Introduction

Polyploidy, the condition in which an organism possesses more than two complete sets of chromosomes, is a compelling phenomenon in plant biology (Soltis et al. 2009). It endows plants with unique morphological, physiological, and genetic characteristics, often enhancing their adaptability and resilience in different ecological niches (Soltis et al. 2009; Wood et al. 2009; Sattler et al. 2016). Morphological characteristics of polyploid individuals were assessed, including parameters such as plant stature, leaf dimensions, floral attributes, and pollen grain size (Urwin et al. 2007; Niu et al. 2016). The incorporation of integrative methodologies that combine molecular analyses with ecophysiological assessments, to contrast the performance of polyploids under controlled laboratory conditions with that of polyploids in their natural habitats holds promise for addressing outstanding questions. Such approaches could provide insights into the ecological and evolutionary significance of polyploidy, thereby advancing our understanding of its ecological and evolutionary implications (Madlung 2013). The induction of artificial polyploidy, an integral aspect of plant breeding and genetic improvement, has widely been explored to harness these advantageous traits (Sattler et al. 2016; Trojak-Goluch et al. 2021). While the in vitro induction of polyploidy allows for the efficient production of pure polyploids, it introduces the possibility of producing mixoploids, characterised as chimeras comprising both diploid and polyploid cells (Chen and Gao 2007). Polyploidisation methods are experiencing a resurgence and are increasingly becoming an integral part of breeding initiatives for several valuable crop species, as well as aromatic and medicinal plants (Dhawan and Lavania 1996; Xing et al. 2011; Bagheri and Mansouri 2015; Švécarová et al. 2018, 2019; Fernandes et al. 2023). In the context of aromatic and medicinal plants, polyploidisation can significantly affect the levels of a wide range of biologically active compounds (Tavan et al. 2015). In particular, such modifications have the potential to facilitate the use of plant extracts in the pharmaceutical, cosmetic, and food processing industries, as well as for botanical pesticide applications (Navrátilová et al. 2022). The positive effect of artificial polyploidisation on the secondary metabolite content of medicinal plant species is summarised in Table 1.

Table 1 Several studies of artificial polyploidisation in medicinal plants with a given group of induced secondary metabolites
Full size table

As a consequence of polyploidy, many polyploid plants show enhanced traits over their diploid progenitors. The best-known phenotypic effect is the ‘gigas’ effect, which is characterised by an increase in cell size resulting in larger organ size due to increased nuclear DNA volume. However, this is not a rule and in cultivated crops, for example, it does not always mean increased biomass production and improved yield (cf. Sattler et al. 2016). Polyploid plants often have lower growth rates, reduced fertility or viable seed production (Stebbins 1971). On the other hand, polyploidisation can have a positive effect on tolerance to environmental and/or biological stresses (Levin 1983, 2002). The complexity of distinguishing ploidy-driven changes from those induced by other factors affecting artificial polyploids adds to the challenge. Processes like chemical induction for genome duplication, protoplast fusion, in vitro regeneration, or genomic shock contribute to uncertainties about the effects of ploidy and the genomic identity between polyploids and their diploid progenitors (Comai 2005; Münzbergová 2017). Further studies focusing on the specific applications and implications of the ‘gigas’ effect in different agricultural contexts are warranted (Sattler et al. 2016; Salma et al. 2017; Eng and Ho 2019).

In fact, the relationship between polyploidy, cell size, and total plant biomass is more complex. While larger cell size is a well-known effect of polyploidisation, it does not always translate into increased adult size or final biomass in plants (Iannicelli et al. 2020). The regulation of cell size and their number serves as an ecological adaptation strategy, and the balance between these factors can impact the growth rate of polyploid cells. Smaller surface area to volume ratios in larger cells can reduce the growth rate, resulting in cases where body size remains unchanged with polyploidisation (Lavania et al. 2012). It is crucial to note that the relationship between increased genetic dose in polyploids and the observed gene expression and enzyme activity is not linear (Iannicelli et al. 2020). The complex interplay of these factors highlights the need for a nuanced understanding of the effects of polyploidisation on plant physiology and secondary metabolite production.

This study explored the interesting field of artificial polyploidisation in the Astragalus membranaceus (Fabaceae), a botanical resource of great importance due to its valuable phytochemical components (Fu et al. 2014; Durazzo et al. 2021). A. membranaceus, commonly known as ‘Huang Qi’ or ‘Milk-vetch root’, is a perennial, diploid herb (the somatic metaphase chromosome number is 2n = 2x = 16; Chen and Gao 2007) traditionally known for its pharmacological properties (D’Avino et al. 2023). The roots (Radix Astragali) are rich in bioactive compounds (esp. triterpene glycosides, flavonoids, saponins and alkaloids; Zhang et al. 2021), and are recognised for their immunomodulatory (Qader et al. 2021), antioxidant (Cui et al. 2022), antibacterial (Samuel et al. 2021), hepatoprotective (Chen et al. 2023b) diuretic (Jia et al. 2021), antidiabetic (Chen et al. 2022), anticancer (Liu et al. 2023), and anti-inflammatory properties (Chen et al. 2023a), making it a staple in traditional medicine (Md Ashif Ikbal et al. 2022). Nevertheless, there is an inherent need to increase productivity and phytochemical content to meet the increasing demands of the pharmaceutical and nutraceutical industries (Wu et al. 2003).

In this study, we use oryzalin, a well-established microtubule-depolymerising agent (Hansen et al. 1998), to induce polyploidy in A. membranaceus under controlled in vitro conditions. Oryzalin’s proven efficacy in promoting polyploidisation in a variety of plant species is the reason for this choice (Tosca et al. 1995; Van Duren et al. 1996; Švécarová et al. 2018, 2019). We employed real-time reverse transcription quantitative polymerase chain reaction (RT-qPCR) to investigate the gene expression profiles associated with flavonoid biosynthesis. To our knowledge, this is the first study to examine gene expression in artificial polyploid plants of A. membranaceus, providing novel insights into how polyploidisation affects the transcriptional regulation of key biosynthetic genes. Specifically, we examined the transcript levels of essential genes involved in the biosynthetic pathways of calycosin and calycosin-7-O-β-D-glucoside, namely AmPAL, AmC4H, and AmI3′H (Kim et al. 2014). The main aim of this study is to elucidate the genetic consequences of artificial polyploidy induction in A. membranaceus, with a particular focus on the potential enhancement of the phytochemical constituents of this botanical resource for pharmaceutical and industrial applications. Although we did not measure root metabolites directly, gene expression analysis is an important first step in understanding the biosynthetic potential of polyploid lines. The expression levels of AmPAL, AmC4H, and AmI3’H serve as indicators of the metabolic capacity for flavonoid production, suggesting that increased transcript levels may correlate with enhanced metabolite accumulation once plants reach maturity (Kim et al. 2014). This choice is driven by practical considerations, as meaningful genotype to control comparisons require 6–8 years of growth, the conventional harvest time for Radix Astragali (Zhang et al. 2021). This extended growth period ensures the accumulation of secondary metabolites at levels suitable for comparison analysis in medicinal plant research. Importantly, our gene expression results will serve as selection markers to identify promising genotypes and guide future biochemical analyses of root secondary metabolites in A. membranaceus. The knowledge gained from this investigation extends the frontiers of plant polyploidy research and holds considerable promise for the pharmacological and industrial sectors.

Materials and methods

Plant material

Mature seeds of A. membranaceus were sourced from the collection of Roman Pavela at the Crop Research Institute (CRI), Prague-Ruzyně, Czech Republic. The seeds have undergone a surface sterilisation process, starting with a 2-minute treatment in 70% ethanol, followed by a 20-minute disinfection using a 5% chloramine T solution enriched with Tween-20. Subsequently, the seeds were meticulously rinsed thrice in sterile distilled water. Germination took place in Petri dishes (6 cm in diameter, each containing 10 seeds), on Murashige and Skoog (MS) (Murashige and Skoog 1962) medium including vitamins, 30 g·L− 1 of sucrose, and 8 g·L− 1 of agar (both from Duchefa, the Netherlands). The medium was further enriched with 0.01 mg·L− 1 of indole-3-butyric acid (IBA) and 0.01 mg·L− 1 of 6-benzyladenine (BA) (both from Duchefa, the Netherlands), and 0.1% Plant Preservative Mixture (PPM) (Plant Cell Technology, Inc.). The pH of the MS medium was adjusted to a value of 5.8.

The seeded Petri dishes were placed in darkness for three days at a constant temperature of 25 °C, regulated by a thermostat, before being transferred to a controlled culture room. This chamber was maintained at 22 ± 2 °C, operating under a photoperiod of 16 h of light and 8 h of darkness, with a light intensity ranging between 32 and 36 µmol·m− 2·s− 1. The germination rate of the seeds was observed to be very low, not exceeding 10%. Subsequent to germination, the resulting plantlets, deemed free of any internal contamination, were subjected to multiplication. The shoots were recurrently trimmed every six weeks into segments measuring 20–30 mm in length and subcultured. This method ensured the emergence of a clone from each seed within Erlenmeyer flasks (100 mL flask volume containing 30 mL of medium and 5–6 shoots). The medium was consistent with the previous formulation, supplemented with 0.1 mg·L− 1 of IBA, 0.1 mg·L− 1 of BA, and 20 mg·L− 1 of ascorbic acid. The side branches were isolated and placed on the medium during the culturing process, facilitating the propagation of a sufficient number of plants from a single seed (approximately 50–60 plants). Prior to initiating the in vitro polyploidisation experiment, the plants were cultivated on MS medium in Erlenmeyer flasks containing 0.01 mg·L− 1 of IBA, 0.01 mg·L− 1 of BA, and 20 mg·L− 1 of ascorbic acid. Out of six genotypes (KP 1–6, each originating from 1 seed), only genotype KP2 was selected for the polyploidisation experiment. The clones of this genotype exhibited the best growth and demonstrated successful micropropagation.

In vitro polyploidisation

Nodal segments from KP2 clones were meticulously chosen as candidates for the in vitro polyploidisation experiment. Explants were derived from the third to fifth nodal segments below the apex of donor plants, each featuring one pair of leaves devoid of visible axillary buds. The oryzalin stock solution (obtained from Sigma-Aldrich, St. Louis, MO, USA) was meticulously prepared according to the protocol outlined by Greplová et al. (2009), with specific oryzalin concentrations subsequently derived from this stock solution. This stock solution was subsequently used to prepare working solutions of various concentrations, which were then employed to immerse nodal segments in the polyploidisation experiment.

Immersion of nodal segments in oryzalin solution for 24 h (ONS)

The immersion of nodal segments in the oryzalin solution was carried out in Erlenmeyer flasks with a meticulous setup to ensure experimental accuracy and consistency. Specifically, each flask contained 7–10 nodal segments, with the exact number depending on the size of the segments. This ensured uniformity of treatment dosage and minimized variability. The immersion experiment was repeated three times to ensure reliability.

Nodal segments (147 clones) were seated on the hormone-free MS medium within Erlenmeyer flasks. Before immersion, the segments were cultivated in a thermostat (for anchoring) for 48 h and entirely immersed in the oryzalin solution for 24 h. The control group (labelled as ‘C’; 20 clones) was submerged in sterile distilled water. Oryzalin solution concentrations varied; clones were exposed to concentrations of 6.92 mg·L− 1 (82 clones) and 13.85 mg·L− 1 (45 clones). The segments were incubated within a thermostat at 25 °C. Following treatment, the oryzalin solution was meticulously removed, and the nodal segments underwent a thorough triple rinse in sterile distilled water before being transferred to MS medium supplemented with 0.01 mg·L− 1 of IBA, 0.01 mg·L− 1 of BA, and 20 mg·L− 1 of ascorbic acid. This transfer occurred within Erlenmeyer flasks, and the culture was maintained for a period of two weeks. Subsequent subculturing activities took place every four weeks within tubes Shoots that emerged from individual axillary buds in response to oryzalin treatment, indicating the ability to regenerate and the possibility of successful polyploidisation, were systematically enumerated and categorised as subclones (resulting in 72 surviving subclones). Throughout the experiment, all procedures were performed under identical environmental conditions to minimize external influences on the results.

Multiplication of polyploid plants, rooting and acclimatisation

The proliferating shoots derived from the oryzalin experiments underwent flow cytometry assessment, and only polyploid and diploid reference plants were selected for further cultivation. Diploid regenerants were deliberately excluded from subsequent analysis. Polyploid plants were propagated on MS medium supplemented with 0.01 mg·L− 1 of IBA, 0.01 mg·L− 1 of BA, and 20 mg·L− 1 of ascorbic acid within culture tubes. These polyploid shoots underwent regular subculturing at intervals of five weeks. Extended shoots were meticulously divided into segments measuring 4–5 cm in length, and any discoloured or blackened leaves were removed. The resulting shoots subsequently underwent the rooting process, which was attempted using various methods. The most successful approach involved using the liquid half-strength MS medium supplemented with 2 mg·L− 1 of 1-Naphthaleneacetic acid (NAA). Subclones in this medium developed young adventitious roots within 2–6 weeks. Once roots were established, the plants were transferred to plastic domes containing perlite for acclimatisation and further rooting. After approximately 5 weeks, the rooted plants were transplanted into pots with substrate and placed outside the domes. Subsequently, these plants were cultivated in a greenhouse for 2 years.

Flow cytometry

The DNA ploidy level (relative genome size; (Suda et al. 2006) of both in vitro and ex vitro plant material was estimated using flow cytometry. Leaf tissue was used as plant material for flow cytometry analysis in both in vitro and ex vitro conditions. This choice was made due to the presence of leaves in both cultivation environments and their suitability for non-destructive sampling. Samples were prepared according the standard protocol using LB01 buffer (Doležel et al. 2007) and were run on the Partec CyFlow ML flow cytometer (Partec GmbH, Münster, German) using fluorochrome DAPI (4′,6-diamidino-2-phenylindole, 5 µg·mL–1). Pisum sativum ‘Ctirad’ (2 C = 9.09 pg; (Doležel et al. 1998) served as an internal standard. For each sample, fluorescence intensity of 3000 particles was recorded, and only histograms with the coefficient of variation (CV) less than 5% were accepted.

The RNA isolation and reverse transcription to cDNA

Total RNA extraction was carried out using the Spectrum Plant Total RNA Kit (Sigma-Aldrich, St. Louis, MO, USA). The harvested biomass (in vitro shoots and leaves of 2-year-old ex vitro plants) was immediately frozen with liquid nitrogen and homogenised using a chilled mortar and pestle with liquid nitrogen on dry ice. During the isolation process, an On-Column DNase I Digestion Set (Sigma-Aldrich, St. Louis, MO, USA) was applied to prevent DNA contamination.

The concentration and purity of the isolated RNA were assessed by measuring the A260/A280 and A260/A230 ratios using a NanoDrop 2000 spectrophotometer (ThermoScientific, Waltham, MA, USA). The RNA quality was verified, and the absence of genomic DNA was confirmed through agarose gel electrophoresis. Subsequently, complementary cDNA was synthesised utilising the Sensi FAST cDNA synthesis kit (Meridian Bioscience Inc., Cincinnati, OH, USA) in an Eppendorf Mastercycler Pro S vapo.protect thermocycler (Eppendorf, Hamburg, Germany).

Real-time reverse transcription quantitative PCR (RT-qPCR) analysis

Primers for all the genes of interest (AmPAL, AmC4H, and AmI3′H) and the housekeeping (HK) gene encoding 18 S ribosomal RNA (Am18S) were taken from the article by Kim et al. (2014).

For detailed information on all the primers used, see Table 2.

Table 2 List of primers for all studied genes. Forward (Fw) and reverse (Rv) primer sequences, amplicon length expressed in base pairs (bp), and PCR efficiency (Eff)
Full size table

The primer sets were initially validated through standard PCR using genomic DNA extracted from both the control and experimental variants. This validation was carried out in a thermocycler Eppendorf Mastercycler Pro S vapo.protect (Eppendorf, Hamburg, Germany), and the resulting PCR products were assessed using agarose gel electrophoresis.

For RT-qPCR, SensiFAST SYBR No-ROX Kit (Meridian Bioscience Inc., Cincinnati, OH, USA) was employed in a 96-well thermocycler, CFX Connect Real-Time PCR Detection System (Bio-Rad Laboratories, Hercules, CA, USA). The expression levels at various sampling cycles were normalised to the 18 S ribosomal RNA (Am18S) gene, serving as a reference HK gene. Before conducting the RT-qPCR analyses for Relative Gene Expression (RGE) profiles of the genes of interest and HK gene, primer pairs underwent an annealing temperature optimisation step to determine PCR reaction efficiency within our system. These efficiencies were subsequently utilised in the calculation of RGE profiles using the Pfaffl method (Pfaffl 2001). The PCR reactions followed these conditions: polymerase activation at 95 °C for 2 min, succeeded by 40 amplification cycles consisting of 95 °C for 5 s, 55 °C for 10 s, and 72 °C for 20 s. A melting curve analysis was then performed, spanning from 65 °C to 95 °C, to confirm the specificity of the PCR products. A non-template (cDNA-free) control was included to detect possible contamination of the reagents. Three independent biological replications were performed for each of the three genes, and each sample was prepared in triplicate.

Statistical analysis

The acquired outcomes of RT-qPCR method underwent statistical analysis utilising the freely available online software ASTATSA. Initial scrutiny involved an analysis of variance, followed by a comparison of means using Tukey’s Honestly Significant Difference (HSD) test within the framework of a one-way ANOVA (Vasavada 2016). Statistically significant distinctions between the control group (represented as ‘C’) and the tetraploid lines are indicated in the figures by asterisks (p ≤ 0.01 ‘**’; p ≤ 0.05 ‘*’).

Results

In vitro polyploidisation and flow cytometric analysis

For clone KP2, 147 nodal segments were subjected to experimental cultivation with oryzalin immersion (ONS). Of these, 63 explants survived this process, of which nine were identified as tetraploids by flow cytometry analysis. Of the 82 clones immersed in the oryzalin solution at a concentration of 6.92 mg·L− 1, 52 subclones were successfully obtained (survival rate of 63.4%), with six confirmed as tetraploids (successful rate of 11.5%, although one later died). When 45 clones were subjected to oryzalin immersion at a concentration of 13.85 mg·L− 1, 11 subclones were obtained (survival rate of 24.4%), and three of them were determined to be tetraploids (successful rate of 27.3%). The tetraploid subclones were denoted by numerical order based on their flow cytometric measurements (i.e. 16, 36, 50, 51, 54, 58 (deceased), 65, 74, and 76), for further analyses.

It was observed that the induction of polyploidy in cultivated nodal segments on hormone-free MS medium at various concentrations resulted in lethality, with affected plants displaying necrotic symptoms.

Verification of the stability of polyploidy after two years under ex vitro conditions

After two years of cultivation of the surviving plants under ex vitro conditions (C, 16, 36, 51, and 74), subsequent flow cytometric analyses were carried out to verify the ploidy level of the newly induced polyploids. Unfortunately, only line 74 was confirmed to be tetraploid. The rest lines were diploids.retained its tetraploid status. The rest of analysed biological material of remaining lines was characterized as diploidic (original ploidy — 2n). However, gene expression analyses using RT-qPCR revealed significant differences between the different lines, even in cases where they were verified to be diploid.

Real-time reverse transcription quantitative PCR (RT-qPCR) analysis

To investigate alterations in the expression of crucial genes between oryzalin-induced neo-tetraploid plants and diploid control plants, we extracted total RNA from ‘KP2’ leaves grown under identical conditions. These samples were collected from plants cultivated under in vitro conditions and matured two-year-old plants in ex vitro conditions within a greenhouse. RT-qPCR analysis was conducted on specific genes of interest (AmPAL, AmC4H, and AmI3′H) associated with the biosynthesis of calycosin and calycosin-7-O-β-D-glucoside, along with the HK gene responsible for 18 S ribosomal RNA (Am18S) (Table 2). The results unveiled notable variations in the expression levels of all three genes of interest between neo-tetraploids and diploid control plants.

RT-qPCR of in vitro plants

The relative expression profiles of the three genes of interest studied in plants grown under in vitro conditions showed statistically significant changes in three experimental tetraploid lines (54, 65, and 74) compared to the control variant, consistently for at least one of the genes examined (Fig. 1). The most pronounced alterations were observed in genotype 54, where all genes examined were statistically significantly upregulated (AmPAL 2.6-fold with p ≤ 0.01; AmC4H 2.5-fold with p ≤ 0.05; AmI3’H 3.4-fold with p ≤ 0.01). Subsequently, genotype 74 exhibited statistically significant upregulation of the AmI3’H gene (2.8-fold with p ≤ 0.01). This upregulation could be attributed to the doubled gene dosage resulting from polyploidisation. However, downregulation of the AmPAL gene was observed in genotypes 51, 65 and 76 (line 65 0.2-fold with p ≤ 0.05).

Fig. 1
figure 1

Quantification of relative expression of AmPAL, AmC4H, and AmI3′H genes in A. membranaceus plants grown under in vitro conditions. Expression data were normalised using Am18S as the housekeeping gene, with the relative control genotype serving as the calibration reference. Significant differences in gene expression are represented by asterisks (p ≤ 0.01 ‘**’; p ≤ 0.05 ‘*’). Note C = diploid ‘Control’ variant

Full size image

RT-qPCR of ex vitro plants

The relative gene expression analysis in matured two-year-old plants cultivated in ex vitro conditions exhibited distinct results (Fig. 2). Genotype 16, although no longer tetraploid, showed the highest relative expression levels among the genotypes examined, with statistically significant upregulation of the AmPAL and AmC4H genes (2.8-fold and 2.1-fold, respectively, both with p ≤ 0.01). Subsequently, the induced autotetraploid genotype 74 displayed a notable and statistically significant upregulation of the AmI3′H gene (2-fold with p ≤ 0.01) and statistical downregulation of gene AmC4H (0.5-fold with p ≤ 0.01). Conversely, the remaining genes in these two genotypes were downregulated, and genotypes 36 and 51 exhibited downregulation across all three examined genes (approximately 0.5-fold for all genes without AmPAL in line 36 with p ≤ 0.01). As mentioned above, only line 74 was confirmed to be tetraploid, the rest lines were diploids. However, the RT-qPCR gene expression analyses revealed substantial variations among the lines, suggesting persistent effects even in lines where tetraploid ploidy level was not detected.

Fig. 2
figure 2

Quantification of relative expression of AmPAL, AmC4H, and AmI3′H genes in matured two-year-old A. membranaceus plants under ex vitro conditions within a greenhouse. Expression data were normalised using Am18S as the housekeeping gene, with the relative control genotype serving as the calibration reference. Significant differences in gene expression are represented by asterisks (p ≤ 0.01 ‘**’; p ≤ 0.05 ‘*’). Note C = diploid ‘Control’ variant

Full size image

Discussion

Polyploidisation is a well-established approach in plant breeding to induce genetic and epigenetic changes. The success of polyploidisation induction depends on various factors such as explant type, genotype, culture medium, antimitotic agents, exposure time, and concentrations used (Dhooghe et al. 2011; Švécarová et al. 2019). Common antimitotic agents include colchicine, oryzalin, and trifluralin (Trojak-Goluch et al. 2021).

Gene dosage manipulation with the whole-genome duplication can alter gene expression and enzyme activity, aiming to achieve desired morphological and phytochemical traits (Niazian and Nalousi 2020). Both in vivo and in vitro polyploidisation can induce changes in morphological and anatomical characteristics, leading to increased production of secondary metabolites (Majdi et al. 2010; Tavan et al. 2015; Pavela 2015). Artificial in vitro polyploidisation has profound effects on gene expression, including amplification, silencing, and regulation of duplicated genes (Niazian and Nalousi 2020).

Polyploidy-induced changes in gene expression involve mechanisms such as sequence elimination, interchromosomal exchange, cytosine methylation, gene repression, novel activation, genetic dominance, subfunctionalisation, and transposon activation (Chen and Ni 2006). These modifications can significantly affect the chemical profile of medicinal plants (Levy 1976; Lavania 2005; Adams and Wendel 2005a; Hannweg et al. 2016). Notably, in Papaver bracteatum, induced tetraploid plants showed increased transcript levels of SAT and TYDCgenes in the benzylisoquinoline biosynthetic pathway (Madani et al. 2019), supporting our use of RT-qPCR to assess gene expression changes in A. membranaceus.

In the context of medicinal plants, the overexpression of genes involved in biosynthetic pathways represents an effective strategy to enhance the production of desired pharmaceutical molecules through metabolic engineering. Whole genome duplication is proving to be a powerful approach, as it affects the entire metabolic pathway, surpassing the manipulation of individual genes (Fuentes et al. 2016; Niazian 2019). However, the consequences of polyploidy induction are diverse, and the desired traits or phenotypes may not always manifest due to gene silencing, genome reorganisation, and dominance effects. Rapid epigenetic changes, chromosome rearrangements, alterations in regulatory networks, and the loss of duplicated genes can all contribute to unexpected outcomes of polyploidy (Osborn et al. 2003; Adams and Wendel 2005a, b; Chen and Ni 2006; Iannicelli et al. 2020).

The careful consideration of these variables is essential for effective induction of polyploidisation in A. membranaceus. The generation of induced autoetraploids in the genus Astragalus has received attention in various studies (Wu et al. 2003; Chen and Gao 2007; Wang et al. 2009; Zhang et al. 2011). However, to our knowledge, none of these studies have focused on examining gene expression changes after polyploidisation. Our choice of oryzalin, a microtubule-depolymerising agent, to induce polyploidy is in line with its proven efficacy in various plant species (Tosca et al. 1995; Švécarová et al. 2018, 2019; Parsons et al. 2019; Navrátilová et al. 2021). The induction of polyploidy in A. membranaceus by oryzalin treatment showed varying degrees of success. The survival rates and successful induction rates at different concentrations of oryzalin highlight the sensitivity of A. membranaceus to polyploidisation induction. The lethality observed, particularly at higher concentrations, is consistent with previous findings in plant polyploidisation studies (Parsons et al. 2019; Navrátilová et al. 2021).

However, to our knowledge, none of these studies have focused on examining gene expression changes after polyploidisation. Our choice of oryzalin, a microtubule-depolymerising agent, to induce polyploidy is in line with its proven efficacy in various plant species (Tosca et al. 1995; Švécarová et al. 2018, 2019; Parsons et al. 2019; Navrátilová et al. 2021). The induction of polyploidy in A. membranaceus by oryzalin treatment showed varying degrees of success. The survival rates and successful induction rates at different concentrations of oryzalin highlight the sensitivity of A. membranaceus to polyploidisation induction. The lethality observed, particularly at higher concentrations, is consistent with previous findings in plant polyploidisation studies (Parsons et al. 2019; Navrátilová et al. 2021).

In the course of this study, our strategic decision to use gene expression analysis rather than direct measurement of root secondary metabolites reflects a methodological choice aligned with our research objectives. This decision was based on practical considerations, as the conventional harvest time for Radix Astragali is six to eight years (Zhang et al. 2021). This longer growth period is essential to ensure the accumulation of secondary metabolites at levels suitable for rigorous analysis in the context of medicinal plant research. It is crucial to note that, specifically for the targeted biochemistry of flavonoids, a minimum cultivation period of three years is required for meaningful biochemical analyses (Ma et al. 2002). However, the ideal cultivation period of six to eight years corresponds to the traditional harvest times of Radix Astragali in China and ensures robust levels of secondary metabolites (Zhang et al. 2021). In particular, the gene expression data generated by our approach are expected to serve as valuable selection markers. These markers will play a pivotal role in identifying and characterising promising genotypes, thereby guiding and informing future biochemical analyses of root secondary metabolites in A. membranaceus. This integrative approach will deepen our understanding of the genetic consequences of artificial polyploidy induction and its potential impact on the phytochemical constituents of this botanical resource, providing insights relevant to both pharmaceutical and industrial applications.

While successful polyploid induction was initially confirmed in several lines, polyploid stability was not uniform over time. The stabilisation of polyploidy in line 74, despite chimeric mosaicism, highlights the complexity of maintaining polyploid states in A. membranaceus. This variability in polyploid stability is consistent with observations in other plant species (McClintock 1984; Feher et al. 1989; Wang et al. 2010; Zhang et al. 2013; Pereira et al. 2017; Lv et al. 2022). Gene expression analyses by RT-qPCR provided insights into the molecular consequences of polyploidisation. Significant changes in the expression levels of key biosynthetic genes (i.e. AmPAL, AmC4H, and AmI3′H) associated with the production of calycosin and calycosin-7-O-β-D-glucoside were observed. The observed up- and downregulation of these genes in different lines suggests a nuanced response to polyploidisation, reflecting the intricate regulatory mechanisms that control gene expression in response to changes in ploidy (Levy 1976; Albuzio et al. 1978; McClintock 1984; Mishra et al. 2010; Yang et al. 2011; Bharadwaj 2015; Ghimire et al. 2016; Zhang et al. 2019). Recent advances in gene expression analysis using RT-qPCR techniques have significantly improved the understanding of how polyploidisation affects metabolite production. Studies focusing on genes responsible for the synthesis of metabolites such as morphine, artemisinin, vindoline, catharanthine, and vinblastine have shown positive regulation and over-expression of many of these genes. These findings indicate that the polyploid state leads to differential modulation of gene expression within biosynthetic pathways (Mishra et al. 2010; Xing et al. 2011; Lin et al. 2011). The observed influence of polyploidy on gene expression can be attributed to the activation of intricate cellular mechanisms that modulate DNA template abundance, thereby influencing translation and transcription processes. Consequently, this dynamic interplay can lead to diverse outcomes, including increased, decreased, or even silenced gene expression (Madani et al. 2021). Our study is consistent with these observations, where several experimental lines exhibited a marked decrease in relative gene expression (genotypes 36 and 51 of matured two-year-old plants in ex vitro conditions exhibited downregulation across all three examined genes– approximately 0.5-fold for all genes without AmPAL in line 36 with p ≤ 0.01). Previous studies have shown that polyploidy can lead to immediate and extensive changes in gene expression, often resulting in the silencing or unequal expression of duplicated genes (Adams and Wendel 2005b). These changes in gene expression can occur as early as the first generation of neo-polyploids and vary between tissues, affecting the stability of gene expression. Factors such as the activation of dormant transpositions and processes such as deacetylation, methylation, and changes in histones and chromatin structure influence gene silencing (Chen and Ni 2006; Song and Chen 2015). In addition, small RNAs (miRNAs and siRNAs) regulate gene expression through RNA interference, which warrant further investigation to understand artificial polyploidisation in medicinal plants (Adams and Wendel 2005a).

Despite the general observation of decreased gene expression in several experimental lines subjected to in vitro polyploidisation, it is noteworthy that our study revealed cases of significantly increased relative gene expression among the examined genes. This increase in gene expression is consistent with findings from other studies showing that polyploidy can lead to differential gene expression patterns, including both up- and downregulation (Adams and Wendel 2005a; Chen and Ni 2006). For example, genotype 54 showed a strong response with all tested genes significantly upregulated (Fig. 1), mirroring observations in other species where polyploidy induced the upregulation of biosynthetic genes involved in secondary metabolite production (Mishra et al. 2010; Lin et al. 2011). Additionally, genotypes 36 and 74 showed significant upregulation of the AmI3’H gene, with a remarkable 2.8-fold increase in genotype 74 (Fig. 1), consistent with the findings of Xing et al. (2011) that polyploidy increased gene expression in biosynthesis pathways.

In mature two-year-old plants under ex vitro conditions, significant upregulation of both AmPAL and AmC4H genes was observed in genotype 16 (Fig. 2). Notably, this genotype was no longer tetraploid, which makes the sustained high expression levels particularly interesting and complex. This may reflect epigenetic, genomic, and mutational changes that persist even after the lines are no longer polyploid, leading to continued increased gene expression. This phenomenon highlights the complex and persistent effects of polyploidisation on gene regulation (Yang et al. 2011; Bharadwaj 2015). In contrast, the induced autotetraploid genotype 74 showed a significant upregulation of the AmI3′H gene and concomitant downregulation of the AmC4H gene (Fig. 2). These results highlight the complex and diverse responses of gene expression to polyploidisation, emphasising the importance of considering specific genotypic contexts and the persistence of polyploid-induced changes over time (Ghimire et al. 2016; Zhang et al. 2019). The unique case of genotype 16 suggests that the regulatory mechanisms governing gene expression after polyploidisation are multi-faceted and warrant further investigation to fully understand the underlying molecular dynamics, providing an interesting avenue for future research.

While our study focused on gene expression changes following polyploidisation induction we did not explore the heritability across subsequent generations. Additionally, our assessment of gene expression relied solely on RT-qPCR analysis, which, while informative, may not capture the full complexity of gene regulatory networks affected by polyploidisation. Incorporating complementary techniques such as RNA sequencing could provide a more comprehensive understanding of the transcriptional landscape post-polyploidisation. The consistent increase in gene expression, particularly in the absence of polyploidy, challenges general expectations and highlights the complexity of the regulatory mechanisms affected by in vitro polyploidisation. While our study observed both up- and downregulation of key biosynthetic genes, the underlying mechanisms driving these expression changes remain speculative. Future research exploring epigenetic modifications, chromatin remodelling, and transcription factor activity after polyploidisation could elucidate the molecular basis of these observed gene expression patterns. These findings emphasise the need for a nuanced understanding of the molecular dynamics involved and encourage further investigation into the underlying molecular mechanisms that maintain altered gene expression across different genotypes and developmental stages.

Lastly, implications for future research should extend beyond our immediate findings to address broader questions regarding the ecological and physiological consequences of artificial polyploidisation in A. membranaceus. Understanding how polyploidisation influences plant fitness, adaptation to environmental stressors, and interactions with symbiotic organisms will be essential for harnessing the full potential of polyploidisation in agricultural and ecological contexts.

Conclusion

In this conclusive analysis of our research on artificial in vitro polyploidisation of A. membranaceus, it is evident that this method can profoundly affect gene expression associated with metabolic pathways of the biosynthesis of calycosin and calycosin-7-O-β-D-glucoside. Despite the observed decrease in gene expression in some lines, it is intriguing that we have identified lines that show a significant increase in relative gene expression, under both in vitro and ex vitro conditions. These striking findings highlight the complexity of regulatory mechanisms affected by artificial polyploidisation in A. membranaceus. However, it is important to acknowledge several limitations of our study. The variability in polyploid stability and gene expression changes between different lines suggests that additional factors, such as epigenetic modifications and environmental conditions, may influence the outcome of polyploidisation. Furthermore, our study focused on a limited number of genes within the flavonoid biosynthetic pathway, and a broader analysis including other metabolic pathways may provide a more comprehensive understanding of the genetic and metabolic effects of polyploidisation. Future research should aim to elucidate these underlying mechanisms and assess the long-term stability of induced polyploids. Investigations into the role of small RNAs, chromatin modifications, and other epigenetic factors in regulating gene expression in polyploid plants are necessary to fully understand the complexity of polyploidisation. In addition, studies that integrate genomic, transcriptomic, and metabolomic data will be essential to identify key regulatory networks and potential targets for enhancing desired traits in polyploid plants. In order to better understand these processes and to optimise in vitro polyploidisation in medicinal plants, further comprehensive studies focusing on the molecular mechanisms influencing gene expression and metabolic pathways in different genotypic lines are essential. The practical applications of these findings extend to pharmaceutical research and agricultural practice, enabling the manipulation of gene expression to enhance the production of pharmacologically significant compounds in plants. This has the potential to improve pharmaceutical development, contribute to the refinement of agricultural strategies to improve crop yield and quality, and ultimately shape the future of both medicine and agriculture. Our findings not only advance the field of plant polyploidy research, but also highlight the potential of artificial polyploidisation as a tool for biotechnological applications, offering promising avenues for the production of high-value phytochemicals and the improvement of crop species.