 6). Table 3 shows DNA content (pg) and the genome sizes (Mbp) of studied Pimpinella species. The 2C-value varied from 2.48 to 5.50 pg (relating to a diploid P. affinis and a tetraploid P. rhodantha, respectively). Among the 16 diploids with chromosome numbers ranging from 18 to 24, a distinction of 0.68 pg in 2C-value (2.48 and 3.16) was observed (Table 3), while the mean 2C-value of a P. rhodantha (2n = 40, 5.50 pg) was determined to be precisely double the value of the two diploids with x = 10 (2n = 20, 2.75 pg). Tukey's test revealed a significant difference in the 2C-value among the 16 diploid species. Meanwhile, the holoploid genome size ranged from 1212.72 Mbp (diploid S1) to 1511.01 Mbp (diploid S13), with a difference of 298.29 Mb, whereas the tetraploid haploid genome size was 2870.43 Mbp. Significant correlation was observed between chromosomal features measured, viz, X, and TCV (r = 0.879** and 0.701**, respectively, in Figures 7A,B), with 2C-values demonstrating linear relationships (b = 0.034** and 1.897**, in Figures 7A,B, respectively).
The differences in chromosome number and chromosome morphology found among species indicate that chromosome structural changes may be used to distinguish species that are very similar to each other and cannot be separated using morphological characters. Among these species, P. khorasanica differs from P. anisactis due to the presence of two "sm" chromosome types and possesses a shorter total chromatin length. These relative differences could be used for breeding programs by facilitating chromosome identification in hybrid populations and derivatives in Pimpinella. Parent combinations with relative differences in the chromosome numbers/type due to chromosome pairing could result in a successful cross (Hamidi et al., 2018;Akbarzadeh et al., 2021). Three species -S3 (P. tragium), S13 (P. anisactis), and S15 (P. khorasanica)are very similar to each other from the viewpoint of anatomy, and we are not able to separate them with anatomical features. In addition, they are morphologically similar to each other. Therefore, the findings of this study can be considered as a guide to separating these species, especially S13 (P. anisactis) and S15 (P. khorasanica), which are distributed in a small area in Khorassan, where they are endemic. Taxonomic criteria, including chromosome number and asymmetric indices, have been used in plant phylogenetic and taxonomic consideration of the genus (Hesamzadeh Hejazi and Ziaei Nasab, 2010). According to Tukey's test, means with different symbol letters in columns are significantly different (P < 0.01). Means with the same letter are not statistically different (P > 0.05). and chromatin lengths (X, µm) (B) at mitotic metaphase. S1 (P. affinis), S2 (P. eriocarpa), S3 (P. tragium), S4 (P. saxifrage), S5 (P. aurea), S6 (P. tragioides), S7 (P. olivieri), S8 (P. khayyamii), S9 (P. kotschyana), S10 (P. deverroides), S11 (P. olivierioides), S12 (P. anthriscoides), S13 (P. anisactis), S14 (P. peucedanifolia), S15 (P. khorasanica), and S16 (P. rhodantha).
Among species with 22 chromosomes, P. anthriscoides can be clearly distinguished from other species (P. khorasanica, P. peucedanifolia, and P. anisactis) by different evolutionary karyotype classification and karyotype formulas. This species displayed a lower 2C-value in comparison with other species with the same number of chromosomes. Also, P. anthriscoides differs from other species of the genus Pimpinella in some morphological traits, such as plant height, tepal and leaf area, and the size of reproductive organs (data not shown). Based on these observations, P. anthriscoides is introduced as a separate species in the distinct genus Pseudopimpinella in this report. Changes in chromosome morphology and genome size have been developed and used as the basic mechanisms in plant taxonomy and phylogenetic consideration of the genus (Bernardos et al., 2003;Navarro et al., 2004;Arslan et al., 2012).
To the best of our knowledge, there has been no cytological report of the presence of the B-chromosome in Pimpinella species. Hence, this study is the first report in this genus. The B-chromosomes are extra chromosomes and smaller than the usual A-chromosomes, of which the origin and functions are not well known (Palestis et al., 2004;Pellicer et al., 2007). The presence of B-chromosomes has been reported in plant taxa (Fregonezi et al., 2004;Felix et al., 2011;Abedi et al., 2015), and they are not necessarily for the survival of the species; however, they may act in either a positive or negative role as an adaptive function or parasitic genome, respectively (Pellicer et al., 2007;Jones, 2012). The recognition of the B-chromosome in a few individuals may favor the hypothesis of a parasitic B-chromosome (Felix et al., 2011). In the present study, the persistent presence of a B-chromosome in all examined individuals of S10 seems to support the hypothesis of an adaptive function of the B-chromosome.
Khorasan, Iran, Figure 1), which was collected from the same geographic regions with only small differences in longitude and latitude. Previous research and our findings may lead us to conclude that the instability in both chromosome number and ploidy levels in studied Pimpinella species is probably due to interspecific hybridization and polyploidization, which, in turn, induces a cascade of subsequent genomic rearrangements. Above all, epigenomic rearrangements (Maletskii, 2004) may also lead to epigenetic silencing (Scheid et al., 1996). These rearrangements may increase the adaptive capacity of certain species.
According to the chromosomal parameters measured in the current study, the two diploid species with 18 chromosomes, the nine with 20 chromosomes, and the four with 22 chromosomes demonstrated intra-and inter-specific variation in X, TL, and TCV. Considering the primary chromosome parameter, S6which was geographically isolated from other species -appears to exhibit the largest chromatin length (X) among other Pimpinella species, either diploids with different chromosome numbers or tetraploids. This might indicate that chromosomal length is affected by geographical and environmental adaptability. In spite of the observed intra-and inter-specific variation, the bulk of karyotypic symmetrical indices suggests that most Pimpinella species, including diploids and tetraploids, possess symmetric and primitive karyotypes, which is most likely due to inter/intra hybridization and polyploidization. Thus, their similar karyotype structure causes their tendency toward crosses and does not cause a disturbance in reproduction. In general, it is believed that asymmetric karyotype can be linked to the evolutionary history of a particular group of plants (Stebbins, 1971). The high value of the A1 index (variable between 0 and 1) is considered a specialized adaptation, whereas the interchromosomal asymmetry index (A2, variable between 0 and ∞) is associated with the relative taxonomic distance between species of different taxa (Romero-Zarco, 1986). The species P. anisactis (S13) had the smallest A1 and A2 index values, which are probably attributable to its strong adaptability to its habitat conditions; it is therefore not particularly specialized. Three species, S1, S10, and S8E2, had the largest values of the A1 and A2 indices, indicating that these may be well-specialized species.
According to our results, FCM was effectively conducted to analyze ploidy level stability of species (Wyman et al., 1992). Different Iranian Pimpinella species were separated based on their nuclear DNA content, indicating inter/intraspecies diversity and confirming the cytological findings. Variability in DNA C−values is a prerequisite for use as a taxonomic character (Ellul et al., 2002). Previously, there was only a single report of the 2Cvalue in the diploid P. saxifrage (2C DNA = 8.52 pg, Temsch et al., 2010). However, this value differs considerably from our data. The reason for this is unknown but could arise from the cell cycle, rate of cell division, radiation sensitivity, ecological demeanor in plant societies and life forms, and differences between methods of DNA content analysis . A surprising finding related to the 2C value is that the two diploid P. khayyamii accessions (S8E1, S8E2) with two different chromosome numbers of 2n = 2x = 20, 24 and two different 2C DNA content were gathered from the same area (North Khorasan) with a slight difference in geographical coordinates ( Table 1). A major question remains unanswered: what causes P. khayyamii from the same geographic regions to differ in four chromosomes? Are there any genetic or ecological attributes that lead to this difference? Systematic investigation into different aspects of this needs to be undertaken.
The mean comparison of 2C value/chromosome between 16-diploids (0.138 pg) and only tetraploid (0.137 pg) was not substantially different (t-value = 0.00032; P-value = 0.74). For monoploid genome size, a similar conclusion was obviously true (13,811 Mbp for diploids, 1344.75 Mbp for tetraploid; t-value = 1.470; P-value = 0.163). In other words, our finding indicated that the 2C DNA-value mean and 1Cx genome size mean in the examined Pimpinella species were not proportional to ploidy level and the nuclear DNA content per basic chromosome set (1 Cx) tended to decrease when a high ploidy level was observed. A broad analysis of the mean basic genome size at different ploidies in angiosperms showed that, while basic genome size decreased with increasing ploidy, those with larger mean genome sizes at the diploid level showed a greater reduction than those with smaller mean genome sizes (Kellogg and Bennetzen, 2004;Leitch and Bennett, 2004). Studying 67 Artemisia species with different ploidy levels showed that 1Cx genome size tended to decrease significantly in polyploids compared with diploids (Pellicer et al., 2007). Exceptions to this are found in the bromegrass germplasm accessions, where only a slight reduction of DNA content was detected as the ploidy level increased (Tuna et al., 2001). Different patterns of genome size variation linked to different kinds of evolutionary mechanisms and the nature of speciation/polyploidization have been shown in previous studies (Bennetzen and Kellogg, 1997;Soltis et al., 2003). Genome size reduction mechanisms, along with allopolyploidization in Aegilops, could be an obligatory adaptation in polyploid genome evolution (Ozkan et al., 2003). Hence, polyploidy is one possible contributor to C-value variation, but the relationship between C-value and ploidy is not straightforward (Leitch and Bennett, 2004;Murray et al., 2005). In this study, there is no relationship between the 2C DNA content of tetraploids and diploids in Pimpinella. Our karyotypic data on Pimpinella species showed a karyotype formula of 6m + 4sm for diploid S8E1 and 6m + 6sm for S8E2. In the karyotype of S8E1, metacentric "m" chromosomes were predominant, and S8E2 varied in the two types of submetacentric "sm" chromosomes. This may help us to deduce that the newly reported 24 chromosomes diploid Pimpinella tends to have a different evolutionary karyotype classification from the "2A" Stebbins karyotype category (relative symmetric karyotype) for most diploids to the "1B" relatively asymmetric karyotype.
In Pimpinella species, the significant positive correlation between the 2C-value and some karyotypic features, including the ploidy level, the total length of chromatin, and total chromosome volume, indicates that changes in nuclear DNA content have accompanied chromosome structural changes. In agreement with our finding, such a relationship between 2C-value and chromosomal parameters has been reported in Tulipa (Abedi et al., 2015); Lathyrus (Karimzadeh et al., 2011), Thymus (Mahdavi and Karimzadeh, 2010), and Helichrysum (Azizi et al., 2014).
Interestingly, in nine species with 20 chromosomes, the 1Cx genome size of S4 (1Cx = 1281.18 Mbp) was considerably lower (8.07%, P < 0.05) than the other diploid S3 (1393.65 Mbp). The average 1Cx genome size of S3 was more than that of either diploids or tetraploids (S16, 1Cx = 1344.75 Mbp), giving us an expanded view of variation among these species. Such considerable variability could be attractive for either polyploid induction or hybrid production among types of studied Pimpinella species. As a result, genome size might be utilized as an effective marker for detecting hybrids (Ellul et al., 2002).
Further work, such as fluorescence in situ hybridization (FISH), using repetitive sequences and rDNA genes and C-banding complementary assessment of karyology and cytology in meiosis would add more data to Pimpinella taxonomic studies. Our data, in combination with additional data on Pimpinella species worldwide, would help to recognize the origin and evolution of this genus and help to protect the endemic and threatened species with different ploidy levels and chromosome numbers. Moreover, knowledge about genome size is helpful in demonstrating any relationship between nuclear DNA content and the ecological niches of Pimpinella species.

DATA AVAILABILITY STATEMENT
The original contributions presented in the study are included in the article/supplementary material, further inquiries can be directed to the corresponding author/s.

AUTHOR CONTRIBUTIONS
SM, GM, JB, and GR conceived and designed this study. SM and AS-E conducted the experiments. SM and MH analyzed the data. SM wrote the manuscript. HN-Z and DE revised the manuscript. All authors have read and approved the published version of the manuscript.

FUNDING
This research was supported by the University of Western Australia, WA, Australia.