Tilianin content and morphological characterization of colchicine-induced autotetraploids in Agastache mexicana

Schematic overview of the experimental program

A schematic overview of the experimental procedure for in vitro polyploidization of Agastache mexicana is presented in Fig. 1. The protocol is initiated with the treatment of seedlings using colchicine, followed by agitation of the treated explants on an orbital shaker before being inoculated in culture media. Putative polyploid shoots (lines) that emerged from treated explants were subsequently subcultured. To confirm the ploidy levels of the resulting plantlets, stomatal characteristic analysis, chromosome counting, and flow cytometry were employed.

Rooted plants were transferred to ex vitro conditions for acclimatization and greenhouse growth. To assess the impact of polyploidization on tilianin accumulation and morphological traits, a comparative study was conducted using ten tetraploid lines and the diploid control. The evaluation encompassed several parameters during the vegetative phase, including tilianin content, chlorophyll levels, and leaf area. Additionally, various floral characteristics were examined.

Polyploid induction

Seeds from Agastache mexicana subsp. mexicana were collected from a wild population in Felipe Neri, Tlanepantla, Morelos, Mexico, at an altitude of 2,823 m above sea level. The plants were authenticated at the HUMO herbarium of the Biodiversity and Conservation Research Center of Universidad Autónoma del Estado de Morelos, and a voucher specimen (No. 35766) was deposited in the same herbarium. Five hundred seeds were surface sterilized using 98% ethanol for 5 min, rinsed with sterile water, and then with 1.5% of commercial sodium hypochlorite (Cloralex®) for 2 min. Subsequently, the seeds were rinsed five times in sterile water and germinated on autoclaved half-strength basal Murashige and Skoog medium (MS 50%, Murashige & Skoog, 1962), containing 50 mg/L myo-inositol and 3.0 g/L Phytagel®. The cultures were maintained at 25 ± 1 °C under a light regime of 16/8 h (light/dark) photoperiod with a light intensity of 2,000 lux.

To observe the effect of the colchicine concentration and exposition time on the survivor rate and polyploidization of A. mexicana subsp. mexicana, we performed an experimental design where 12 seedlings at the stage of cotyledon emergence were dipped in each of the colchicine concentrations 0.0, 0.1, 0.3, and 0.5% (w/v) at three times: 6, 12 and, 24 h (n = 144). Colchicine was sterilized by filtration through a 0.22 μm Millipore syringe filter. Seedlings were incubated with constant shaking at 100 rpm, in the dark at 25 °C to avoid deterioration of the antimitotic agent under light (Eng, Ho & Ling, 2021). Dimethyl sulfoxide (DMSO) was added at a non-toxic concentration of 2% to help the colchicine penetrate through cell walls (Głowacka, Jeżowski & Kaczmarek, 2009; Salma, Kundu & Mandal, 2017). At the end of each treatment, the seedlings were rinsed eight times with sterile water and then transferred to the same medium, supplemented with 20 g/L sucrose. After 2 months, putative polyploid shoots (8–10 cm length) from surviving seedlings were excised and individually cultured into 250 mL glass jars containing 50 mL of MS 50% medium and subcultures, every 20 days. The jars were sealed with aluminum foil caps with tiny holes covered with a piece of 3 M Micropore™ tape that enables reducing humidity in the jars, while increasing gas exchange to minimize the effect caused by hyperhydricity or vitrification (Zarate-Salazar et al., 2020). The cultures were incubated under the same conditions as described previously. Plantlets obtained by in vitro culture of single shoots from colchicine-treated and untreated seedlings were identified as putative polyploid lines (Amx) and diploid control, respectively. After 8 weeks of culture, the stomatal characteristics, such as length, density, and stomatal index, were assessed. Finally, chromosome duplication of in vitro plants was confirmed by chromosome counts in young root tips and flow cytometry of leaf nuclei.

Acclimatization

After 15 months of successive subcultures and confirmation of the ploidy level, the in vitro plantlets of diploid controls and tetraploid lines were transferred to a commercial substrate (peat moss, leaf soil, coconut fiber, tezontle and agrolite) in six-inch plastic pots and grown under greenhouse conditions (25 ± 2 °C and natural photoperiod conditions) with daily irrigation. At least five plants per line were grown, and subsequently, leaf area, chlorophyll content, flower characteristics, and tilianin content were analyzed.

Stomatal characteristics

Five fully expanded leaves were excised at node position three from the shoots of three diploid/or control plants and from three individuals randomly chosen from five selected putative polyploid lines of the in vitro cultures (n = 90, three plants × five leaves × six lines including the control). The abaxial leaf surfaces were coated with a thin layer of nail polish. After 10 min, the dried polish was removed by applying a strip of transparent one-sided adhesive tape. The dry polish samples, along with the adhered sticky tape were mounted permanently on glass microscope slides, and the stomata length, stomata density (number of stomata per visual field, PVF), and stomata index were recorded using a Leica DM500 optical microscope (Leica Microsystems, Wetzlar, Germany). Images were analyzed using image J software (https://imagej.net). The fields of view were located in the middle portion of leaf lamina, and three fields of vision were investigated for each leaf. The stomatal index was determined from the formula: SI = [S/(S + E)] × 100, where S is the number of stomata in the microscopic field and E is the number of epidermal cells per unit leaf area (Mishra, 1997). Data is presented as the mean of 15 leaf observations within each line or control plants for length, density and stomatal index.

Chromosome count

Mitotic chromosomes were prepared from young root meristems, following the method described by Las Peñas, Bernardello & Kiesling (2008), with minor modifications. In vitro root tips (∼0.5–1 cm) were excised from putative polyploid lines and diploid controls between 7 and 8 o’clock in the morning and pretreated with 0.002 mM 8-hydroxyquinoline (8-HQ) for 24 h at 4 °C in the dark and rinsed with distilled water for 5 min to remove 8-HQ. The root tips were then fixed in Farmer’s solution (ethyl alcohol: acetic acid, 3:1 v/v) for 24 h at room temperature and rinsed with distilled water. Subsequently, samples were hydrolyzed in 1N HCl at 60 °C for 10 min and stained with Schiff’s reagent for 1 h in the dark. Finally, each hydrolyzed root was crushed in a drop of 45% (w/v) acetocarmine and 45% (w/v) acetic acid, and the number of chromosomes in mitotic cells was determined using a light microscope at 100x magnification (DM500®; Leica Microsystems, Wetzlar, Germany). A total of 10 representative photomicrographs were analyzed from three root tips from each line, including the control.

Flow cytometry

Fresh apical leaves from the selected in vitro lines, including the control, were chopped in 1.5 mL Galbraith’s modified buffer (45 mM MgCl₂, 30 mM sodium citrate, 20 mM 4-morpholine propane sulfonate (MOPS) and 0.5% (v/v) TritonX-100, pH 7.0) (Galbraith et al., 1983). After filtration through a 30 μm nylon mesh, crude nuclear samples were stained with 10 mg of propidium iodide. Nuclear DNA content was determined, using the method described by Arumuganathan & Earle (1991), which employs an Attune® Acoustic Focusing Flow Cytometer blue/violet (Applied Biosystems, Foster City, CA, United States of America). Fresh leaves from Solanum lycopersicum (2C = 1.96 pg DNA) were used as an internal standard (Doležel, Greilhuber & Suda, 2007) and more than 5,000 nuclei per sample were analyzed. Three independent replicates were performed for each analysis.

Floral characteristics

To study the effect of polyploidy on some floral characteristics, the size of the flower, as well as the color were assessed in ten tetraploid lines and one control line. Three plants were chosen per each line and within each plant two inflorescences at peak bloom were randomly. Eighteen individual, fully opened flowers from the top, middle, and bottom of the inflorescences were selected for flower length and maximum calyx length, using a digital vernier caliper. Color coordinates were performed using a sprectrophotometer (X-Rite SP64; X-Rite, Grand Rapids, MI, USA) (McGuire, 1992). The X-Rite SP64 was positioned with minimal pressure, perpendicular to the lower lip of each flower and the data were reported in the L* (luminosity 0 = black, 100 = white), C* (chromaticity, saturation level of h) and h (tone angle: 0° = red, 90° = yellow, 180° = green, 270° = blue, 300° magenta) colorimetric system according to Commission Internationale de l’éclairage (CIE) (Schanda, 2007). For accuracy comparison Royal Horticultural Society (RHS) Colour Charts (Royal Horticultural Society, 2015) were used to compare with the X-Rite SP64 sensor readings. Data reported represent the averages for three measurements per flower.

Leaf area and chlorophyll content

Ten tetraploid lines were evaluated under greenhouse conditions to assess the effect of polyploidy on leaf area and chlorophyll content. Leaf area was determined with a leaf area meter (LI-3100C AREA METER; LI-COR® Bio Sciences Instrument, Lincoln, NE, USA) and chlorophyll content (Chl a, b and total) using a portable chlorophyllmeter ClorofiLOG (model CFL 1030; ClorofiLOG, Falker, Brazil). For each line and control (diploid), five plants in vegetative phase were selected at random, and triplicate measurements were performed on two fully developed leaves, taken from the middle section of the shoots (n = 11 lines including the control × five plants × two leaves) (Xavier et al., 2021).

Quantification of tilianin by HPLC

The quantification of tilianin from tetraploid and diploid lines was based on the method established by Hernández-Abreu et al. (2009). For each line in vegetative phase of plant development, 10 g of finely ground dried plant material (leaves and stems) was subjected to continuous maceration (1:10 w/v) with hexane (C₆H₁₄), dichloromethane (CH₂Cl₂), and methanol (CH₃0H) three times for 72 h at room temperature. The tilianin content was determined in the methanolic extract using Waters HPLC equipment with a photodiode array detector, Zorbax C18 SB-CN (4.6 mm × 250 mm, 5-μm particle size; Agilent®), and data analysis was performed using Empower 2002 software. The mobile phase consisted of methanol-water at a ratio of 61:39 (v/v) at a flow rate of 0.7 ml/min and a wavelength of 260 nm. The quantification of tilianin was defined according to the corresponding calibration curve at concentrations of 103.258, 34.264, 27.693, 20.173, 13.027, and 6.442 μg/mL, using highly purified tilianin as reference. All samples were assayed in triplicate.

Statistical analysis

The effect of different concentration of colchicine (0.0, 0.1, 0.3, 0.5%) and exposure time (6, 12, 24 h) on seedling survival rate was analyzed by applying a multiple linear regression (Rossi, 2022). To determine possible phenotypic differences between ten polyploid lines (genotypes) and diploid control, we performed one-way analysis of variance (ANOVA) to define stomata length, stomata density, stomata index, leaf area, chlorophyll a, chlorophyll b, total chlorophyll and tilianin content. Differences were analyzed by applying the Tukey multiple comparisons test (p < 0.05). Principal component analysis (PCA) was carried out to determine potential relationships between genome size and phenotypic traits, previously mentioned. All analyses were performed in R version 4.2.3 (R Core Team, 2023) using ade4 (PCA, Dray & Dufour, 2007) and ggplot2 (Wickham, 2016) libraries, which were used to generate graphs.

Identification of tetraploids and ploidy stability

After 11 days of in vitro culture, 29.8% of the A. mexicana seeds germinated. Seedlings at the stage of cotyledon emergence were selected for the in vitro polyploid induction using colchicine. The first visible effect of colchicine was reflected in the browning and delayed growth rate of treated seedlings. Significant differences between treatments were observed after 2 weeks of culture, and multiple linear regression analysis (y = 1.278-colchicine (2.279)-time exposure (0.005)) revealed a strong negative correlation between seedling survival and colchicine dosage (p = 0.003), but no significant correlation was found with exposure time. The second observed effect of colchicine was the inhibition of two processes: (1) shoot formation from axillary buds in seedlings, and (2) in vitro rooting of excised shoots to produce plantlets. The number of shoots per seedling decreased with increasing colchicine dose, resulting in a shoot count of 0.40 ± 0.329 for colchicine-treated seedlings and 1.79 ± 0.318 for untreated (diploid control) seedlings after 4 weeks in culture. Overall, higher colchicine doses corresponded to lower survival rates and fewer plantlets produced (putative polyploid lines) (Table 1).

In this study, 29 putative polyploid lines of A. mexicana were obtained from colchicine treatments. Generally, after 8 weeks of in vitro culture, the putative polyploid plantlets showed markedly different morphological characteristics from the diploid controls, manifesting smaller or larger leaves, often with a rolled structure, and ranging in color from purple to dark green color. Significant differences were observed between the stomatal characteristics of the in vitro diploid control and tetraploid lines. As shown in Fig. 2, the leaves from tetraploid lines exhibited large stomata, with significantly higher stomatal lengths than leaves from diploid control (F_5,24 = 4.613, p < 0.01; 4x = 37.94 ± 7.81 μm; 2x = 31.46 ± 4.00 μm), as well as, a lower stomatal density (F_5,24 = 21.82, p < 0.001; 4x = 117.5 ± 60.26 PVF; 2x = 161.24 ± 19.43 PVF) and stomatal index (F_5,24 = 18.47, p < 0.001; 4x = 10.25 ± 2.22%; 2x = 16.23 ± 1.85%).

Studies of mitotic cells in actively growing root tips clearly indicated that the polyploidy of the samples was due to chromosome doubling in diploid seedlings induced by colchicine treatments. Figure 3 shows that the number of chromosomes in the polyploid plantlets is 2n = 4x = 36, whereas the number of chromosomes in the diploid plantlets (controls) is 2n = 2x = 18. Of the 29 lines evaluated; 23 (79.3%) were tetraploids, 5 (17.3%) diploids, and 1 (3.4%) mixoploid, containing a mixture of diploid and tetraploid cells. In order to confirm the ploidy level of plantlets using flow cytometry, as there have been no reports to date that estimate genome size of A. mexicana, it was necessary to determine the genomic content, by comparing the mean position of the G0/G1 peak of the internal standard of S. lycopersicum with the mean position of the peak of the diploid sample of A. mexicana. Eight independent measurements were performed to conclude that diploid control plants (2n = 18) have an average genomic content of 1.433 ± 0.025 pg. Flow cytometry DNA histograms of the polyploid lines revealed that the peak position was twice that of the genome DNA of diploid (Fig. 4). It was thus concluded that the DNA content of the tetraploid cells (2.660 ± 0.236 pg) equals twice that diploid cells. The three groups of ploidy levels: diploid, tetraploid, and mixoploid plantlets based on flow cytometry, fully concurred with the results obtained from counting chromosomes. The highest percentage of tetraploid production efficiency (91.66%) was achieved with 0.1% colchicine for 6 h, without generating any mixoploids (Table 1). Chromosome counting and flow cytometry analysis indicated that DNA ploidy levels remain stable in all tetraploid lines after 15 months of in vitro culture.

During in vitro culture, the mixoploid line and several tetraploid lines showed symptoms of hyperhydricity, low growth capacity, and reduced root development (Fig. 5). Consequently, during the acclimatization process, the mixoploid plants did not survive the transfer from in vitro conditions to greenhouse conditions. In contrast, ten tetraploid lines exhibited accelerated growth and development in comparison to the diploid control. These lines were selected to evaluate the effect of ploidy on leaf and flower traits, as well as and on tilianin content in A. mexicana.

Flower and leaf characteristics of A. mexicana grown under greenhouse condition

Inflorescence emergence began earlier in the tetraploid plants; after only 3 months in the greenhouse, compared to 4 months for diploid control. All tetraploid lines, except for line Amx2, exhibited significantly increased flower and calyx lengths compared to the diploid control (Table 2; Fig. 6). In tetraploid lines, the mean lengths of flowers and calyxes were 30.659 ± 2.401 mm and 11.732 ± 0.801 mm, respectively, so significantly greater than those in diploid lines (25.252 ± 2.947 mm and 11.271 ± 2.334 mm, respectively; p < 0.0001). Also, there were significant differences in the color components of flowers of tetraploids and diploid control (Table 2). However, in both tetraploid lines and diploids, the value of L* tends to be neutral, the color tends to be magenta and slightly more opaque (C* = less than 30) in the tetraploid lines, compared to the diploid control, which tends to be violet. This resembles the visual evaluation provided by the RHS, which identified the color of the diploid flowers as deep strong reddish purple C (NN78) and the tetraploid flowers as light reddish purple D (NN78).

Differences in leaf area were observed between the diploid and the tetraploid lines Amx3, Amx7, and Amx10, with the tetraploids manifesting a higher chlorophyll content and smaller leaf area (Fig. 2, Table 3). Tetraploid plants presented a dark green leaf color.

Tilianin content

Statistically significant differences in tilianin content were found between the tetraploid lines and diploid control in methanolic extract (F_{10, 22} = 23,806, p < 0.0001). The diploid control produced an average of 6.388 ± 0.005 mg of tilianin per gram of dry weight (mg/g DW). Out of the tetraploid lines, the greatest tilianin accumulation was recorded in Amx3 (32.964 ± 0.004 mg/g DW) and Amx7 (32.392 ± 0.110 mg/g DW). Compared to the control (as shown in Fig. 7), Amx8 (3.195 ± 0.005 mg/g DW) exhibited less accumulation of tilianin.

Principal component analysis

Principal component analysis (PCA) was performed to investigate how the patterns of variance of morphological traits and tilianin content correlated with ploidy level. This was also used to determine their possible contribution to the content of tilianin and to identify particular traits that distinguish diploid from tetraploid plants. Two principal components, PC1 and PC2, were determined, based on their degree of contribution. The first two components accounted for 57.05% of total variance, with 33.88% relating to PC1 and 23.17% relating to PC2 (Fig. 6). PC1 correlated positively with C* (0.711) and negatively with calyx length (−0.805), DNA (−0.796), flower length (−0.760) and h (−0.498). PC2 showed a positive correlation with tilianin (0.714), and negative correlation with L (−0.748) and leaf area (−0.589). The control group is separated from the tetraploid lines in Fig. 8B.