The Content of Phenolic Compounds in Stevia rebaudiana (Bertoni) Plants Derived from Melatonin and NaCl Treated Seeds

Stevia is a plant with many beneficial properties. It contains not only steviol glycosides, which are used as non-caloric natural sweeteners, but also a number of metabolites with antioxidant properties. This study examined the content of both phenolic acids and flavonoids in stevia leaves as an effect of treating seeds with melatonin and conducting germination in NaCl conditions. The results of our research indicated higher amounts of phenolic acids compared to flavonoids in stevia leaves. Among these acids, isochlorogenic, rosmarinic, and chlorogenic acids were accumulated in the largest amounts, regardless of the germination conditions. For 5 and 100 µM of melatonin treatments, the content of both phenolic acids and flavonoids increased. However, in salinity conditions (50 mM NaCl), 500 µM of melatonin had the most favorable effect on the synthesis of phenolic acids. The phenolic acids in that case reached a level three-times higher than that in the samples with the same melatonin concentration but without NaCl. We also found that the content of phenolic compounds varied depending on the age of the leaves. To the best of our knowledge, this is the first study to describe the effect of melatonin and NaCl on the synthesis on phenolic acids and flavonoids in stevia.


Introduction
Stevia rebaudiana (Bertoni) is a perennial herb belonging to the Asteraceae family. The natural areas of occurrence of stevia are Paraguay and Brazil. Currently, the population of stevia in natural sites is significantly limited, mainly due to the excessive grazing of animals and plant exploitation [1]. Interest in stevia results mainly from the content of steviol glycosides (SGs), which are sweet, zero-calorie metabolites that can successfully replace traditional sugar in human diets. These compounds are derivatives of tetracyclic diterpenes, and the main part of their molecules is steviol linked by glycosidic bonds with the sugar residue [2]. About 30 SGs have been identified in stevia; these differ in the number of sugar elements, with the main sweet-tasting compounds being stevioside and rebaudioside A [3][4][5][6][7][8]. Steviol glycosides are found mainly in leaves, and their synthesis begins in plastids along the methylerythritol phosphate pathway (MEP). In leaves, the estimated concentrations of stevioside is 6.5-9.1% and rebaudioside A is 2.3-3.8% [9,10]. Smaller amounts are also found in the stem and flowers. The SG content varies depending on the stage of plant development. Angelini et al. [8] showed that the greatest amounts of SGs accumulate in the leaves during the flowering period. Steviol glycosides are not only natural sweeteners, they also have proven health-promoting properties; since they lower the glycemic index, they are especially recommended for people with diabetes. The sweetening properties of stevia have been known for a long time, but it was not until the 20th century that SGs were approved for use in the human diet by the EU Commission [11].
Although various approaches for applying MEL have been tested, there is no research related to phenolic acid and flavonoid contents in S. rebaudiana as a result of seed treatment. Therefore, this study is the first to describe the content of phenolic compounds in stevia plants derived from seeds treated with various concentrations of MEL and additionally germinated on NaCl.

The Effect of Melatonin on Phenolic Acid and Flavonoid Content
In previous research, we indicated the effect of MEL treatment of seeds on the development and growth of stevia plants as well as on the synthesis of SGs [46]. For this study, we investigated its effects on the contents of phenolic acids and flavonoids. As previously indicated, MEL could enhance the level of flavonoids and different forms of isoflavone during soybean germination [58]. Xu et al. [59] demonstrated that melatonin treatment increased the contents of total phenols (for 18 of 22 phenolic compounds), flavonoids, anthocyanins, and proanthocyanidins in berries. Ptak et al. [52] found that MEL enhanced the galanthamine biosynthesis in the in vitro culture of Leucojum aestivum, with the most promising effect shown for the concentration of 5 µM of MEL. In the experiment presented here, treatment of seeds before germination with 5 µM and 100 µM of MEL also increased the contents of total phenolic acids (5.5-fold and 5.2-fold, respectively) and total flavonoids (1.9-fold and 2.6-fold, respectively) in stevia leaves compared to those in the control. A higher concentration of MEL (500 µM) had no such beneficial effects (Figure 2a

The Effect of Melatonin on Phenolic Acid and Flavonoid Content
In previous research, we indicated the effect of MEL treatment of seeds on the development and growth of stevia plants as well as on the synthesis of SGs [46]. For this study, we investigated its effects on the contents of phenolic acids and flavonoids. As previously indicated, MEL could enhance the level of flavonoids and different forms of isoflavone during soybean germination [58]. Xu et al. [59] demonstrated that melatonin treatment increased the contents of total phenols (for 18 of 22 phenolic compounds), flavonoids, anthocyanins, and proanthocyanidins in berries. Ptak et al. [52] found that MEL enhanced the galanthamine biosynthesis in the in vitro culture of Leucojum aestivum, with the most promising effect shown for the concentration of 5 µM of MEL. In the experiment presented here, treatment of seeds before germination with 5 µM and 100 µM of MEL also increased the contents of total phenolic acids (5.5-fold and 5.2-fold, respectively) and total flavonoids (1.9-fold and 2.6-fold, respectively) in stevia leaves compared to those in the control. A higher concentration of MEL (500 µM) had no such beneficial effects (Figure 2a,b). Regardless of the concentration of MEL, the total content of phenolic acids was higher than that of flavonoids, which is consistent with the results of other studies [23,60].
In this study, the content of phenolic compounds were analyzed separately in older and younger stevia leaves. Nevertheless, we did not observe any effect of the leaf maturity stage on the total phenolic acid content. Despite this, Liu et al. [61] previously showed that younger leaves of tea plants contained twice as much phenolic acid as older ones. However, we did find that the stage of leaf development had an effect on the content of flavonoids in relation to the concentration of MEL. In older leaves, we observed the highest combined amounts of the three flavonoids (hyperoside, isoquercetin, and quercitrin) for 100 µM of MEL. In younger and control leaves, these compounds accumulated in the highest amounts for plants derived from seeds treated with 5 µM of melatonin. The highest concentration of melatonin (500 µM) used for seed treatment had no effect on the synthesis of the determined flavonoids in stevia leaves (Figure 2a,b).
Despite no differences in the qualitative composition of phenolic acids and flavonoids, on the basis of the peak area of each compound, we found that melatonin concentrations and the maturity stage of the leaves significantly affected the content of individual phenolic compounds ( Table 2). Regardless of the concentration of MEL, the total content of phenolic acids was higher than that of flavonoids, which is consistent with the results of other studies [23,60]. In this study, the content of phenolic compounds were analyzed separately in older and younger stevia leaves. Nevertheless, we did not observe any effect of the leaf maturity stage on the total phenolic acid content. Despite this, Liu et al. [61] previously showed that younger leaves of tea plants contained twice as much phenolic acid as older ones. However, we did find that the stage of leaf development had an effect on the content of flavonoids in relation to the concentration of MEL. In older leaves, we observed the highest combined amounts of the three flavonoids (hyperoside, isoquercetin, and quercitrin) for 100 µM of MEL. In younger and control leaves, these compounds accumulated in the highest amounts for plants derived from seeds treated with 5 µM of melatonin. The highest concentration of melatonin (500 µM) used for seed treatment had no effect on the synthesis of the determined flavonoids in stevia leaves (Figure 2a,b).
Despite no differences in the qualitative composition of phenolic acids and flavonoids, on the basis of the peak area of each compound, we found that melatonin concentrations and the maturity stage of the leaves significantly affected the content of individual phenolic compounds (Table 2). Table 2. Variance analysis for the content of phenolic compounds with regard to melatonin concentration and the maturity of stevia leaves. The analyses were carried out separately for plants obtained from seeds treated with melatonin and germinated under control conditions (−NaCl) and for plants obtained from melatonin treated seeds and germinated on a medium with the addition of NaCl (+NaCl).  Table 2. Variance analysis for the content of phenolic compounds with regard to melatonin concentration and the maturity of stevia leaves. The analyses were carried out separately for plants obtained from seeds treated with melatonin and germinated under control conditions (−NaCl) and for plants obtained from melatonin treated seeds and germinated on a medium with the addition of NaCl (+NaCl).  Padda and Picha [62] also described the different levels of certain phenolic acids in sweet potato leaves at different stages of maturity. However, no studies have shown the effect of melatonin on the content of individual phenolic compounds and flavonoids. In our research, the highest phenolic acid concentrations were found for isochlorogenic, rosmarinic, and chlorogenic acids, and the highest flavonoid concentration was for quercitrin. The average content of each of these compounds was 3.18, 2.38, and 0.62 of dry weight (DW) for isochlorogenic, rosmarinic, and chlorogenic acids, respectively, and 0.58 mg/g for quercitrin (Table 3).
Regarding isochlorogenic and chlorogenic acids, their concentrations were significantly higher in younger leaves compared to older ones in all tested samples with 5 µM of MEL. The contents of chlorogenic acid was 1.9-fold and isochlorogenic acid was 3.3-fold higher compared to those in the control. With an increase in MEL concentration, the content of both acids decreased slightly for both older and younger leaves. Our research confirms the literature indicating that the main phenolic acids in stevia leaves are isochlorogenic and chlorogenic acid [22,31]. However, this is the first study to show the influence of MEL on the biosynthesis of these acids. Regarding rosmarinic acid, when MEL was used for the seed treatment before germination, it significantly increased the content in older leaves compared to that in younger leaves at all tested concentrations; however, this effect was best seen with lower melatonin concentrations (Table 3). Among the identified flavonoids, quercitrin was accumulated in the greatest amount. In that case, the highest concentration (about 2.7-times higher than that in the control) was detected in older leaves under 100 µM of MEL (0.87 mg/g of DW). However, 500 µM of MEL significantly decreased the quercitrin concentration, specifically in younger leaves ( Table 3).
The effect of MEL on plants seems to be long-term [63]. Therefore, a specific concentration of MEL used for seed treatment before germination may be a good stimulator of the biosynthesis of phenolic compounds in S. rebaudiana leaves. For the extraction of specific compounds, attention should be paid to the use of leaves in the appropriate maturity stage. Table 3. The content of individual phenolic acids and flavonoids (mg/g of DW) in stevia leaves obtained from melatonin and NaCl treated seeds. Results are presented as a mean of three replications (n = 3) ± SD. Different letters in each column (separately for −NaCl and +NaCl) indicate a significant difference of p < 0.05 according to Duncan's test; for mean value of each compound, homogeneous groups were designated separately and marked with capital letters. 0 MEL, 5 MEL, 100 MEL, 500 MEL-different concentrations of melatonin used for seed treatment: 0 µM, 5 µM, 100 µM, and 500 µM, respectively; O-older leaves, Y-younger leaves. Monitored compounds: neochlorogenic acid (NCA); protocatechuic acid (PCA); chlorogenic acid (CGA); cryptochlorogenic acid (CCGA); isoferulic acid (IFA); isochlorogenic acid (ICQA); rosmarinic acid (RA); hyperoside (HYP); isoquercetin (IQC); quercitrin (QR).

The Effect of Melatonin on Phenolic Acid and Flavonoid Content under NaCl
Salinity is one of the most significant abiotic factors that negatively influences plant efficiency and the production of many crops. Around 19.5% of irrigated agricultural land is considered to be saline [64]. Stevia plantlet growth decreases as salinity increases [65][66][67]. Our previous study also showed that stevia germination and growth is impacted by salt stress [46]. Although a low level of salinity (50 mM) had a positive impact on the morphological parameters, 150 mM of sodium chloride (NaCl) exerted an unfavorable effect on seedling development, resulting in growth inhibition and even death [46]. Our previous study also indicated that 50 mM of NaCl increased the stevioside content but decreased rebaudioside A [46]. This finding was in agreement with other research showing that low levels of NaCl increased SG, chlorophyll a, carotenoids, and total sugars [68,69]. However, no previous studies have shown the effect of NaCl on the content of individual phenolic compounds in stevia. In this study, we found that NaCl at a concentration of 50 mM increased the contents of both phenolic acids and total flavonoids (Figure 2c,d). Similarly, Zhu et al. [70] found that in Taraxacum officinale, the contents of chlorogenic and isochlorogenic acids were higher under salt stress conditions (≤1 g kg −1 ) compared to those in the control, with the opposite effect observed at higher concentrations of NaCl. An interesting solution limiting the negative impact of NaCl is the use of MEL. Melatonin is a molecule with a wide spectrum of activity, and in many species, it has been proven to have a beneficial effect on growth under stress conditions, including salt stress. In Malus hupehensis L., exogenous melatonin modulated the flavonoid content in response [71]. Bistgani et al. [72] found that foliar application of MEL improved the total phenolic compounds in Thymus daenensis L. leaves under salinity stress. In this study, we observed that the treatment of seeds with 5 or 500 µM of MEL before germination on 50 mM of NaCl affected the total phenolic acid concentration in stevia leaves. However, 5 µM of MEL had the best effect on younger leaves, while 500 µM of MEL had the best effect on older leaves (Figure 2c). Additionally, the highest sum of flavonoids was observed for 500 µM of MEL. We observed that 500 µM of MEL increased the total flavonoid content in older leaves, while 5 µM and no melatonin (0 MEL) increased the content in younger leaves (Figure 2d). In a previous study [73,74], melatonin affected the concentrations of the flavonoids and total phenolic acids. In our research, MEL also had a significant effect on the content of individual phenolic compounds under NaCl conditions ( Table 2). The highest accumulations were recorded for isochlorogenic, rosmarinic, and chlorogenic acids (average contents of 3.46, 2.78, and 0.81 mg/g of DW, respectively) as well as quercitrin (0.72 mg/g of DW). The highest contents of rosmarinic and isochlorogenic acids were recorded in older leaves obtained from seeds exposed to 500 µM of MEL (9.28 and 7.74 mg/g of the DW, respectively), which were about 13-times and 10-times higher, respectively, than their amounts in the older leaves of control plants ( Table 3). The results described above indicate that in the presence of NaCl, melatonin regulates the content of individual phenolic compounds in a slightly different way compared to conditions without NaCl.

The Effect of Melatonin and NaCl on the Accumulation of Phenolic Compounds
To examine the linkage between the effects of melatonin and NaCl in terms of the studied biochemical compounds, Ward's [75] hierarchical clustering was applied. Based on the average content of the identified phenolic compounds, two clusters formed on the dendrogram (Figure 3). The similarities shown in both clusters indicate the important roles of both melatonin and NaCl in inducing a specific response in secondary metabolism. At a lower melatonin concentration (5 µM) and in the absence of melatonin (0 MEL), the presence of NaCl had no significant effect on the content of phenolic compounds in stevia plants. However, the presence of NaCl in the germination medium had an effect on seeds treated with melatonin at higher concentrations (100 and 500 µM); samples that germinated in the presence and absence of NaCl in the medium were located in different clusters (Figure 3). This may indicate that in addition to the adverse effects of NaCl observed in many plant species, higher concentrations of melatonin may also induce a stress response in stevia plants [76]. roles of both melatonin and NaCl in inducing a specific response in secondary metabolism. At a lower melatonin concentration (5 µM) and in the absence of melatonin (0 MEL), the presence of NaCl had no significant effect on the content of phenolic compounds in stevia plants. However, the presence of NaCl in the germination medium had an effect on seeds treated with melatonin at higher concentrations (100 and 500 µM); samples that germinated in the presence and absence of NaCl in the medium were located in different clusters (Figure 3). This may indicate that in addition to the adverse effects of NaCl observed in many plant species, higher concentrations of melatonin may also induce a stress response in stevia plants [76].  Principal component analysis (PCA) showed that the first component (PC1) explained nearly 53% of the variation, while the second component (PC2) explained about 23% of the variation. PC1 was significantly affected by almost all variables except the protocatechuic acid content. In contrast, PC2 was strongly negatively influenced by the content of protocatechuic acid, cryptochlorogenic acid, quercitrin, and isoquercetin. However, the contents of neochlorogenic acid and hyperoside had a significant positive effect on PC2 (Figure 4). The position of the loading vectors for the individual variables made it possible to indicate several relationships. A significant positive correlation was found between the contents of neochlorogenic acid and hyperoside. Additionally, a significant relationship was found between the contents of isochlorogenic, isoferulic, and rosmarinic acids. A significant relationship was also seen between the contents of cryptochlorogenic acid and quercitrin and the content of isoquercetin. The biplot chart indicates the relationship between the studied treatments and the content of the analyzed phenolics. From the distribution of the data, it can be concluded that the control, characterized by the absence of melatonin treatment and NaCl, was not related to the variation in any of the analyzed acids. In contrast, the plants treated only with NaCl were characterized by a particular increase in protocatechuic acid content not observed in other treatments. Melatonin at a dose of 5 µM was specifically associated with an increase in the content of neochlorogenic acid and a higher content of hyperoside. A similar distribution of phenolic compound contents was found in the plants obtained from seeds treated with MEL at a dose of 100 µM and at a dose of 500 µM in combination with NaCl, especially for chlorogenic, isochlorogenic, isoferulic, and rosmarinic acids (Figure 4). acids. In contrast, the plants treated only with NaCl were characterized by a particular increase in protocatechuic acid content not observed in other treatments. Melatonin at a dose of 5 µM was specifically associated with an increase in the content of neochlorogenic acid and a higher content of hyperoside. A similar distribution of phenolic compound contents was found in the plants obtained from seeds treated with MEL at a dose of 100 µM and at a dose of 500 µM in combination with NaCl, especially for chlorogenic, isochlorogenic, isoferulic, and rosmarinic acids (Figure 4).

Plant Material
The stevia plants were obtained from seeds incubated in an aqueous solution of MEL prior to germination.
The seeds were surface-sterilized according to Simlat et al. [77] before treatment. Three concentrations of MEL were used: 5 µM (5 MEL), 100 µM (100 MEL), and 500 µM (500 MEL). For the control, the seeds were incubated in water (0 MEL). After incubation, half of the seeds were germinated in in vitro conditions on agar gel and the other half on agar gel supplemented with 50 mM of NaCl. After three weeks of germination experiment, seedlings were transferred to the Murashige and Skoog (MS) [78] medium for one month and then well-developed plantlets were transplanted into pots (15 × 15 cm) containing soil, peat, and sand in a 3:1:1 ratio. The plants used for analysis were grown for six months in controlled conditions (25 • C, fluorescent light with intensity expressed as Photosynthetic Photon Flux Density of 320 µmol m −2 s −1 for 16 h/day, and 70% ± 5% relative humidity; Adaptis-A1000AR, Conviron, Winnipeg, MB, Canada).

Statistical Analyses
The data were reported as mean ± SD. To study the effect of MEL and NaCl on the phenolic acid and flavonoid content, factorial ANOVA (TIBCO Statistica, Palo Alto, CA, USA) was applied. To estimate the significance of differences between the means, we performed Duncan's multiple range test at a significance level of p < 0.05. To examine the linkage between the effects of MEL and NaCl in terms of the studied biochemical compounds, Ward's [75] hierarchical clustering was applied. The dendrogram was generated based on the phenolic acid and flavonoid contents. To determine the relationship between treatment and the phenolic content, the data were analyzed with PCA using R v. 3.6.1 [80].

Conclusions
Taking into account the present results as well as previously published studies on S. rebaudiana, it is reasonable to conclude that stevia is an abundant source of secondary metabolites with health benefits. The presence of bioactive compounds in stevia leaves belonging to phenolic and flavonoid groups tend to justify their medicinal properties and their application in both the food and pharmaceutical industries.
Furthermore, our studies pointed out that MEL used at the stage of seed germination seems to be a good stimulator of the biosynthesis of phenolic compounds in stevia plants, and low dose of NaCl may be an additional stimulating factor. This has a very practical aspect: MEL treated seeds can be used in the establishment of plantations for obtaining high quality plants useful for the food and pharmaceutical industries.
Although significant progress has been made in understanding the role of MEL in plants, the MEL signaling pathway under salt stress remains unclear. Melatonin's mitigation of salt stress damages can be achieved through multiple mechanisms: (1) MEL reduces excessive production of reactive oxygen species (ROS) by facilitating electron leakage as a result of increasing the efficiency of the mitochondrial electron transport chain; (2) MEL regulates ion homeostasis; (3) MEL modulates the activity of transcription factors; and (4) MEL may also mitigate cell damage and reduce the increase in osmotic pressure caused by salt stress by increasing the accumulation of the phenolic compounds and all phenolic acids.
Comprehensive and well-designed future research on the role of MEL in biosynthesis of phenolic compounds under salt stress in S. rebaudiana would be valuable. Our further research will focus on the expression of genes involved in the metabolism of phenylpropanoids and the analysis of the activity of relevant enzymes. We believe that these anal-